Tumor Immunology - Hindawi Publishing...

Tumor Immunology

Clinical and Developmental Immunology

Tumor Immunology

Copyright © 2010 Hindawi Publishing Corporation. All rights reserved.

This is a focus issue published in volume 2010 of “Clinical and Developmental Immunology.” All articles are open access articlesdistributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in anymedium, provided the original work is properly cited.


Editorial Board

B. Dicky Akanmori, GhanaScott Antonia, USAR. Baughman, USAStuart Berzins, AustraliaBengt Bjorksten, SwedenK. Blaser, SwitzerlandFederico Bussolino, ItalyRobert B. Clark, USAMario Clerici, ItalyRobert E. Cone, USABernhard Fleischer, GermanyRichard L. Gallo, USARonald B. Herberman, USAD. Craig Hooper, USAH. Inoko, JapanDavid Kaplan, USAW. Kast, USA

Taro Kawai, JapanHiroshi Kiyono, JapanDennis Klinman, USAGuido Kroemer, FranceYang Liu, USAEnrico Maggi, ItalyStuart Mannering, AustraliaEiji Matsuura, JapanC. J. M. Melief, The NetherlandsJiri Mestecky, USAC. Morimoto, JapanBernhard Moser, UKHiroshi Nakajima, JapanT. Nakayama, JapanGraham Ogg, UKG. Opdenakker, BelgiumI. Pastan, USA

C. D. Pauza, USABerent Prakken, The NetherlandsNima Rezaei, IranClelia M. Riera, ArgentinaCharles R. Rinaldo, USALuigina Romani, ItalyB. T. Rouse, USAE. M. Shevach, USAS. Sozzani, ItalyGeorge B. Stefano, USAHelen Su, USAAlain Tedgui, FranceBan-Hock Toh, AustraliaJ. F. Urban, USAYvette Van Kooyk, The NetherlandsY. Yoshikai, Japan

Contents

The Initial Immune Reaction to a New Tumor Antigen Is Always Stimulatory and Probably Necessary forthe Tumor’s Growth, Richmond T. PrehnVolume 2010, Article ID 851728, 5 pages

Mycobacterium bovis Bacillus Calmette-Guerin-Induced Macrophage Cytotoxicity against BladderCancer Cells, Yi Luo and Matthew J. KnudsonVolume 2010, Article ID 357591, 6 pages

Regulation of Antitumor Immune Responses by the IL-12 Family Cytokines, IL-12, IL-23, and IL-27,Mingli Xu, Izuru Mizoguchi, Noriko Morishima, Yukino Chiba, Junichiro Mizuguchi,and Takayuki YoshimotoVolume 2010, Article ID 832454, 9 pages

Potential Target Antigens for a Universal Vaccine in Epithelial Ovarian Cancer, Renee Vermeij,Toos Daemen, Geertruida H. de Bock, Pauline de Graeff, Ninke Leffers, Annechien Lambeck, Klaske A. tenHoor, Harry Hollema, Ate G. J. van der Zee, and Hans W. NijmanVolume 2010, Article ID 891505, 8 pages

Vaccines and Immunotherapeutics for the Treatment of Malignant Disease, Joel F. Aldrich,Devin B. Lowe, Michael H. Shearer, Richard E. Winn, Cynthia A. Jumper, and Ronald C. KennedyVolume 2010, Article ID 697158, 12 pages

Glucocorticoid-Induced TNFR-Related (GITR) Protein and Its Ligand in Antitumor Immunity:Functional Role and Therapeutic Modulation, Theresa Placke, Hans-Georg Kopp, and Helmut Rainer SalihVolume 2010, Article ID 239083, 10 pages

Unexpected High Response Rate to Traditional Therapy after Dendritic Cell-Based Vaccine in AdvancedMelanoma: Update of Clinical Outcome and Subgroup Analysis, Laura Ridolfi, Massimiliano Petrini,Laura Fiammenghi, Anna Maria Granato, Valentina Ancarani, Elena Pancisi, Emanuela Scarpi,Massimo Guidoboni, Giuseppe Migliori, Stefano Sanna, Francesca Tauceri, Giorgio Maria Verdecchia,Angela Riccobon, and Ruggero RidolfiVolume 2010, Article ID 504979, 9 pages

Overview of Cellular Immunotherapy for Patients with Glioblastoma, Elodie Vauleon, Tony Avril,Brigitte Collet, Jean Mosser, and Veronique QuillienVolume 2010, Article ID 689171, 18 pages

Clinicopathologic Significance of HIF-α, CXCR4, and VEGF Expression in Colon Cancer, Yugang Wu,Min Jin, Huanbai Xu, Zhang Shimin, Songbing He, Liang Wang, and Yanyun ZhangVolume 2010, Article ID 537531, 10 pages

The SSX Family of Cancer-Testis Antigens as Target Proteins for Tumor Therapy, Heath A. Smith andDouglas G. McNeelVolume 2010, Article ID 150591, 18 pages

Tumor Antigen Cross-Presentation and the Dendritic Cell: Where it All Begins?, Alison M. McDonnell,Bruce W. S. Robinson, and Andrew J. CurrieVolume 2010, Article ID 539519, 9 pages

CpG Oligodeoxynucleotides Enhance the Efficacy of Adoptive Cell Transfer Using Tumor InfiltratingLymphocytes by Modifying the Th1 Polarization and Local Infiltration of Th17 Cells, Lin Xu,Chunhong Wang, Zhenke Wen, Ya Zhou, Zhongmin Liu, Yongjie Liang, Zengguang Xu, and Tao RenVolume 2010, Article ID 410893, 9 pages

Regulation of Tumor Immunity by Tumor/Dendritic Cell Fusions, Shigeo Koido, Sadamu Homma,Eiichi Hara, Yoshihisa Namiki, Akitaka Takahara, Hideo Komita, Eijiro Nagasaki, Masaki Ito,Toshifumi Ohkusa, Jianlin Gong, and Hisao TajiriVolume 2010, Article ID 516768, 14 pages

NGcGM3 Ganglioside: A Privileged Target for Cancer Vaccines, Luis E. Fernandez, Mariano R. Gabri,Marcelo D. Guthmann, Roberto E. Gomez, Silvia Gold, Leonardo Fainboim, Daniel E. Gomez,and Daniel F. AlonsoVolume 2010, Article ID 814397, 8 pages

Dendritic Cell-Based Immunotherapy for Prostate Cancer, Hanka Jahnisch, Susanne Fussel,Andrea Kiessling, Rebekka Wehner, Stefan Zastrow, Michael Bachmann, Ernst Peter Rieber,Manfred P. Wirth, and Marc SchmitzVolume 2010, Article ID 517493, 8 pages

Harnessing the Effect of Adoptively Transferred Tumor-Reactive T Cells on Endogenous (Host-Derived)Antitumor Immunity, Yolanda Nesbeth and Jose R. Conejo-GarciaVolume 2010, Article ID 139304, 11 pages

B7-H3 and Its Role in Antitumor Immunity, Martin Loos, Dennis M. Hedderich, Helmut Friess,and Jorg Kleeff

Volume 2010, Article ID 683875, 7 pages

Changes of Immunological Profiles in Patients with Chronic Myeloid Leukemia in the Course ofTreatment, Zuzana Humlova, Hana Klamova, Ivana Janatkova, Karin Malıckova, Petra Kralıkova,Ivan Sterzl, Zdenek Roth, Eva Hamsıkova, and Vladimır VonkaVolume 2010, Article ID 137320, 17 pages

DNA Vaccination: Using the Patient’s Immune System to Overcome Cancer, Georg Eschenburg,Alexander Stermann, Robert Preissner, Hellmuth-Alexander Meyer, and Holger N. LodeVolume 2010, Article ID 169484, 14 pages

Gene Carriers and Transfection Systems Used in the Recombination of Dendritic Cells for EffectiveCancer Immunotherapy, Yu-Zhe Chen, Xing-Lei Yao, Yasuhiko Tabata, Shinsaku Nakagawa,and Jian-Qing GaoVolume 2010, Article ID 565643, 12 pages

Theoretical Modeling Techniques and Their Impact on Tumor Immunology, Anna Lena Woelke,Manuela S. Murgueitio, and Robert PreissnerVolume 2010, Article ID 271794, 11 pages

Prognostic and Diagnostic Value of Spontaneous Tumor-Related Antibodies, Sebastian Kobold,Tim Luetkens, Yanran Cao, Carsten Bokemeyer, and Djordje AtanackovicVolume 2010, Article ID 721531, 8 pages

Focus on Adoptive T Cell Transfer Trials in Melanoma, Liat Hershkovitz, Jacob Schachter,Avraham J. Treves, and Michal J. BesserVolume 2010, Article ID 260267, 11 pages

T Cell-Tumor Interaction Directs the Development of Immunotherapies in Head and Neck Cancer,A. E. Albers, L. Strauss, T. Liao, T. K. Hoffmann, and A. M. KaufmannVolume 2010, Article ID 236378, 14 pages

Immunotherapy for Renal Cell Carcinoma, Momoe Itsumi and Katsunori TatsugamiVolume 2010, Article ID 284581, 8 pages

Long-Term Follow-Up of HLA-A2+ Patients with High-Risk, Hormone-Sensitive Prostate CancerVaccinated with the Prostate Specific Antigen Peptide Homologue (PSA146-154), Supriya Perambakam,Hui Xie, Seby Edassery, and David J. PeaceVolume 2010, Article ID 473453, 11 pages

Archaeosome Adjuvant Overcomes Tolerance to Tumor-Associated Melanoma Antigens InducingProtective CD8+ T Cell Responses, Lakshmi Krishnan, Lise Deschatelets, Felicity C. Stark, Komal Gurnani,and G. Dennis SprottVolume 2010, Article ID 578432, 13 pages

Immune Response in Ovarian Cancer: How Is the Immune System Involved in Prognosis and Therapy:Potential for Treatment Utilization, Nikos G. Gavalas, Alexandra Karadimou, Meletios A. Dimopoulos,and Aristotelis BamiasVolume 2010, Article ID 791603, 15 pages

Engagement of the Mannose Receptor by Tumoral Mucins Activates an Immune Suppressive Phenotypein Human Tumor-Associated Macrophages, P. Allavena, M. Chieppa, G. Bianchi, G. Solinas, M. Fabbri,G. Laskarin, and A. MantovaniVolume 2010, Article ID 547179, 10 pages

Identification of Two Novel HLA-A∗0201-Restricted CTL Epitopes Derived from MAGE-A4,Zheng-Cai Jia, Bing Ni, Ze-Min Huang, Yi Tian, Jun Tang, Jing-Xue Wang, Xiao-Lan Fu, and Yu-Zhang WuVolume 2010, Article ID 567594, 7 pages

Tumor Antigen-Dependent and Tumor Antigen-Independent Activation of Antitumor Activity in T Cellsby a Bispecific Antibody-Modified Tumor Vaccine, Philippe Fournier and Volker SchirrmacherVolume 2010, Article ID 423781, 12 pages

Monoclonal Antibodies for Non-Hodgkin’s Lymphoma: State of the Art and Perspectives, Giulia Motta,Michele Cea, Eva Moran, Federico Carbone, Valeria Augusti, Franco Patrone, and Alessio NencioniVolume 2010, Article ID 428253, 14 pages

Immunotherapy of Brain Cancers: The Past, the Present, and Future Directions, Lisheng Ge, Neil Hoa,Daniela A. Bota, Josephine Natividad, Andrew Howat, and Martin R. JadusVolume 2010, Article ID 296453, 19 pages

Immune Suppression in Head and Neck Cancers: A Review, Anaelle Duray, Stephanie Demoulin,Pascale Hubert, Philippe Delvenne, and Sven SaussezVolume 2010, Article ID 701657, 15 pages

Hindawi Publishing CorporationClinical and Developmental ImmunologyVolume 2010, Article ID 814397, 8 pagesdoi:10.1155/2010/814397

Review Article

NGcGM3 Ganglioside: A Privileged Target for Cancer Vaccines

Luis E. Fernandez,1 Mariano R. Gabri,2 Marcelo D. Guthmann,3 Roberto E. Gomez,3

Silvia Gold,4 Leonardo Fainboim,5 Daniel E. Gomez,2 and Daniel F. Alonso2

1 Vaccine Department, Center of Molecular Immunology, Havana 11600, Cuba2 Laboratory of Molecular Oncology, Quilmes National University, Roque Saenz Pena 352, Bernal B1876BXD Buenos Aires, Argentina3 Elea Laboratories, C1417AZE Buenos Aires, Argentina4 Chemo-Romikin, C1061ABC Buenos Aires, Argentina5 Division of Immunogenetics, Jose de San Martın Clinics Hospital, University of Buenos Aires, C1120AAF Buenos Aires, Argentina

Correspondence should be addressed to Daniel F. Alonso, [email protected]

Received 12 July 2010; Accepted 24 September 2010

Academic Editor: Taro Kawai

Copyright © 2010 Luis E. Fernandez et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Active specific immunotherapy is a promising field in cancer research. N-glycolyl (NGc) gangliosides, and particularly NGcGM3,have received attention as a privileged target for cancer therapy. Many clinical trials have been performed with the anti-NGc-containing gangliosides anti-idiotype monoclonal antibody racotumomab (formerly known as 1E10) and the conjugatedNGcGM3/VSSP vaccine for immunotherapy of melanoma, breast, and lung cancer. The present paper examines the role of NGc-gangliosides in tumor biology as well as the available preclinical and clinical data on these vaccine products. A brief discussion onthe relevance of prioritization of cancer antigens in vaccine development is also included.

1. Ganglioside-Based Cancer Vaccines

The field of cancer vaccines definitively changed after April2010 when the US Food and Drug Administration (FDA)approved Provenge (sipuleucel-T), a vaccine for advancedprostate cancer patients [1]. For the first time sipuleucel-T convincingly increased overall survival by about fourmonths in a randomized Phase III trial conducted in 512patients. While this immunotherapeutic agent is relativelyinconvenient as a personalized vaccine, it seems that selectingthe recombinant version of the prostatic acid phosphatase(PAP)—expressed in 95% of prostatic tumor cells—asantigen was critical [2], reaffirming target selection as a keyfeature in cancer vaccine design.

Reasoning in the same way, gangliosides, a broad familyof structurally related glycolipids, were firstly suggested aspotential targets for cancer immunotherapy [3, 4] based ontheir higher abundance in tumors when compared with thematched normal tissues. Disappointingly, at the beginningof the present century the failure of the best developedganglioside-based cancer vaccine for that time, GMK [5],irradiated an unfavorable risky perception to all the projects

in the field, even those not related with the target antigen, theGM2 ganglioside.

Nevertheless, in these days ganglioside cancer vacci-nologists were facing a kind of “polarization” of posi-tions in view of new facts. Eggermont et al. reported anearlier stop of the Phase III GMK vaccine clinical trialin stage II melanoma patients because the IndependentData Monitoring Committee detected an inferior survivalrate for the vaccine arm [6]. While difficult to interpretwith respect to potential detrimental effects, this result wasconsidered as a significant setback for ganglioside-basedspecific immunotherapy in melanoma. In the opposite side,a more optimistic outlook of gangliosides as targets for activeimmunotherapy of cancer came from the recent work byCheever et al., who reported in 2009 the National CancerInstitute pilot project for prioritization of cancer antigens[7]. From the selected 75 representative antigens, 4 weregangliosides (GD2, GD3, fucosyl-GM1, and N-acetyl-GM3),ranking between positions 12 and 48.

As an overall, at the present moment only two N-glycolyl (NGc) ganglioside-based vaccines, the focus of thispaper, are currently tested in Phase III clinical trials [8].

2 Clinical and Developmental Immunology

Normal tissue Tumor tissue

Normal cell Tumor cell

Gangliosides

NAc

Golgi

CMAH

CMAH mRNA

CMAH gene

Alu

CMAH mRNA

CMAH

CMAH gene

Alu

Golgi

Gangliosides

NGc

DietaryNGc

DietaryNGc

Vessel

Figure 1: NGc gangliosides in human tumor biology. Although NGcGM3 is practically undetectable in healthy human tissues as a result ofan Alu-mediated inactivation of the gene, the ganglioside is highly expressed in several human cancer cells presumably due to incorporationof dietary NGc.

These include racotumomab (formerly known as 1E10)and NGcGM3/VSSP vaccine (Table 1). Racotumomab isan anti-idiotype murine monoclonal antibody (mAb) toNGc-containing gangliosides. An anti-idiotype mAb, such asracotumomab, is the mirror image of the original antibodyformed against specific surface antigens. Thus, anti-idiotypeantibodies can act as antigens, inducing a response againstthe original antigen. On the other hand, the NGcGM3/VSSPvaccine results from the conjugation of the ganglioside intovery small size proteoliposomes (VSSP) derived from N.meningitidis.

2. NGc Gangliosides in Tumor Biology

The most common sialic acids in mammals are N-acetylneuraminic acid and N-glycolylneuraminic acid, usu-ally found as terminal constituents of different membraneglycoconjugates such as the GM3 ganglioside. The onlystructural difference between them consists of a single oxy-gen atom at the C-5 position of N-glycolylneuraminic acid,

catalyzed by the cytidine monophospho-N-acetylneuraminicacid hydroxylase (CMAH) [9]. In contrast to most mammals,including our closest relatives: the great apes, NGc ispractically undetectable in healthy human tissues and fluids[10], since human cells lack the presence of CMAH [11]. It isknown that this absence is due to the loss of a 92-bp segmentin the exon 6, which results in a frameshift mutation of thehuman gene by an Alu-mediated inactivation [12, 13] dated2.5–3 million years ago, prior to brain expansion duringhuman evolution [14, 15].

It is noteworthy that the monosialic acid gangliosideNGcGM3 is highly expressed in several human cancer cells[16]. Although initially it was suggested that NGc couldbe expressed in human tissues by an alternative metabolicpathway [17], nowadays plenty of evidence suggests that thepresence of this sialic acid in human cancer is the resultof the metabolic incorporation of dietary NGc [18, 19], asillustrated in Figure 1. We reported that cultured mousetumor cells lacking CMAH expression are able to processand incorporate NGc from different sources such as bovineserum, NGc-rich mucins, or purified N-glycolylneuraminic

Clinical and Developmental Immunology 3

Table 1: N-glycosylated (NGc) ganglioside-based cancer vaccines.

Vaccine Product description Target antigen Potential indications Clinical Phase

Racotumomab Anti-idiotype murinemAb (1E10 antibody)

NGc-containinggangliosides

Lung cancerBreast cancerPediatric tumors?

Ongoing Phase III

NGcGM3/VSSP

Gangliosideconjugated withbacterialproteoliposomes

NGcGM3 gangliosideBreast cancerMelanoma Sarcoma?

Ongoing Phase III

acid, thus, promoting the metastatic phenotype [20]. More-over, genetically modified mice expressing a human-likeCMAH mutation showed no endogenous NGc, as in humans[21].

The significance of NGc overexpression in human canceris still under investigation. Taking into consideration that ananti-NGc antibody response was detected in several cancerpatients, Varki [22] recently hypothesized that antibody-mediated inflammation could facilitate tumor progression.However, it is accepted that high titers of these antibodiescan kill tumor cells [22]. In addition, experimental resultsobtained by de Leon et al. indicated that growth-stimulatingfeatures of the NGc on tumor cells can be explained byimmune system down modulation [23].

3. Preclinical Data

Our team analyzed the antitumor activity and the preclinicaltoxicity of the mAb racotumomab and the NGcGM3/VSSPvaccine using different animal models. Considering that themost important feature of an anti-idiotype mAb (Ab2) isits biological effect, racotumomab was evaluated in two syn-geneic murine tumor models, the F3II mammary carcinoma(BALB/c mice) and B16 melanoma (C57BL/6 mice). Bothcell lines are positive for the idiotype mAb P3 (Ab1), whichspecifically reacts with NGc-containing gangliosides on cellsurface [24, 25]. In BALB/c mice, vaccination with severalintraperitoneal doses, at 14-day intervals of racotumomabcoupled to keyhole limpet hemocyanin in Freund’s adjuvant,significantly reduced subcutaneous tumor growth of F3IImammary carcinoma cells and the formation of spontaneouslung metastases [24]. Similarly, intravenous administrationof uncoupled racotumomab, as a biological response mod-ifier, dramatically inhibited metastatic lung colonization byB16 melanoma cells in C57BL/6 mice [24].

Vaccination with aluminum hydroxide-precipitatedracotumomab induced antimetastatic effects in the 3LL-D122 Lewis lung carcinoma, a poorly immunogenic andhighly metastatic model in C57BL/6 mice [37]. The effectwas associated to T cell infiltration, enhancement of tumorapoptosis, and reduction of new blood vessels formation inlung nodules. The 3LL-D122 lung carcinoma is an antigen-positive, validated model for the NGcGM3 ganglioside. Themodel evidenced an increased expression of such specificantigen from primary tumors to metastatic lesions [38].Immunization with the NGcGM3/VSSP vaccine, preparedeither with synthetic or natural source-derived ganglioside,

showed similar immunogenicity profiles and antitumoreffects in the 3LL-D122 model [38].

Racotumomab also demonstrated a potent antitumoreffect in combination with chemotherapy in preclinicalstudies, providing a rationale for chemo-immunotherapycombinations in solid cancers. Administration of low-dosecyclophosphamide together with subcutaneous immuniza-tion with racotumomab in alum significantly reduced F3IItumor growth [25]. The antitumor response was comparableto that obtained with standard high-dose chemotherapy insuch breast cancer model, but without overt signs of toxic-ity. Interestingly, combinatory chemo-immunotherapy pro-moted CD8+ lymphocyte tumor infiltration and increasedtumor apoptosis [25].

Ganglioside immunotherapies with racotumomab andNGcGM3/VSSP vaccine were well tolerated in animals [24,38]. In preclinical toxicology studies, the immunizationprotocol did not affect body weight gain, food and water con-sumption or induce other signs of overt toxicity in murinemodels. Subacute toxicity after continuous daily treatmentwas expressed by an excessive activation of extramedullarymyelopoiesis in the spleen and liver in all mice and astrong inflammatory reaction in the lungs, showing denseneutrophil infiltrates in the interalveolar septa [24].

4. Expression of NGc in Human Tumors

Tumor-specific expression of NGc-containing gangliosidesin some human tumors suggests that the induction ofan effective immune response against these antigens maybe useful for patients with antigen-positive tumors. Theganglioside NGcGM3 has been described in human neo-plasms, including breast carcinoma [33, 34] and melanoma[31], but is usually not detected in normal human cells.This fact defines NGcGM3 as an interesting target forimmunotherapy.

As described by Tangvoranuntakul et al. [34] using amonospecific antibody against NGc, staining showed celltype-specific reactivity in adult human tissues. The overallpattern of expression was summarized as prominent insecretory epithelia and associated secretions and present inmany blood vessels. In addition, in fetal tissues NGc canbe detected in epithelial cells or secretions as well as theplacental villus blood vessels [34].

van Cruijsen et al. [35] assessed the possible associationof NGcGM3 expression with angiogenesis in lung cancer.They examined 176 samples of nonsmall cell lung cancer


Table 2: Relevant characteristics of NGcGM3 ganglioside as a cancer antigen, according to the antigen prioritization criteria described byCheever et al. [7].

Criteria Data on NGcGM3

Therapeutic function Clinical data showing that a vaccine-induced clinical responses in at least a small numberof patients [26–28]

Immunogenicity T cell [29] and antibody [26, 27, 29, 30] responses elicited in clinical trials, spontaneousantibody observed in some patients [31, 32]

Oncogenicity Increased expression in adult [31, 33–35] and pediatric [36] solid tumors, to bedetermined a clear association with oncogenic process or tissue differentiation

Specificity Overexpressed in cancer with little or no expression in normal adult tissues [34]

Expression level and % positive cells Highly expressed on most cancer cells in patients designated for treatment [31, 33, 35, 36]

Stem cell expression Expression on most cancer cells [35, 36] but without information about putative stemcells

No. of patients with antigen-positive cancers High level of expression in >80% of patients with a particular tumor type [35, 36]

No. of antigen epitopes Short antigenic segment with one or few epitopes [31]

Cellular location of antigen expression Expressed on the cell surface [31, 33, 35, 36] with little or no circulating antigen [10, 11]

(NSCLC) by immunohistochemistry in tissue microarrayand found that NGcGM3 is widely expressed in morethan 90% of the cases. Microvessel density, as determinedby CD34 staining, was lower in NSCLC tissues with highNGcGM3 expression, suggesting that the presence of theganglioside may favor an antiangiogenic response. Moreover,based on the expression of CD83 which is a marker of maturedendritic cells, NGcGM3 appeared to be involved in tumor-induced dendritic cell suppression [35].

More recently, Scursoni et al. [36] reported for the firsttime the expression of NGcGM3 in a pediatric solid tumor.They detected the ganglioside in 88% of the cases of Wilms’tumor (nephroblastoma), using the specific anti-NGcGM3mAb 14F7 and a peroxidase-labeled polymer conjugatedto secondary antibodies on postchemotherapy samples.Wilms tumor is considered an “embryonic tumor” of thekidney, being a mimicry of various elements in normal orabnormal nephrogenesis and presenting a diverse spectrumof histologic appearance. In this regard, Wilms tumor givesthe unique opportunity to learn about the expression ofNGc gangliosides in diverse transformed cell lineages, com-prising epithelial, stromal, and blastemal elements [39]. Thestrongest expression was found in the epithelial componentof Wilms’ tumor, and the lower percentage of positive tumorcells was observed in the stromal subtype [36]. Similar resultswere obtained in a preliminary study with P3, a less-specificmAb that recognizes different NGc-containing gangliosidesand sulfatides, including NGcGM3 [40]. More than 70% ofWilms tumors showed a positive staining for NeuGc residuesusing the P3 antibody [36].

5. Immunological Response to NGc in Humans

Targeting ganglioside antigens has been a matter of concerndue to the possibility of inducing autoimmune responses.Indeed, in most neuropathies of immunological origin,endogenous gangliosides have been shown to be the tar-get of the autoimmune reactions [41–44]. Antigangliosideantibodies may also affect nonneural tissues, as occuring

in systemic lupus erythematosus [45], rheumatoid arthritis,or Sjogren’s syndrome [41]. Antibodies reactive to NGc-containing glycolipids have been found to be induced inpatients after repeated transfusions with sera from otherspecies [46], in rheumatoid arthritis [47], and also inmelanoma patients [32]. Melanoma cells express NGc-containing gangliosides, including NGcGM3 [31], and natu-ral anti-NGc antibodies are increased in melanoma patients.Most interestingly, a significantly higher level of anti-NGcantibodies was demonstrated in those patients who were freeof disease more than 5 years after surgery than in those whorelapsed within 2 years [32].

In spite of what could be expected, no induction ofdetrimental autoimmune reactions have been described overdecades of clinical development of cancer vaccines targetingendogenous gangliosides such as GM2, GD2, and GD3.In this respect, the heterophilic nature of NGcGM3 isconsidered an additional asset of this tumor antigen, since, asreviewed in the following sections, the absence of significantexpression in normal tissue allows for increased immuneresponses to immunization while precluding self-targetedreactions.

Breast cancer patients were treated with a regime of 5biweekly IM injections of NGcGM3/VSSP vaccine followedby monthly boosters. Anti-NGcGM3 IgM and IgG responseswere detected in all patients who completed the first 5injections, collectively termed as “induction phase”. The timecourse of antibody production showed an overall increaseacross the 32-week followup, and the maximal recordedtiters reached 164,000 for both IgM and IgG. The functionalrelevance of the induced antibodies was underscored by theircapacity to react against an NeuGcGM3+ murine tumorcell line and mediated complement-dependent cytotoxic-ity. Further confirming their specificity for the targetedganglioside, the induced antibodies reacted as well withhuman mammary ductal carcinoma cells without staining ofsurrounding normal tissue [26].

In a second clinical trial, 21 advanced melanoma patientsreceived the same vaccination schedule [30]. Two dose levels


were examined: 0.2 and 0.4 mg of ganglioside per dose. Theimmunogenicity of NGcGM3/VSSP was confirmed, withall evaluable patients eliciting IgM and IgG responses aftertreatment. Antibody class switching to IgA was detected aswell. Delayed-type hypersensitivity (DTH) responses werealso evaluated with ID injections of NGcGM3/VSSP. DTHresponses were observed in 46% of patients at the 0.2 mg doselevel and in 78% of patients at the 0.4 mg dose level. The gan-glioside contribution to the specificity of the hypersensitivityreactions was, however, not assessed. No clear relationshipcould be established between immunological response andthe clinical outcome, and the lower dose level was selectedfor future clinical development due to its safer toxicity profile[30].

The immune response elicited by the murine anti-idiotypic mAb racotumomab was monitored in greater detailand in additional clinical settings. The vaccination regime forracotumomab administration consisted of a 6-dose induc-tion phase followed by monthly boosters. Anti-gangliosideresponses were induced in 16/17 melanoma patients [48],16/16 breast cancer patients, and 16/20 NSCLC patients[27]. The specificity of the induced Ab3 antibodies wasassessed by adsorption and analysis of the reactivity of thenonadsorbed fraction. Adsorption with an isotype-matchedmonoclonal antibody (IgG1) abrogated a small fraction ofthe racotumomab-directed reactivity, which is an indicationof the immunodominance of the racotumomab idiotypeover the rest of the IgG1 molecule [27]. Most interestingly,adsorption of Ab3 antibodies with racotumomab preserved50 to 90% of the reactivity for NGcGM3 in the non-adsorbed fraction, suggesting that the idiotype (Id)+/antigen(Ag)+ and Id−/Ag+ specificities are present on separateantibody molecules [48]. Such Id−/Ag+ antibodies couldreflect the activation of an autologous idiotypic cascade inthe patient’s immune system [27]. The antibodies inducedby racotumomab are NGc specific. No cross-reaction wasobserved to NAcGM3 [48]. The time course of NGcGM3-specific antibodies over 50 weeks in breast cancer patientsunder racotumomab treatment showed sustained antibodytiters. Some patients had detectable ganglioside-specific IgGonly by week 30, suggesting that the extended vaccinationregime not only is undetrimental to the immune response,but is favorable for late responders as well [29].

As described for NGcGM3/VSSP vaccine, the anti-bodies induced by racotumomab treatment reacted withan NGcGM3+ murine tumor cell line. In addition,racotumomab-induced antibodies were able to react withNGcGM3+ lung carcinoma tissue sections [49]. No signifi-cant differences were found in the antibody response acrossthe dose levels examined (0.5, 1, and 2 mg). Maximal titersreached about 10,000 in the three dose levels for both IgMand IgG, with no significant differences in the titer meansbetween dose levels [29]. The 1 mg dose level was chosen forfurther clinical investigation [50].

NGcGM3-specific cellular responses were assessed inracotumomab-treated breast cancer patients. Cryopreservedperipheral blood mononuclear cells were challenged in vitrowith ganglioside-loaded, CD1d+, autologous monocyte-derived dendritic cells, and the response was measured

with an interferon-γ immunospot assay. A low frequency ofganglioside-specific interferon-γ-secreting cells was detectedin 5/13 patients. These cytokine responses were undetectableat baseline and became detectable by weeks 14 to 42,thereby confirming the convenience of an extended vac-cination schedule to elicit a ganglioside-specific response[29].

6. Toxicity and Preliminary Clinical Outcomes

The main toxicities observed in stage III/IV breast cancerpatients receiving the NGcGM3/VSSP vaccine (200 μg perdose) were erythema and induration at the injection site,occasionally associated with mild pain and fever. In spiteof the fact that this trial was not adequate for efficacyassessments, a remarkable progression-free survival time wasobserved in 2 patients with lung metastases [26]. Similarly,in another trial in advanced cutaneous and ocular malignantmelanomas, 7 patients treated with the NGcGM3/VSSPvaccine remained alive for more than 2 years after inclusionin the study [30].

In stage III/IV melanoma patients administered withbiweekly doses of the anti-idiotype mAb racotumomab(2 mg per dose), the tolerance was satisfactory and no unex-pected or serious adverse events were reported. The morefrequent adverse event was the local reaction with indurationand erythema at the injection site [48]. In a clinical trial inpatients with stage III/IV breast cancer, doses of 1 or 2 mg ofracotumomab were well tolerated. There were no differencesbetween the two levels of doses tested in toxicity [50]. Inanother Phase I trial, the toxicity profile of racotumomabwas investigated using an extended vaccination protocol of 6biweekly intradermal injections (induction phase), followedby 10 monthly boosters (maintenance). Nineteen patientswith high-risk (stage III) or metastatic breast cancer werevaccinated with different dose levels of 0.5, 1, and 2 mg.Vaccination was relatively well tolerated; local skin reactionsgrades I and II represented the most common adverse eventfollowed by mild flu-like symptoms lasting for 1 to 2 days.Similar safety results were observed with the 3 tested doselevels [29].

In a compassionate-use basis study, 34 stage IIIb and37 stage IV NSCLC patients were vaccinated with racotu-momab, after receiving standard chemotherapy and radio-therapy [28]. Patients were administered with 5 biweeklyinjections of 1 mg of racotumomab, other 10 doses at 28-day intervals, and later the patients who maintained a goodperformance status continued to be immunized at this sametime interval. No evidence of unexpected or serious adverseeffects was reported. The median survival time of patients,who entered the study with partial response or diseasestabilization and with a performance status (PS) 1 after thefirst line of chemo/radiotherapy, was 11.50 months sincethe start of vaccination. In contrast, the median survivaltime calculated for patients who started vaccination withprogressive disease and/or a PS2 was 6.50 months [28].A statistically significant correlation was observed betweenanti-ganglioside response and survival time in a subset of


20 NSCLC patients from this study. Nonresponder patients(n = 4) had a median survival time of 6.35 months (95%CI, 4.97–9.67 months), whereas patients who developedIgG and/or IgM antibodies against NGcGM3 had a mediansurvival time of 14.26 months (95% CI, 5.95–17.3 months;P < .01, log rank) [27].

Even though NGcGM3 is a glycolipid, its heterophilicexpression allowed a specific immune response whenpatients were immunized with either a conjugate vaccine(i.e., NGcGM3/VSSP) or an anti-idiotype mAb targetingNGc gangliosides (i.e., racotumomab). Evidence of an NGc-specific cytokine response was observed in some few patients,and its correlation with clinical outcomes remains to beestablished. The absence of cross-reaction with endogenousgangliosides is inline with the absence of autoimmunetoxicity and overall safety profile of both conjugated vaccineand the racotumomab mAb. On the other hand, hightiter ganglioside-specific antibody responses were observedin most of the patients, in correlation with survival ina group of NSCLC patients accrued after completion offirst-line chemotherapy [27], as described above. The pos-sible involvement of the induced antibodies in a protec-tive antitumor activity is actively being pursued. Recentsequencing and modelling studies suggest that racotumomabmight selectively induce Ab3 antibodies with conservedgermline sequences specific for heterophilic saccharideantigens [51]. Randomized controlled trials are presentlyunderway and are expected to provide further insight intothe role of the induced immune response on the clinicaloutcome.

In summary, the above-mentioned studies indicate thatboth vaccines targeted to NGc gangliosides have acceptablesafety outcomes and are able to induce specific humoraland cellular immune responses. The response to vaccinationseems to be stronger in those patients with lower tumorburden, better performance status, and a good responseto previous oncospecific treatment. Also, preliminary evi-dence suggested that these vaccines may have a positiveinfluence on survival in patients with immune responseto NGcGM3 antigen. The current Phase III trials that arebeing conducted at present will give a definitive answer tothe potential clinical benefits offered by these therapeuticvaccines. Furthermore, studies will be required to determinethe more efficient combination with chemotherapy or otherimmune interventions to prevail over the tumor-inducedimmunosuppression.

7. Perspectives

This paper deals with NGc gangliosides—and particularlyNGcGM3—as a target for cancer immunotherapy. With thenumerous antigens that can be used in immunotherapy,the decision-making process for researchers, hospitals, andcompanies, in whether or not to invest resources in aspecific antigen, has been always a very complicated matter.Fortunately, in a recent work by the National CancerInstitute Translational Research Working Group, Cheeveret al. developed a method for prioritization of cancer

antigens paving the way to take more rational, informeddecisions [7]. Such work aimed to develop a priority-ranked list of cancer vaccine target antigens based onpredefined and preweighted objective criteria. An additionalaim was testing a new approach for prioritizing transla-tional research opportunities based on an analytic hierarchyprocess (AHP), a structured technique, and a mathemat-ical model for dealing with complex decisions. Antigenprioritization involved developing a list of “ideal” cancerantigen criteria/characteristics, assigning relative weights tothose criteria using pairwise comparisons. The result of thecriteria weighting, in descending order, was as follows: (a)therapeutic function, (b) immunogenicity, (c) role of theantigen in oncogenicity, (d) specificity, (e) expression leveland percent of antigen-positive cells, (f) stem cell expression,(g) number of patients with antigen-positive cancers, (h)number of antigen epitopes, and (i) cellular location ofantigen expression.

Having that work as a reference, we rethought of ourexperimentation with NGc, and, although is neither on thescope of this paper nor our prerogative to position theantigen in the ranking, we can affirm that NGc somehowmatches all of the criteria considered (Table 2), at leastin some proportion—as described throughout this paper—whose relative weight should be evaluated by panels ofexternal experts.

Some authors have recently enunciated the introductionof potential biases in the National Cancer Institute PilotProject [52]. Lang et al. affirmed that the methodology used(AHP) is not well described and is subject to several sourcesof possible bias, such as participant selection, number ofantigens chosen for prioritization, errors in rank order,redundancy, and internal validity. First of all, we differ withLang et al. in the fact that AHP is not well described, dueto being a very well-known technique properly used in awide variety of settings, including cancer clinical decisions[53, 54]. Cheever et al. clearly described the method by citingthe popular work of Bhushan [55] and how AHP is used ina web-based tool [56]. AHP is a powerful tool, used widelyin science, and, although it has had some detractors overthe years, Forman and Gass . carried out an in-depth paperdiscussing and rebutting the academic criticisms of AHP[57].

Furthermore, at the time of Cheever’s paper publicationno cancer vaccine had yet been approved by FDA. However,recent approval of sipuleucel-T for men with advancedprostate cancer, targeting PAP antigen, gave us a valuablelesson on this matter [58]. Interestingly, PAP ranked 26out of 75 antigens in the ranking of cancer antigen pilotprioritization [7], confirming its capacity to somehow “fore-cast” those antigens more likely to be translated to patients.Although the ranking is dynamic, given that priorities changeas knowledge accrues from new studies, we must reinforcethe idea that the associated lists of weighted criteria informinvestigators as to what experimental evidence is requiredto advance antigens to higher priority levels. In this line,those criteria helped us to evaluate that NGcGM3 gangliosidecomprised most if not all relevant characteristics as a cancerantigen for vaccine development.


Acknowledgments

This work was partially supported by the National Agencyof Scientific and Technological Promotion (Argentina). Thesupport of Recombio (Spain) is also acknowledged. M. R.Gabri, L. Fainboim, D. E. Gomez and D. F. Alonso aremembers of the National Research Council (CONICET,Argentina). D. E. Gomez and D. F. Alonso contributedequally to this work.

References

[1] C. S. Higano, P. F. Schellhammer, E. J. Small et al.,“Integrated data from 2 randomized, double-blind, placebo-controlled, phase 3 trials of active cellular immunotherapywith sipuleucel-T in advanced prostate cancer,” Cancer, vol.115, no. 16, pp. 3670–3679, 2009.

[2] D. G. McNeel, E. J. Dunphy, J. G. Davies et al., “Safety andimmunological efficacy of a DNA vaccine encoding prostaticacid phosphatase in patients with stage D0 prostate cancer,”Journal of Clinical Oncology, vol. 27, no. 25, pp. 4047–4054,2009.

[3] R. F. Irie, K. Irie, and D. L. Morton, “A membrane antigencommon to human cancer and fetal calf brain,” CancerResearch, vol. 36, pp. 3510–3517, 1976.

[4] P. O. Livingston, “Approaches to augmenting the immuno-genicity of melanoma gangliosides: from whole melanomacells to ganglioside-KLH conjugate vaccines,” ImmunologicalReviews, no. 145, pp. 147–166, 1995.

[5] J. M. Kirkwood, J. G. Ibrahim, J. A. Sosman et al., “High-dose interferon α-2b significantly prolongs relapse-free andoverall survival compared with the GM2-KLH/QS-21 vaccinein patients with resected stage IIB-III melanoma: results ofintergroup trial E1694/S9512/C509801,” Journal of ClinicalOncology, vol. 19, no. 9, pp. 2370–2380, 2001.

[6] A. M. M. Eggermont, “Therapeutic vaccines in solid tumours:can they be harmful?” European Journal of Cancer, vol. 45, no.12, pp. 2087–2090, 2009.

[7] M. A. Cheever, J. P. Allison, A. S. Ferris et al., “Theprioritization of cancer antigens: a National Cancer Institutepilot project for the acceleration of translational research,”Clinical Cancer Research, vol. 15, no. 17, pp. 5323–5337, 2009.

[8] L. E. Fernandez, D. F. Alonso, D. E. Gomez, and A.M. Vazquez, “Ganglioside-based vaccines and anti-idiotypeantibodies for active immunotherapy against cancer,” ExpertReview of Vaccines, vol. 2, no. 6, pp. 817–823, 2003.

[9] R. Schauer, “Achievements and challenges of sialic acidresearch,” Glycoconjugate Journal, vol. 17, no. 7–9, pp. 485–499, 2000.

[10] E. A. Muchmore, S. Diaz, and A. Varki, “A structural differencebetween the cell surfaces of humans and the great apes,”American Journal of Physical Anthropology, vol. 107, no. 2, pp.187–198, 1998.

[11] A. Irie, S. Koyama, Y. Kozutsumi, T. Kawasaki, and A. Suzuki,“The molecular basis for the absence of N-glycolylneuraminicacid in humans,” Journal of Biological Chemistry, vol. 273, no.25, pp. 15866–15871, 1998.

[12] H.-H. Chou, H. Takematsu, S. Diaz et al., “A mutationin human CMP-sialic acid hydroxylase occurred after theHomo-Pan divergence,” Proceedings of the National Academyof Sciences of the United States of America, vol. 95, no. 20, pp.11751–11756, 1998.

[13] T. Hayakawa, Y. Satta, P. Gagneux, A. Varki, and N. Takahata,“Alu-mediated inactivation of the human CMP-N-acetylneu-

raminic acid hydroxylase gene,” Proceedings of the NationalAcademy of Sciences of the United States of America, vol. 98, no.20, pp. 11399–11404, 2001.

[14] H.-H. Chou, T. Hayakawa, S. Diaz et al., “Inactivation ofCMP-N-acetylneuraminic acid hydroxylase occurred prior tobrain expansion during human evolution,” Proceedings of theNational Academy of Sciences of the United States of America,vol. 99, no. 18, pp. 11736–11741, 2002.

[15] A. Varki and T. K. Altheide, “Comparing the human andchimpanzee genomes: searching for needles in a haystack,”Genome Research, vol. 15, no. 12, pp. 1746–1758, 2005.

[16] Y. N. Malykh, R. Schauer, and L. Shaw, “N-Glycolylneuraminicacid in human tumours,” Biochimie, vol. 83, no. 7, pp. 623–634, 2001.

[17] J. Vamecq, N. Mestdagh, J.-P. Henichart, and J. Poupaert,“Subcellular distribution of glycolyltransferases in rodent liverand their significance in special reference to the synthesis ofN-glycolylneuraminic acid,” Journal of Biochemistry, vol. 111,no. 5, pp. 579–583, 1992.

[18] M. Hedlund, P. Tangvoranuntakul, H. Takematsu et al., “N-glycolylneuraminic acid deficiency in mice: implications forhuman biology and evolution,” Molecular and Cellular Biology,vol. 27, no. 12, pp. 4340–4346, 2007.

[19] M. Bardor, D. H. Nguyen, S. Diaz, and A. Varki, “Mechanismof uptake and incorporation of the non-human sialic acidN-glycolylneuraminic acid into human cells,” Journal ofBiological Chemistry, vol. 280, no. 6, pp. 4228–4237, 2005.

[20] M. R. Gabri, L. L. Otero, D. E. Gomez, and D. F. Alonso,“Exogenous incorporation of neugc-rich mucin augments n-glycolyl sialic acid content and promotes malignant phenotypein mouse tumor cell lines,” Journal of Experimental andClinical Cancer Research, vol. 28, no. 1, article no. 146, 2009.

[21] M. Hedlund, V. Padler-Karavani, N. M. Varki, and A. Varki,“Evidence for a human-specific mechanism for diet andantibody-mediated inflammation in carcinoma progression,”Proceedings of the National Academy of Sciences of the UnitedStates of America, vol. 105, no. 48, pp. 18936–18941, 2008.

[22] A. Varki, “Uniquely human evolution of sialic acid geneticsand biology,” Proceedings of the National Academy of Sciences ofthe United States of America, vol. 107, supplement 2, pp. 8939–8946, 2010.

[23] J. de Leon, A. Fernandez, M. Clavell et al., “Differentialinfluence of the tumour-specific non-human sialic acidcontaining GM3 ganglioside on CD4+CD25− effector andnaturally occurring CD4+CD25+ regulatory T cells function,”International Immunology, vol. 20, no. 4, pp. 591–600, 2008.

[24] A. M. Vazquez, M. R. Gabri, A. M. Hernandez et al., “Anti-tumor properties of an anti-idiotypic monoclonal antibodyin relation to N-glycolyl-containing gangliosides,” OncologyReports, vol. 7, no. 4, pp. 751–756, 2000.

[25] D. Fuentes, J. Avellanet, A. Garcia et al., “Combined therapeu-tic effect of a monoclonal anti-idiotype tumor vaccine againstNeuGc-containing gangliosides with chemotherapy in a breastcarcinoma model,” Breast Cancer Research and Treatment, vol.120, no. 2, pp. 379–389, 2010.

[26] A. Carr, E. Rodrıguez, M. D. C. Arango et al., “Immunother-apy of advanced breast cancer with a heterophilic ganglioside(NeuGcGM3) cancer vaccine,” Journal of Clinical Oncology,vol. 21, no. 6, pp. 1015–1021, 2003.

[27] A. M. Hernandez, D. Toledo, D. Martınez et al., “Characteriza-tion of the antibody response against NeuGcGM3 gangliosideelicited in non-small cell lung cancer patients immunized withan anti-idiotype antibody,” Journal of Immunology, vol. 181,no. 9, pp. 6625–6634, 2008.


[28] S. Alfonso, R. M. Diaz, A. De La Torre et al., “1E10 anti-idiotype vaccine in non-small cell lung cancer: experience instage IIIb/IV patients,” Cancer Biology and Therapy, vol. 6, no.12, pp. 1847–1852, 2007.

[29] M. D. Guthmann, M. A. Castro, G. Cinat et al., “Cellularand humoral immune response to N-glycolyl-GM3 elicitedby prolonged immunotherapy with an anti-idiotypic vaccinein high-risk and metastatic breast cancer patients,” Journal ofImmunotherapy, vol. 29, no. 2, pp. 215–223, 2006.

[30] M. Osorio, E. Gracia, E. Rodrıguez et al., “HeterophilicNeuGcGM3 ganglioside cancer vaccine in advancedmelanoma patients: results of a Phase Ib/IIa study,” CancerBiology and Therapy, vol. 7, no. 4, pp. 488–495, 2008.

[31] A. Carr, A. Mullet, Z. Mazorra et al., “A mouse IgG1monoclonal antibody specific for N-glycolyl GM3 gangliosiderecognized breast and melanoma tumors,” Hybridoma, vol. 19,no. 3, pp. 241–247, 2000.

[32] H. Nakarai, P. J. Chandler, K. Kano, D. L. Morton, and R.F. Irie, “Hanganutziu-Deicher antigen as a possible targetfor immunotherapy of melanoma,” International Archives ofAllergy and Applied Immunology, vol. 91, no. 3, pp. 323–328,1990.

[33] G. Marquina, H. Waki, L. E. Fernandez et al., “Gangliosidesexpressed in human breast cancer,” Cancer Research, vol. 56,no. 22, pp. 5165–5171, 1996.

[34] P. Tangvoranuntakul, P. Gagneux, S. Diaz et al., “Humanuptake and incorporation of an immunogenic nonhumandietary sialic acid,” Proceedings of the National Academy ofSciences of the United States of America, vol. 100, no. 21, pp.12045–12050, 2003.

[35] H. van Cruijsen, M. Ruiz, P. van der Valk, T. D. de Gruijl, andG. Giaccone, “Tissue micro array analysis of ganglioside N-glycolyl GM3 expression and signal transducer and activatorof transcription (STAT)-3 activation in relation to dendriticcell infiltration and microvessel density in non-small cell lungcancer,” BMC Cancer, vol. 9, article no. 180, 2009.

[36] A. M. Scursoni, L. Galluzzo, S. Camarero et al., “Detectionand characterization of N-glycolyated gangliosides in Wilmstumor by immunohistochemistry,” Pediatric and Developmen-tal Pathology, vol. 13, no. 1, pp. 18–23, 2010.

[37] Y. Diaz, A. Gonzalez, A. Lopez, R. Perez, A. M. Vazquez,and E. Montero, “Anti-ganglioside anti-idiotypic monoclonalantibody-based cancer vaccine induces apoptosis and antian-giogenic effect in a metastatic lung carcinoma,” CancerImmunology, Immunotherapy, vol. 58, no. 7, pp. 1117–1128,2009.

[38] M. Labrada, M. Clavell, Y. Bebelagua et al., “Direct vali-dation of NGcGM3 ganglioside as a new target for cancerimmunotherapy,” Expert Opinion on Biological Therapy, vol.10, no. 2, pp. 153–162, 2010.

[39] S. C. Kaste, J. S. Dome, P. S. Babyn et al., “Wilms tumour:prognostic factors, staging, therapy and late effects,” PediatricRadiology, vol. 38, no. 1, pp. 2–17, 2008.

[40] A. M. Vazquez, M. Alfonso, B. Lanne et al., “Generationof a murine monoclonal antibody specific for N- glycolyl-neuraminic acid-containing gangliosides that also recognizessulfated glycolipids,” Hybridoma, vol. 14, no. 6, pp. 551–556,1995.

[41] A. S. Bansal, B. Abdul-Karim, R. A. Malik et al., “IgMganglioside GM1 antibodies in patients with autoimmunedisease or neuropathy, and controls,” Journal of ClinicalPathology, vol. 47, no. 4, pp. 300–302, 1994.

[42] I. Niebroj-Dobosz, P. Janik, and H. Kwiecinski, “Serum IgManti-GM1 ganglioside antibodies in lower motor neuron

syndromes,” European Journal of Neurology, vol. 11, no. 1, pp.13–16, 2004.

[43] S. Kusunoki, K.-I. Kaida, and M. Ueda, “Antibodies againstgangliosides and ganglioside complexes in Guillain-Barresyndrome: new aspects of research,” Biochimica et BiophysicaActa, vol. 1780, no. 3, pp. 441–444, 2008.

[44] H. J. Willison, “Gangliosides as targets for autoimmune injuryto the nervous system,” Journal of neurochemistry, vol. 103,supplement 1, pp. 143–149, 2007.

[45] D. L. V. Greenwood, V. M. Gitlits, F. Alderuccio, J. W. Sentry,and B.-H. Toh, “Autoantibodies in neuropsychiatric lupus,”Autoimmunity, vol. 35, no. 2, pp. 79–86, 2002.

[46] J. M. Merrick, K. Zadarlik, and F. Milgrom, “Characterizationof the Hanganutziu-Deicher (serum-sickness) antigen as gan-gliosides containing N-glycolylneuraminic acid,” InternationalArchives of Allergy and Applied Immunology, vol. 57, no. 5, pp.477–480, 1978.

[47] R. Kasukawa, K. Kano, M. L. Bloom, and F. Milgrom,“Heterophile antibodies in pathologic human sera resemblingantibodies stimulated by foreign species sera,” Clinical andExperimental Immunology, vol. 25, no. 1, pp. 122–132, 1976.

[48] M. Alfonso, A. Dıaz, A. M. Hernandez et al., “An anti-idiotypevaccine elicits a specific response to N-glycolyl sialic acidresidues of glycoconjugates in melanoma patients,” Journal ofImmunology, vol. 168, no. 5, pp. 2523–2529, 2002.

[49] E. Neninger, R. M. Dıaz, A. De La Torre et al., “Activeimmunotherapy with 1E10 anti-idiotype vaccine in patientswith small cell lung cancer: report of a phase I trial,” CancerBiology and Therapy, vol. 6, no. 2, pp. 145–150, 2007.

[50] A. Dıaz, M. Alfonso, R. Alonso et al., “Immune responsesin breast cancer patients immunized with an anti-idiotypeantibody mimicking NeuGc-containing gangliosides,” ClinicalImmunology, vol. 107, no. 2, pp. 80–89, 2003.

[51] A. Talavera, A. Eriksson, M. Okvist et al., “Crystal structure ofan anti-ganglioside antibody, and modelling of the functionalmimicry of its NeuGc-GM3 antigen by an anti-idiotypicantibody,” Molecular Immunology, vol. 46, no. 16, pp. 3466–3475, 2009.

[52] J. M. Lang, A.-C. Andrei, and D. G. McNeel, “Prioritization ofcancer antigens: keeping the target in sight,” Expert Review ofVaccines, vol. 8, no. 12, pp. 1657–1661, 2009.

[53] Y. Katsumura, H. Yasunaga, T. Imamura, K. Ohe, and H.Oyama, “Relationship between risk information on totalcolonoscopy and patient preferences for colorectal cancerscreening options: analysis using the Analytic HierarchyProcess,” BMC Health Services Research, vol. 8, 2008.

[54] J. G. Dolan and S. Iadarola, “Risk communication formatsfor low probability events: an exploratory study of patientpreferences,” BMC Medical Informatics and Decision Making,vol. 8, article no. 14, 2008.

[55] N. Bhushan and K. Rai, Strategic Decision Making: Applyingthe Analytic Hierarchy Process, Springer, London, UK, 2004.

[56] W. C. Olson, W. D. W. Heston, and A. K. Rajasekaran, “Clinicaltrials of cancer therapies targeting prostate-specific membraneantigen,” Reviews on Recent Clinical Trials, vol. 2, no. 3, pp.182–190, 2007.

[57] E. H. Forman and S. I. Gass, “The analytic hierarchy process—An exposition,” Operations Research, vol. 49, no. 4, pp. 469–486, 2001.

[58] A. Bot, “The landmark approval of provenge, what it meansto immunology and ‘in this issue’: the complex relationbetween vaccines and autoimmunity,” International Reviews ofImmunology, vol. 29, pp. 235–238, 2010.


Review Article

Dendritic Cell-Based Immunotherapy for Prostate Cancer

Hanka Jahnisch,1 Susanne Fussel,2 Andrea Kiessling,3 Rebekka Wehner,1 Stefan Zastrow,2

Michael Bachmann,1, 4 Ernst Peter Rieber,1 Manfred P. Wirth,2 and Marc Schmitz1, 4

1 Institute of Immunology, Medical Faculty, Technical University of Dresden, Fetscherstr. 74, 01307 Dresden, Germany2 Department of Urology, Medical Faculty, Technical University of Dresden, 01307 Dresden, Germany3 Translational Sciences and Safety, Novartis Biologic, 4002 Basel, Switzerland4 Center for Regenerative Therapies Dresden, 01307 Dresden, Germany

Correspondence should be addressed to Marc Schmitz, [email protected]

Received 30 June 2010; Accepted 7 October 2010

Academic Editor: Yang Liu

Copyright © 2010 Hanka Jahnisch et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Dendritic cells (DCs) are professional antigen-presenting cells (APCs), which display an extraordinary capacity to induce, sustain,and regulate T-cell responses providing the opportunity of DC-based cancer vaccination strategies. Thus, clinical trials enrollingprostate cancer patients were conducted, which were based on the administration of DCs loaded with tumor-associated antigens.These clinical trials revealed that DC-based immunotherapeutic strategies represent safe and feasible concepts for the inductionof immunological and clinical responses in prostate cancer patients. In this context, the administration of the vaccine sipuleucel-Tconsisting of autologous peripheral blood mononuclear cells including APCs, which were pre-exposed in vitro to the fusion proteinPA2024, resulted in a prolonged overall survival among patients with metastatic castration-resistent prostate cancer. In April 2010,sipuleucel-T was approved by the United States Food and Drug Administration for prostate cancer therapy.

1. Introduction

Prostate cancer (PCa) represents the most common noncuta-neous cancer and the second leading cause of cancer-relateddeaths in the United States with an estimated incidence of192,280 cases and an estimated number of 27,360 deaths in2009 [1]. In Europe, PCa is also the most frequent cancerdiagnosed in men with an estimated number of 345,900cases in 2006 [2]. Most of the patients are diagnosed withorgan-confined disease, for which radical prostatectomy,radiotherapy, and brachytherapy are effective treatmentmodalities [3, 4]. Active surveillance, which includes activemonitoring of the disease and start of treatment at pre-defined thresholds for progression, represents an alternativesince most of these tumors would never become of vitalclinical importance if they had not been detected. Althoughthe majority of patients are successfully treated with radicalprostatectomy or radiation therapy, approximately 30% ofpatients develop recurrent disease [5].

Androgen deprivation represents an effective treatmentmodality for recurrent PCa [3, 4]. Bisphosphonates can

increase bone mineral density and reduce the risk ofbone fractures, which are typical side effects of androgendeprivation therapy [6]. Similar effects are expected for thetreatment with denosumab, a monoclonal antibody againstreceptor activator of NF-kappaB ligand, which acts as a keymediator for osteoclast function, activation, and survival [7].

Therapeutic options for patients with progressive diseaseunder androgen deprivation therapy comprise secondaryhormonal manipulation and nonhormonal therapy suchas chemotherapy [3]. In the management of metastatichormone-refractory PCa (HRPC), chemotherapy with doc-etaxel serves as reference treatment due to the demonstratedsignificant survival benefit [8, 9]. In patients with HRPC, bis-phosphonates are useful for the treatment of skeletal compli-cations and pain relief thereby improving quality of life andalso providing a suitable medication for palliative care [3].

Despite of the therapeutic benefit of these approachesand the achieved prolongation of overall survival, additionaltreatment strategies are needed to prevent progression fromlocalized to advanced disease and to further improve survivaloutcomes for patients with advanced PCa.


2. The Important Role of Dendritic Cells inAntitumor Immunity

Dendritic cells (DCs) are professional antigen-presentingcells (APCs), which display a unique capacity to induce,sustain, and regulate T-cell responses [10, 11]. In tumor set-ting, DCs circulate through the blood and migrate to tumortissues, where they interact with malignant cells. ImmatureDCs are particularly efficient in the uptake of tumor-derived material. DC maturation is induced by tumor-derived molecules such as heat shock proteins and high-mobility-group box 1 protein as well as proinflammatorycytokines produced by various tumor-infiltrating immunecells. During maturation DCs migrate from tumor tissuesto T-cell-rich areas of secondary lymphoid organs, wherethey activate tumor-reactive CD8+ cytotoxic T lymphocytes(CTLs) and CD4+ T cells. CD8+ CTLs efficiently recognizeand destroy tumor cells, which expose peptides derivedfrom tumor-associated antigens (TAAs) in the complexwith human leukocyte antigen (HLA) class I molecules[12]. Clinical studies focusing on the adoptive transfer ofcytotoxic effector cells revealed tumor regression in cancerpatients [13]. CD4+ T cells recognizing peptides in thecontext of HLA class II molecules also play an importantrole in antitumor immunity [14]. CD4+ T cells improvethe capacity of DCs to induce CTLs by the interactionbetween CD40 on DCs and CD40 ligand on activatedCD4+ T cells. In addition, CD4+ T cells provide helpfor the maintenance and expansion of CTLs by secretingcytokines such as interleukin (IL)-2 and can eradicate tumorcells directly. Besides their extraordinary capacity to induceand stimulate T-cell responses, DCs efficiently improve theimmunomodulatory and cytotoxic potential of natural killercells, which essentially contribute to the elimination of tumorcells [15–17]. Furthermore, DCs can also directly mediatetumor-directed cytotoxicity [18–20]. Owing to their variousantitumor effects, DCs evolved as promising candidates forvaccination protocols in cancer therapy [21, 22].

3. Prostate Cancer-Associated Antigens forDC-Based Immunotherapy

Based on the crucial role of T cells in the elimination oftumor cells, much attention has been paied to the identifica-tion of tumor-associated proteins, that may provide targetsof tumor-reactive T cells, and on the definition of peptidemotifs within these proteins serving as T-cell epitopes. Here,we focus on PCa-associated target antigens, which havealready been used for DC-based vaccination trials enrollingPCa patients. A summary of these CD8+ T cell epitopes isdemonstrated in Table 1.

Prostate-specific antigen (PSA), a kallikrein-like serin-protease, is almost exclusively expressed by prostate epithelialcells, can be detected in the majority of PCa tissues, andrepresents the most widely used serum marker for diagnosisand monitoring of PCa [39–42]. The identification ofHLA-A2-restricted PSA-derived peptides was driven by invitro approaches using peptide-pulsed or RNA-transfectedAPCs to activate tumor-reactive CTLs [23–25, 27, 43].

By combining several previously identified and novel PSApeptides in an oligopeptide, Correale et al. demonstrated thepossibility of simultaneous induction of CTLs specific fordifferent epitopes dependent on the HLA repertoire of thepatient [26].

The integral membrane glycoprotein prostate-specificmembrane antigen (PSMA) represents a marker for normalprostate cells and can be detected in the majority of prostatetumors, particularly in undifferentiated, metastatic HRPC[44, 45]. Several HLA-A2-restricted peptides were shown toinduce tumor-reactive CTL responses in vitro and in vivo[28, 29].

Prostatic acid phosphatase (PAP) is a glycoprotein withenzymatic activity, which can be mainly detected in prostatetissue [46]. Peshwa et al. identified an HLA-A2-binding,endogenously generated, immunogenic peptide that inducedtumor-directed CTLs in vitro [30].

Prostate stem cell antigen (PSCA) is a glycosylphos-phatidylinositol-anchored cell surface glycoprotein, that ismainly expressed in the prostate [47]. PSCA expression isdetectable in more than 80% of primary PCa samples andbone metastases. It is increased in both androgen-dependentand -independent prostate tumors when compared to thecorresponding normal prostate tissues, particularly in carci-nomas of high stages and Gleason Scores [47, 48]. We andothers identified an HLA-A2-restricted PSCA peptide, whichinduced tumor-reactive CTL responses in vitro [31, 32].Increased frequencies of CD8+ T cells specific for this peptidewere found in the blood of PCa patients indicating therelevance of this epitope in vivo [32].

Prostein represents a transmembrane protein of the Golgiwith unique specificity for normal and malignant prostatetissues [49, 50]. Our group found abundant expression inmalignant and normal prostate tissues and maintained oreven elevated transcript levels in 87% of the primary tumorscompared to autologous nonmalignant tissue samples [33].By in vitro stimulation of CD8+ T lymphocytes with peptide-loaded DCs, we identified an autochthonously generated,HLA-A2-presented peptide, that was capable of activatingtumor-reactive CTLs [33].

The gene transient receptor potential (trp)-p8 encodes aseven-span transmembrane protein with significant homol-ogy to a family of Ca2+ channel proteins [51]. Trp-p8 ismainly detected in the prostate and shows an overexpressionin PCa of early stages and low grades [34, 51]. We identifiedan HLA-A2-binding peptide, which is able to stimulatetumor-reactive CTLs in vitro [34].

Potential target structures, which are overexpressed intumors of different origin including PCa comprise thehuman telomerase reverse transcriptase (hTERT), which isthe catalytic subunit of telomerase, and survivin. hTERT isundetectable in most nontransformed somatic cells but isexpressed in more than 85% of human tumors including PCa[52]. Several naturally generated CTL epitopes efficientlyinducing peptide-specific and tumor-reactive CTLs in vitroand in vivo have been described. Thus, the generationof HLA-A2-restricted hTERT peptide-specific CTLs, whichare able to lyse hTERT-expressing tumor cells of diversehistological origin including PCa cells has been reported


Table 1: PCa-associated antigen-derived CD8+ T-cell epitopes used for DC-based immunotherapy.

Antigen HLA restriction element Peptide position Amino acid sequence References

Prostate-specific antigen (PSA) HLA-A2146–154 KLQCVDLHV [23, 24]

141–150 FLTPKKLQCV [25, 26]

154–163 VISNDVCAQV [25–27]

Prostate-specific membrane antigen (PSMA) HLA-A24–12 LLHETDSAV [28, 29]

711–719 ALFDIESKV [29]

Prostatic acid phosphatase (PAP) HLA-A2 299–307 ALDVYNGLL [30]

Prostate stem cell antigen (PSCA) HLA-A2 14–22 ALQPGTALL [31, 32]

Prostein HLA-A2 31–39 CLAAGITYV [33]

Transient receptor potential-p8 (Trp-p8) HLA-A2 187–195 GLMKYIGEV [34]

Human telomerase reverse transcriptase (hTERT) HLA-A2 540–548 ILAKFLHWL [35, 36]

Survivin HLA-A2 95–104 ELTLGEFLKL [37, 38]

[35, 36]. In addition, the peptide proved to be immunogenicin vivo, since immunization of HLA-A2.1 transgenic micegenerated a specific CTL response [36]. Immunogenicityin mice could be markedly increased by an amino acidsubstitution at an HLA anchoring position [53].

Survivin, an inhibitor of apoptosis, is highly overex-pressed in many human tumors including PCa, and itsexpression correlates with aggressiveness and poor prognosisof tumor disease [54, 55]. The wide expression in cancerand the functional role for tumor cell survival make survivina promising target for T-cell-based immunotherapy. Wedescribed an endogenously produced HLA-A2-restrictedpeptide, which induced specific CTL responses in vitro[37]. Specific T-cell reactivity against this peptide motifwas detected in the peripheral blood of chronic lymphaticleukemia patients and in tumor-infiltrated lymph nodes ofmelanoma patients [38].

4. Dendritic Cell-Based Immunotherapy forProstate Cancer

DCs play a critical role for the induction of innate andadaptive antitumor immune responses. Due to their variousantitumor effects, DCs emerged as attractive candidatesfor vaccination protocols in cancer therapy (Figure 1).Animal models demonstrated that TAA-presenting DCs arecapable of inducing protective and therapeutic antitumorresponses [56, 57]. Clinical trials enrolling B-cell lymphoma,melanoma, or renal cancer patients revealed promisingimmunologic and clinical responses of TAA-loaded DCsadministered as a vaccine against cancer [58–61].

In prostate cancer setting, the administration of DCspulsed with TAA-derived peptides was well tolerated andresulted in the induction of immunological and clinicalresponses in patients. Thus, a phase-I trial was initiatedto evaluate the vaccination of DCs loaded with PSMA-derived peptides in HRPC patients [29, 62]. DCs weregenerated from monocytes in the presence of granulocyte-macrophage colony-stimulating factor (GM-CSF) and IL-4.Subsequently, the monocyte-derived DCs were pulsed withthe HLA-A2-restricted PSMA-derived peptides PSM-P1 or

PSM-P2. Nineteen patients received at least two infusionsof up to 2 × 107 peptide-loaded DCs at six- to eight-week intervals. Treatment was well tolerated except a mildto moderate transient hypotension. Five partial respondersbased on National Prostate Cancer Project criteria and a>50% reduction of PSA level were observed. Subsequently,a phase II trial was conducted to further investigate thetherapeutic efficiency of PSMA peptide-pulsed DCs [63].Six infusions of monocyte-derived DCs pulsed with PSM-P1 and PSM-P2 were administered at six week intervals.In addition, 17 patients received subcutaneous injections ofGM-CSF. Nine partial responders based on National ProstateCancer Project criteria and a >50% reduction of PSA levelwere identified in a group of 33 HRPC patients.

In another clinical study, monocyte-derived DCs pulsedwith a hTERT-derived peptide and keyhole limpet hemo-cyanin were administered to five patients with metastaticHRPC [64]. DCs were subcutaneously injected every otherweek for up to six vaccinations. Peptide-reactive T cells wereinduced in two patients after vaccination. All four evaluablepatients had stabilization of disease. Recently, we conducteda clinical study to evaluate the potential of DCs loadedwith a cocktail consisting of HLA-A2-restricted peptidesderived from PSA, PSMA, survivin, prostein, and trpp8[65]. Immature DCs were generated from monocytes in thepresence of GM-CSF and IL-4. For maturation, DCs wereincubated with GM-CSF, IL-4, IL-1β, IL-6, tumor necro-sis factor (TNF)-α, and prostaglandin E2. Subsequently,the mature monocyte-derived DCs were pulsed with fiveHLA-A∗0201-restricted TAA-derived peptides. Eight HRPCpatients received four vaccinations of every other week.Peptide-pulsed DCs were simultaneously injected intrader-mally and intravenously. One patient displayed a partialresponse. Three other patients showed stable disease over 4to 17 weeks. Three of these four PSA responders exhibitedspecific T-cell responses against prostein, survivin, or PSMA.In another clinical trial, six HRPC patients were treatedwith mature monocyte-derived DCs pulsed with a cocktailconsisting of HLA-A2-restricted peptides derived from PSA,PSCA, PSMA, and PAP [66]. Treatment was well tolerated.Three patients displayed specific T-cell responses against


Tumor peptideTumor protein

RNA codingfor tumor protein

DC

ActivationApoptosis

HLA I TCR

B7 CD28

CTL

TU

HLA II TCR

B7CD28

CD4

CTLIL-2

IFN-γCD40L-CD40

TU

Activation

Apoptosis

DC

Activation

Antigenpresentation

Figure 1: DC-based immunotherapeutic strategies for prostate cancer. DCs display a unique capacity to induce and maintain T-cellresponses and emerged as promising candidates for vaccination strategies in prostate cancer therapy. Thus, DCs are loaded with PCa-associated antigen-derived peptides, protein, or RNA. Due to their high surface expression of HLA-peptide-complexes and costimulatorymolecules, DCs efficiently activate and expand CD8+ CTLs and CD4+ T cells. CD8+ CTLs possess a profound capability to recognize anddestroy tumor cells. CD4+ T cells enhance the capacity of DCs to induce CTLs by the interaction between CD40 on DCs and CD40 ligandon activated CD4+ T cells. In addition, they provide help for the maintenance and expansion of CTLs by secreting cytokines and are able toeradicate tumor cells directly. CTLs: cytotoxic T cells; DCs: dendritic cells; HLA: human leukocyte antigen; IL: interleukin; IFN: interferon;TCR: T cell receptor; TU: tumor cells.

all antigens. Clinically, DC vaccination was associated withan increase in PSA doubling time. Thomas-Kaskel et al.initiated a clinical study to evaluate the vaccination ofDCs pulsed with PSA- and PSMA-derived peptides in 12patients with hormone- and chemotherapy-refractory PCa[67]. Patients received four vaccinations with a medianof 2.7 × 107 peptide-loaded mature monocyte-derivedDCs subcutaneously in biweekly intervals. Six patientshad stable disease and five patients developed delayed-type hypersensitivity (DTH) reactions. DTH-positivity wasassociated with superior survival. A significant correlationbetween DTH reactions and progression-free survival wasnot observed. Hildenbrand et al. conducted a clinical trialenrolling 12 HRCP patients, which was based on thecombination of interferon (IFN)-γ and mature monocyte-derived DCs pulsed with three different HLA-A2-restrictedPSA peptides [68]. Treatment consisted of the subcuta-neous injection of IFN-γ followed by three intracutaneousadministrations of 2 × 106 peptide-loaded DCs. Vaccinationwas applied four times at three-week intervals. No severeside effects were observed. One patient displayed a partialresponse showing regression of lymph node metastases,four patients showed stable disease, one patient exhibiteda mixed response, and six patients displayed progressivedisease.

Further clinical trials evaluated immunological andtherapeutic efficiency of protein-loaded DCs in PCa patients.Thus, Fong et al. administered DCs loaded with recombinantmurine PAP protein to 21 patients with metastatic PCa[69]. Patients received two injections monthly with a meandose of 11,2 × 106 cells per vaccination. Treatment waswell tolerated. All patients developed T-cell immunity tomouse PAP and 11 patients to the homologous self-antigenhuman PAP. Six patients displayed clinical stabilization oftheir previously progressing PCa as determined by PSA levelmonitoring, computerized tomography, and bone scans. Allthese patients developed T-cell proliferation in response tohuman PAP. Another clinical trial was performed to evaluatethe efficiency of mature monocyte-derived DCs pulsed withhuman recombinant PSA protein for the treatment of PCapatients in biochemical relapse after radical prostatectomy[70]. Twenty-four patients received nine administrations ofPSA-loaded DCs by combined intravenous, subcutaneous,and intradermal routes over 21 weeks. No severe side effectswere observed and 11 patients exhibited a transient PSAdecrease.

A particular promising immunotherapeutic strategy foradvanced PCa patients is based on the administration ofAPCs pre-exposed in vitro to PA2024, a fusion protein con-sisting of human GM-CSF and PAP (APC8015, sipuleucel-T,


Provenge). To generate sipuleucel-T, autologous peripheralblood mononuclear cells including APCs such as DCs werecollected by two sequential buoyant density centrifugationsteps and incubated with PA2024. Small et al. conductedsequential phase I and phase II trials including 31 HRPCpatients to determine the safety and efficacy of sipuleucel-T [71]. Patients were treated intravenously with sipuleucel-T on weeks 0, 4, 8, and 24. Treatment was well tolerated.No patient had pre-existing T-cell responses or antibodiesto PAP. After treatment, 38% of patients developed a T-cell response to PAP and 53% of patients had antibodies.Three patients had a more than 50% decline in PSA leveland additional 3 patients displayed 25% to 49% decreasesin PSA. In another phase II trial, 21 HRPC patients werevaccinated with sipuleucel-T [72]. In this study, the vaccinewas administered intravenously twice, on weeks 0 and 2. Themedian number of cells was 2.7 × 109 for the first infusionand 3.2× 109 for the second infusion. Subsequently, patientsreceived three subcutaneous injections of PA2024 at weeks4, 8, and 12. Two patients exhibited a 25% to 50% transientdecrease in PSA level. For a third patient, PSA droppedto undetectable levels by week 24. The PSA level remainedundetectable for 52 months and the metastatic adenopathyresolved.

Rini et al. performed a clinical trial, which was basedon the administration of sipuleucel-T and bevacizumabto 22 patients with recurrent PCa after definitive localtherapy [73]. Bevacizumab is a recombinant antibody againstvascular endothelial growth factor, that represents a proan-giogenic protein with inhibitory effects on APCs. Patientsreceived sipuleucel-T intravenously on weeks 0, 2, and 4and bevacizumab on weeks 0, 2, and 4 and every two weeksthereafter until toxicity or disease progression were observed.Nine patients displayed a decrease of PSA, ranging from 6%to 72%.

Following the results of the previous studies, a phase IIIstudy (D9901) enrolling 127 metastatic HRPC patients wasconducted to determine the safety and therapeutic efficiencyof sipuleucel-T in a placebo-controlled trial [74]. Patientswere randomized to receive three infusions of the vaccine orplacebo every two weeks with primary endpoint of time todisease progression. The median time to disease progressionwas not statistically significant at 11.7 weeks in the vaccinegroup compared with 10.0 weeks in the placebo group.However, a statistically significant increase in median overallsurvival was observed (25.9 months in the vaccine group;21,4 months in the placebo group).

More recently, Higano et al. performed an integral dataanalysis of the formerly described phase III study (D9901)and a second phase III trials (D9902A), which was also basedon the administration of sipuleucel-T to HRPC patients[75]. Altogether, 225 patients were randomized to receivethree infusions of sipuleucel-T (147 patients) or placebo(78 patients) every two weeks. Of the 147 patients inthe sipuleucel-T arms, 5 patients showed a PSA reductionof >50% and two additional patients of >25%. Patientsrandomized to sipuleucel-T had a 21% reduction in the riskof disease progression and a 33% reduction in the risk ofdeath compared with patients randomized to placebo. The

median survival was of 23,2 months in the sipuleucel-Tarms and 18,9 months in the placebo arms. The percentageof patients alive at 36 months was 33% in the sipuleucel-T arms and 15% in the placebo arms. Treatment was welltolerated. The overall incidence of adverse events was similarbetween patients treated with sipuleucel-T and patientstreated with placebo. The most common adverse events werechills, pyrexia, headache, asthenia, dyspnea, vomiting, andtremor. Taken together, the integrated results of D9901 andD9902A demonstrate a survival benefit for patients treatedwith sipuleucel-T compared to patients treated with placebo.To further confirm the therapeutic efficiency of sipuleucel-Tanother randomised, placebo-controlled, multicenter phaseIII trial enrolling 512 patients with metastatic HRPC wasconducted [76]. Patients on the sipuleucel-T treatment armexperienced a relative reduction of 22% in the risk of deathcompared with the placebo group. The median survival was25,8 months in the sipuleucel-T group and 21,7 months inthe placebo group. Based on these promising clinical results,the United States Food and Drug Administration recentlyapproved sipuleucel-T for the treatment of asymptomatic orminimally symptomatic, metastatic HRPC.

Further DC-based immunotherapeutic strategies forprostate cancer were evaluated in clinical trials. Thus, thepotential of RNA-transfected DCs for PCa therapy wasinvestigated. In this context, Heiser et al. vaccinated PSARNA-transfected DCs to 13 metastatic PCa patients [77].DCs were generated from monocytes in the presence ofGM-CSF and IL-4 and subsequently transfected with PSARNA. The PSA RNA-transfected DCs were administered atthree escalating dose levels. Dose escalation was performedthrough an intravenous route with 1 × 107, 3 × 107, or 5 ×107 cells applied at weeks 2, 4, and 6. For optimization, a con-comitant dose of 1× 107 cells was given intradermally at eachvaccination cycle. Induction of PSA-specific T-cell responseswas found in all evaluated patients. Six of seven evaluatedpatients displayed a significant decrease in the log slope PSA.In another clinical trial, hTERT RNA-transfected DCs wereadministered to 20 metastatic PCa patients [78]. Vaccinationresulted in an expansion of hTERT-specific T cells in 19patients, was associated with a reduction of PSA velocity, anda molecular clearance of circulating tumor cells. Mu et al.conducted a clinical study enrolling 20 HRPC patients, whichwas based on the administration of monocyte-derived DCstransfected with RNA from allogeneic PCa cell lines [79].Each patient received at least four weekly injections with 2× 107 transfected DCs either intranodally or intradermally.Thirteen of 19 patients that completed vaccination displayeda decrease in log slope PSA.

Another vaccination strategy for prostate cancer is basedon the administration of c-fms-like tyrosine kinase 3 (Flt3)ligand. This immunostimulatory agent efficiently promotesthe differentiation and expansion of DCs in vitro and invivo. Higano et al. conducted a trial evaluating the efficiencyof Flt3 ligand in HRPC patients [80]. Treatment was welltolerated. Flt3 ligand application resulted in a markedincrease of DC number in the peripheral blood. Eleven of 31patients showed a decrease or only a minor increase (<25%)in PSA levels.


5. Conclusion

DCs play a crucial role for the induction of innate andadaptive antitumor immune responses. Thus, they effi-ciently activate and expand tumor-reactive CD8+ CTLs andCD4+ T cells. In addition, DCs can markedly improve theimmunomodulatory and cytotoxic potential of natural killercells and can directly mediate tumor-directed cytotoxicity.Due to their various antitumor effects, DCs emerged aspromising candidates for the treatment of PCa patients.Consequently, several clinical trials enrolling PCa patientswere conducted, which were based on the administration ofDCs pulsed with TAA-derived peptides, protein, or RNA.These studies demonstrated that DC-based immunother-apeutic strategies represent safe and feasible concepts forthe induction of immunological and clinical responses inPCa patients. Recently, sipuleucel-T consisting of PA2024fusion protein-loaded APCs was approved by the UnitedStates Food and Drug Administration for the treatmentof asymptomatic or minimally symptomatic, metastaticHRPC. Despite these promising clinical effects the efficiencyof the various DC-based treatment modalities for manypatients with advanced PCa is still limited. Therefore,further improvement is required, which may be achieved bycombining DC-based vaccination strategies with antibody-,radio-, hormone-, chemo-, or antiangiogenic therapy.

Conflict of Interests

The authors have no conflict of interests.

References

[1] A. Jemal, R. Siegel, E. Ward, Y. Hao, J. Xu, and M. J. Thun,“Cancer statistics, 2009,” CA Cancer Journal for Clinicians, vol.59, no. 4, pp. 225–249, 2009.

[2] J. Ferlay, P. Autier, M. Boniol, M. Heanue, M. Colombet, andP. Boyle, “Estimates of the cancer incidence and mortality inEurope in 2006,” Annals of Oncology, vol. 18, no. 3, pp. 581–592, 2007.

[3] A. Heidenreich, G. Aus, M. Bolla et al., “EAU guidelines onprostate cancer,” European Urology, vol. 53, no. 1, pp. 68–80,2008.

[4] J.-E. Damber and G. Aus, “Prostate cancer,” The Lancet, vol.371, no. 9625, pp. 1710–1721, 2008.

[5] K. A. Roehl, M. Han, C. G. Ramos, J. A. V. Antenor,and W. J. Catalona, “Cancer progression and survival ratesfollowing anatomical radical retropubic prostatectomy in3,478 consecutive patients: long-term results,” The Journal ofUrology, vol. 172, no. 3, pp. 910–914, 2004.

[6] J.-B. Lattouf and F. Saad, “Bone complications of androgendeprivation therapy: screening, prevention, and treatment,”Current Opinion in Urology, vol. 20, no. 3, pp. 247–252, 2010.

[7] M. R. Smith, B. Egerdie, N. H. Toriz et al., “Denosumab in menreceiving androgen-deprivation therapy for prostate cancer,”The New England Journal of Medicine, vol. 361, no. 8, pp. 745–755, 2009.

[8] I. F. Tannock, R. De Wit, W. R. Berry et al., “Docetaxel plusprednisone or mitoxantrone plus prednisone for advancedprostate cancer,” The New England Journal of Medicine, vol.351, no. 15, pp. 1502–1512, 2004.

[9] D. P. Petrylak, C. M. Tangen, M. H. A. Hussain et al.,“Docetaxel and estramustine compared with mitoxantroneand prednisone for advanced refractory prostate cancer,” TheNew England Journal of Medicine, vol. 351, no. 15, pp. 1513–1520, 2004.

[10] J. Banchereau, F. Briere, C. Caux et al., “Immunobiology ofdendritic cells,” Annual Review of Immunology, vol. 18, pp.767–811, 2000.

[11] R. M. Steinman and J. Banchereau, “Taking dendritic cells intomedicine,” Nature, vol. 449, no. 7161, pp. 419–426, 2007.

[12] S. A. Rosenberg, “Cancer vaccines based on the identificationof genes encoding cancer regression antigens,” ImmunologyToday, vol. 18, no. 4, pp. 175–182, 1997.

[13] S. A. Rosenberg and M. E. Dudley, “Adoptive cell therapy forthe treatment of patients with metastatic melanoma,” CurrentOpinion in Immunology, vol. 21, no. 2, pp. 233–240, 2009.

[14] R.-F. Wang, “The role of MHC class II-restricted tumorantigens and CD4+ T cells in antitumor immunity,” Trends inImmunology, vol. 22, no. 5, pp. 269–276, 2001.

[15] N. C. Fernandez, A. Lozier, C. Flament et al., “Dendritic cellsdirectly trigger NK cell functions: cross-talk relevant in innateanti-tumor immune responses in vivo,” Nature Medicine, vol.5, no. 4, pp. 405–411, 1999.

[16] G. Ferlazzo, M. L. Tsang, L. Moretta, G. Melioli, R. M.Steinman, and C. Munz, “Human dendritic cells activateresting natural killer (NK) cells and are recognized viathe NKp30 receptor by activated NK cells,” The Journal ofExperimental Medicine, vol. 195, no. 3, pp. 343–351, 2002.

[17] R. Wehner, B. Lobel, M. Bornhauser et al., “Reciprocalactivating interaction between 6-sulfo LacNAc+ dendritic cellsand NK cells,” International Journal of Cancer, vol. 124, no. 2,pp. 358–366, 2009.

[18] N. A. Fanger, C. R. Maliszewski, K. Schooley, and T. S. Griffith,“Human dendritic cells mediate cellular apoptosis via tumornecrosis factor-related apoptosis-inducing ligand (TRAIL),”The Journal of Experimental Medicine, vol. 190, no. 8, pp.1155–1164, 1999.

[19] M. Schmitz, S. Zhao, Y. Deuse et al., “Tumoricidal potentialof native blood dendritic cells: direct tumor cell killing andactivation of NK cell-mediated cytotoxicity,” The Journal ofImmunology, vol. 174, no. 7, pp. 4127–4134, 2005.

[20] G. Stary, C. Bangert, M. Tauber, R. Strohal, T. Kopp, and G.Stingl, “Tumoricidal activity of TLR7/8-activated inflamma-tory dendritic cells,” The Journal of Experimental Medicine, vol.204, no. 6, pp. 1441–1451, 2007.

[21] J. Banchereau and A. K. Palucka, “Dendritic cells as therapeu-tic vaccines against cancer,” Nature Reviews Immunology, vol.5, no. 4, pp. 296–306, 2005.

[22] E. Gilboa, “DC-based cancer vaccines,” The Journal of ClinicalInvestigation, vol. 117, no. 5, pp. 1195–1203, 2007.

[23] B.-H. Xue, Y. Zhang, J. A. Sosman, and D. J. Peace, “Inductionof human cytotoxic T lymphocytes specific for prostate-specific antigen,” Prostate, vol. 30, no. 2, pp. 73–78, 1997.

[24] S. Perambakam, B.-H. Xue, J. A. Sosman, and D. J. Peace,“Induction of Tc2 cells with specificity for prostate-specificantigen from patients with hormone-refractory prostate can-cer,” Cancer Immunology, Immunotherapy, vol. 51, no. 5, pp.263–270, 2002.

[25] P. Correale, K. Walmsley, C. Nieroda et al., “In vitro generationof human cytotoxic T lymphocytes specific for peptidesderived from prostate-specific antigen,” Journal of the NationalCancer Institute, vol. 89, no. 4, pp. 293–300, 1997.

[26] P. Correale, K. Walmsley, S. Zaremba, M. Zhu, J. Schlom, andK. Y. Tsang, “Generation of human cytolytic T lymphocyte


lines directed against prostate-specific antigen (PSA) employ-ing a PSA oligoepitope peptide,” The Journal of Immunology,vol. 161, no. 6, pp. 3186–3194, 1998.

[27] A. Heiser, P. Dahm, D. R. Yancey et al., “Human dendritic cellstransfected with RNA encoding prostate-specific antigen stim-ulate prostate-specific CTL responses in vitro,” The Journal ofImmunology, vol. 164, no. 10, pp. 5508–5514, 2000.

[28] B. Tjoa, A. Boynton, G. Kenny, H. Ragde, S. L. Misrock, and G.Murphy, “Presentation of prostate tumor antigens by dendriticcells stimulates T-cell proliferation and cytotoxicity,” Prostate,vol. 28, no. 1, pp. 65–69, 1996.

[29] G. P. Murphy, B. Tjoa, H. Ragde, G. Kenny, and A. Boynton,“Phase I clinical trial: T-cell therapy for prostate cancer usingautologous dendritic cells pulsed with HLA-A0201-specificpeptides from prostate-specific membrane antigen,” Prostate,vol. 29, no. 6, pp. 371–380, 1996.

[30] M. V. Peshwa, J. D. Shi, C. Ruegg, R. Laus, and W. C. A.Van Schooten, “Induction of prostate tumor-specific CD8+cytotoxic T-lymphocytes in vitro using antigen-presentingcells pulsed with prostatic acid phosphatase peptide,” Prostate,vol. 36, no. 2, pp. 129–138, 1998.

[31] J. Dannull, P.-A. Diener, L. Prikler et al., “Prostate stemcell antigen is a promising candidate for immunotherapy ofadvanced prostate cancer,” Cancer Research, vol. 60, no. 19, pp.5522–5528, 2000.

[32] A. Kiessling, M. Schmitz, S. Stevanovic et al., “Prostatestem cell antigen: identification of immunogenic peptidesand assessment of reactive CD8+ T cells in prostate cancerpatients,” International Journal of Cancer, vol. 102, no. 4, pp.390–397, 2002.

[33] A. Kiessling, S. Stevanovic, S. Fussel et al., “Identification ofan HLA-A∗0201-restricted T-cell epitope derived from theprostate cancer-associated protein prostein,” British Journal ofCancer, vol. 90, no. 5, pp. 1034–1040, 2004.

[34] A. Kiessling, S. Fussel, M. Schmitz et al., “Identification ofan HLA-A∗0201-restricted T-cell epitope derived from theprostate cancer-associated protein trp-p8,” Prostate, vol. 56,no. 4, pp. 270–279, 2003.

[35] R. H. Vonderheide, W. C. Hahn, J. L. Schultze, and L.M. Nadler, “The telomerase catalytic subunit is a widelyexpressed tumor-associated antigen recognized by cytotoxic Tlymphocytes,” Immunity, vol. 10, no. 6, pp. 673–679, 1999.

[36] B. Minev, J. Hipp, H. Firat, J. D. Schmidt, P. Langlade-Demoyen, and M. Zanetti, “Cytotoxic T cell immunity againsttelomerase reverse transcriptase in humans,” Proceedings of theNational Academy of Sciences of the United States of America,vol. 97, no. 9, pp. 4796–4801, 2000.

[37] M. Schmitz, P. Diestelkoetter, B. Weigle et al., “Generationof survivin-specific CD8+ T effector cells by dendritic cellspulsed with protein or selected peptides,” Cancer Research, vol.60, no. 17, pp. 4845–4849, 2000.

[38] M. H. Andersen, L. O. Pedersen, J. C. Becket, and P. ThorStraten, “Identification of a cytotoxic T lymphocyte responseto the apoptosis inhibitor protein survivin in cancer patients,”Cancer Research, vol. 61, no. 3, pp. 869–872, 2001.

[39] A. Lundwall and H. Lilja, “Molecular cloning of humanprostate specific antigen cDNA,” FEBS Letters, vol. 214, no. 2,pp. 317–322, 1987.

[40] J. E. Oesterling, “Prostate specific antigen: a critical assessmentof the most useful tumor marker for adenocarcinoma of theprostate,” The Journal of Urology, vol. 145, no. 5, pp. 907–923,1991.

[41] S. P. Balk, Y.-J. Ko, and G. J. Bubley, “Biology of prostate-specific antigen,” Journal of Clinical Oncology, vol. 21, no. 2,pp. 383–391, 2003.

[42] S. J. Freedland, J. M. Hotaling, N. J. Fitzsimons et al., “PSA inthe new millennium: a powerful predictor of prostate cancerprognosis and radical prostatectomy outcomes—results fromthe SEARCH database,” European Urology, vol. 53, no. 4, pp.758–766, 2008.

[43] H. Terasawa, K.-Y. Tsang, J. Gulley, P. Arlen, and J. Schlom,“Identification and characterization of a human agonistcytotoxic T-lymphocyte epitope of human prostate-specificantigen,” Clinical Cancer Research, vol. 8, no. 1, pp. 41–53,2002.

[44] G. P. Murphy, A.-A. A. Elgamal, S. L. Su, D. G. Bostwick, andE. H. Holmes, “Current evaluation of the tissue localizationand diagnostic utility of prostate specific membrane antigen,”Cancer, vol. 83, no. 11, pp. 2259–2269, 1998.

[45] M. Kawakami and J. Nakayama, “Enhanced expression ofprostate-specific membrane antigen gene in prostate cancer asrevealed by in situ hybridization,” Cancer Research, vol. 57, no.12, pp. 2321–2324, 1997.

[46] T. Solin, M. Kontturi, R. Pohlmann, and P. Vihko, “Geneexpression and prostate specificity of human prostatic acidphosphatase (PAP): evaluation by RNA blot analyses,”Biochimica et Biophysica Acta, vol. 1048, no. 1, pp. 72–77, 1990.

[47] R. E. Reiter, Z. Gu, T. Watabe et al., “Prostate stem cellantigen: a cell surface marker overexpressed in prostatecancer,” Proceedings of the National Academy of Sciences of theUnited States of America, vol. 95, no. 4, pp. 1735–1740, 1998.

[48] Z. Gu, G. Thomas, J. Yamashiro et al., “Prostate stem cellantigen (PSCA) expression increases with high gleason score,advanced stage and bone metastasis in prostate cancer,”Oncogene, vol. 19, no. 10, pp. 1288–1296, 2000.

[49] J. Xu, M. Kalos, J. A. Stolk et al., “Identification andcharacterization of prostein, a novel prostate-specific protein,”Cancer Research, vol. 61, no. 4, pp. 1563–1568, 2001.

[50] M. Kalos, J. Askaa, B. L. Hylander et al., “Prostein expressionis highly restricted to normal and malignant prostate tissues,”Prostate, vol. 60, no. 3, pp. 246–256, 2004.

[51] L. Tsavaler, M. H. Shapero, S. Morkowski, and R. Laus, “Trp-p8, a novel prostate-specific gene, is up-regulated in prostatecancer and other malignancies and shares high homology withtransient receptor potential calcium channel proteins,” CancerResearch, vol. 61, no. 9, pp. 3760–3769, 2001.

[52] N. W. Kim, M. A. Piatyszek, K. R. Prowse et al., “Specificassociation of human telomerase activity with immortal cellsand cancer,” Science, vol. 266, no. 5193, pp. 2011–2015, 1994.

[53] M. Thorn, M. Wang, H. Kløverpris et al., “Identificationof a new hTERT-derived HLA-A∗0201 restricted, naturallyprocessed CTL epitope,” Cancer Immunology, Immunotherapy,vol. 56, no. 11, pp. 1755–1763, 2007.

[54] G. Ambrosini, C. Adida, and D. C. Altieri, “A novel anti-apoptosis gene, survivin, expressed in cancer and lymphoma,”Nature Medicine, vol. 3, no. 8, pp. 917–921, 1997.

[55] H. Koike, Y. Sekine, M. Kamiya, H. Nakazato, and K.Suzuki, “Gene expression of survivin and its spliced isoformsassociated with proliferation and aggressive phenotypes ofprostate cancer,” Urology, vol. 72, no. 6, pp. 1229–1233, 2008.

[56] J. I. Mayordomo, T. Zorina, W. J. Storkus et al., “Bone marrow-derived dendritic cells pulsed with synthetic tumour pep-tides elicit protective and therapeutic antitumour immunity,”Nature Medicine, vol. 1, no. 12, pp. 1297–1302, 1995.

[57] S. K. Nair, A. Heiser, D. Boczkowski et al., “Inductionof cytotoxic T cell responses and tumor immunity against


unrelated tumors using telomerase reverse transcriptase RNAtransfected dendritic cells,” Nature Medicine, vol. 6, no. 9, pp.1011–1017, 2000.

[58] F. J. Hsu, C. Benike, F. Fagnoni et al., “Vaccination ofpatients with B-cell lymphoma using autologous antigen-pulsed dendritic cells,” Nature Medicine, vol. 2, no. 1, pp. 52–58, 1996.

[59] F. O. Nestle, S. Alijagic, M. Gilliet et al., “Vaccination ofmelanoma patients with peptide- or tumor lysate-pulseddendritic cells,” Nature Medicine, vol. 4, no. 3, pp. 328–332,1998.

[60] B. Thurner, I. Haendle, C. Roder et al., “Vaccinationwith Mage-3A1 peptide-pulsed nature, monocyte-deriveddendritic cells expands specific cytotoxic T cells andinduces regression of some metastases in advanced stage IVmelanoma,” The Journal of Experimental Medicine, vol. 190,no. 11, pp. 1669–1678, 1999.

[61] J. Wierecky, M. R. Muller, S. Wirths et al., “Immunologicand clinical responses after vaccinations with peptide-pulseddendritic cells in metastatic renal cancer patients,” CancerResearch, vol. 66, no. 11, pp. 5910–5918, 2006.

[62] B. A. Tjoa, S. J. Erickson, V. A. Bowes et al., “Follow-upevaluation of prostate cancer patients infused with autologousdendritic cells pulsed with PSMA peptides,” Prostate, vol. 32,no. 4, pp. 272–278, 1997.

[63] B. A. Tjoa, S. J. Simmons, V. A. Bowes et al., “Evaluation ofphase I/II clinical trials in prostate cancer with dendritic cellsand PSMA peptides,” Prostate, vol. 36, no. 1, pp. 39–44, 1998.

[64] R. H. Vonderheide, S. M. Domchek, J. L. Schultze et al.,“Vaccination of cancer patients against telomerase inducesfunctional antitumor CD8+ T lymphocytes,” Clinical CancerResearch, vol. 10, no. 3, pp. 828–839, 2004.

[65] S. Fuessel, A. Meye, M. Schmitz et al., “Vaccination ofhormone-refractory prostate cancer patients with peptidecocktail-loaded dendritic cells: results of a phase I clinicaltrial,” Prostate, vol. 66, no. 8, pp. 811–821, 2006.

[66] Y. Waeckerle-Men, E. Uetz-Von Allmen, M. Fopp etal., “Dendritic cell-based multi-epitope immunotherapy ofhormone-refractory prostate carcinoma,” Cancer Immunology,Immunotherapy, vol. 55, no. 12, pp. 1524–1533, 2006.

[67] A.-K. Thomas-Kaskel, R. Zeiser, R. Jochim et al., “Vaccinationof advanced prostate cancer patients with PSCA and PSApeptide-loaded dendritic cells induces DTH responses thatcorrelate with superior overall survival,” International Journalof Cancer, vol. 119, no. 10, pp. 2428–2434, 2006.

[68] B. Hildenbrand, B. Sauer, O. Kalis et al., “Immunotherapyof patients with hormone-refractory prostate carcinoma pre-treated with interferon-gamma and vaccinated with autol-ogous PSA-peptide loaded dendritic cells—a pilot study,”Prostate, vol. 67, no. 5, pp. 500–508, 2007.

[69] L. Fong, D. Brockstedt, C. Benike et al., “Dendritic cell-basedxenoantigen vaccination for prostate cancer Immunotherapy,”The Journal of Immunology, vol. 167, no. 12, pp. 7150–7156,2001.

[70] B. Barrou, G. Benoıt, M. Ouldkaci et al., “Vaccination ofprostatectomized prostate cancer patients in biochemicalrelapse, with autologous dendritic cells pulsed with recombi-nant human PSA,” Cancer Immunology, Immunotherapy, vol.53, no. 5, pp. 453–460, 2004.

[71] E. J. Small, P. Fratesi, D. M. Reese et al., “Immunotherapyof hormone-refractory prostate cancer with antigen-loadeddendritic cells,” Journal of Clinical Oncology, vol. 18, no. 23,pp. 3894–3903, 2000.

[72] P. A. Burch, G. A. Croghan, D. A. Gastineau et al.,“Immunotherapy (APC8015, provenge) targeting prostaticacid phosphatase can induce durable remission of metastaticandrogen-independent prostate cancer: a phase 2 trial,”Prostate, vol. 60, no. 3, pp. 197–204, 2004.

[73] B. I. Rini, V. Weinberg, L. Fong, S. Conry, R. M. Hershberg,and E. J. Small, “Combination immunotherapy with prostaticacid phosphatase pulsed antigen-presenting cells (provenge)plus bevacizumab in patients with serologic progression ofprostate cancer after definitive local therapy,” Cancer, vol. 107,no. 1, pp. 67–74, 2006.

[74] E. J. Small, P. F. Schellhammer, C. S. Higano et al.,“Placebo-controlled phase III trial of immunologic therapywith Sipuleucel-T (APC8015) in patients with metastatic,asymptomatic hormone refractory prostate cancer,” Journal ofClinical Oncology, vol. 24, no. 19, pp. 3089–3094, 2006.


[76] P. W. Kantoff, C. S. Higano, N. D. Shore et al., “Sipuleucel-Timmunotherapy for castration-resistent prostate cancer,” TheNew England Journal of Medicine, vol. 363, no. 5, pp. 411–422,2010.

[77] A. Heiser, D. Coleman, J. Dannull et al., “Autologous dendriticcells transfected with prostate-specific antigen RNA stimulateCTL responses against metastatic prostate tumors,” TheJournal of Clinical Investigation, vol. 109, no. 3, pp. 409–417,2002.

[78] Z. Su, J. Dannull, B. K. Yang et al., “Telomerase mRNA-transfected dendritic cells stimulate antigen-specific CD8+and CD4+ T cell responses in patients with metastatic prostatecancer,” The Journal of Immunology, vol. 174, no. 6, pp. 3798–3807, 2005.

[79] L. J. Mu, J. A. Kyte, G. Kvalheim et al., “Immunotherapy withallotumour mRNA-transfected dendritic cells in androgen-resistant prostate cancer patients,” British Journal of Cancer,vol. 93, no. 7, pp. 749–756, 2005.

[80] C. S. Higano, N. J. Vogelzang, J. A. Sosman, A. Feng, D. Caron,and E. J. Small, “Safety and biological activity of repeated dosesof recombinant human Flt3 ligand in patients with bone scan-negative hormone-Refractory prostate cancer,” Clinical CancerResearch, vol. 10, no. 4, pp. 1219–1225, 2004.


Review Article

Harnessing the Effect of Adoptively Transferred Tumor-Reactive TCells on Endogenous (Host-Derived) Antitumor Immunity

Yolanda Nesbeth1 and Jose R. Conejo-Garcia2

1 Department of Microbiology and Immunology, Dartmouth Medical School, Lebanon, NH 03756, USA2 The Immunology Program, The Wistar Institute, Philadelphia, PA 19104, USA

Correspondence should be addressed to Jose R. Conejo-Garcia, [email protected]

Received 18 June 2010; Accepted 5 August 2010

Academic Editor: Kurt Blaser

Copyright © 2010 Y. Nesbeth and J. R. Conejo-Garcia. This is an open access article distributed under the Creative CommonsAttribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work isproperly cited.

Adoptive T cell transfer therapy, the ex vivo activation, expansion, and subsequent administration of tumor-reactive T cells, isalready the most effective therapy against certain types of cancer. However, recent evidence in animal models and clinical trialssuggests that host conditioning interventions tailored for some of the most aggressive and frequent epithelial cancers will be neededto maximize the benefit of this approach. Similarly, the subsets, stage of differentiation, and ex vivo expansion procedure of tumor-reactive T cells to be adoptively transferred influence their in vivo effectiveness and may need to be adapted for different typesof cancer and host conditioning interventions. The effects of adoptively transferred tumor-reactive T cells on the mechanisms ofendogenous (host-derived) antitumor immunity, and how to maximize their combined effects, are further discussed.

1. Introduction

It has been more than 50 years now since Thomas andBurnet first proposed the hypothesis that the immune systemcould identify and eradicate transformed or malignant cells,confirming earlier observations by Paul Ehrlich that an“overwhelming frequency” of carcinomas could be repressedby the immune system. This intrinsic ability of the immunesystem to provide control against malignancies has sincebeen refined and termed immunosurveillance [1–4]. Despitethe presence of immunosurveillance properties within theimmune system, immunocompetent patients still developcancers, yet these tumors are often less immunogenic thanthose that develop in immunosuppressed hosts. These andother observations led to the demonstration that tumorsare imprinted by their immune environment, and thisimprinting facilitates their transformation into populationsthat can more effectively resist the pressure exerted by theimmune system to eradicate them [5–7]. This process, inwhich the immune system acts both positively to inhibitthe progression of tumors and negatively to mold theestablishment of tumors that can evade its recognition, orworse to promote the advancement of tumor development,

is referred to as immunoediting [3, 8]. Thus, the immunesystem can prevent or promote tumor progression.

2. Myeloid Leukocytes Accumulate at TumorLocations and Induce Immunosuppression

Professional antigen presenting cells (APCs) with ade-quate stimulatory capacity are necessary within the tumormicroenvironment (TME) to induce sufficient effector cellsor cytokines to maintain their tumor-fighting capacity.However, tumor-bearing hosts do not appropriately presenttumor antigens. Instead, they mobilize immature myeloidcells that include precursors of macrophages, dendritic cells(DCs), and neutrophils. These cells, generically termedMyeloid-Derived Suppressor Cells [9] (MDSCs), massivelyaccumulate at splenic and solid tumor locations, wherethey contribute to tumor progression by providing growthfactors, as well as paracrine support for the formation ofblood vessels [10–15]. Most importantly, MDSCs abrogateantitumor immune responses through multiple mechanismsthat include, at least, the production of L-Arginase, NOand reactive oxygen species [10, 16–22], and the tyrosine


nitration of the T cell receptor [23]. Because of the het-erogeneous nature of the precursors recruited to tumorlocations as immature MDSCs, more differentiated butstill immunosuppressive macrophages or dendritic cells arealso frequently found in the tumor microenvironment.In tumors, the precise categorization of myeloid cells istherefore complicated by a high degree of phenotypic overlapand also depends on specific microenvironments. In ovariancancer, for instance, we have repeatedly demonstrated thatthe most abundant leukocyte subset in the SOLID tumormicroenvironment in humans, and in both tumor massesand ascites in mice, expresses low but detectable levels ofphenotypic markers of bona fide DCs, including CD11c,DEC205, CD86, and MHC-II (10, 13–15, 22, 24, 25).Irrespective of their overlapping phenotypic characteriza-tion, we have repeatedly demonstrated that when thesetumor leukocytes receive specific activating signals, they canfunctionally process full-length OVA in vitro [14, 24] andin vivo [22, 25], as well as effectively present processedSIINFEKL to T cells [10, 15, 22, 25].

Yet, while DCs are also abundant in the microenviron-ment of many other tumors, functional mature DCs capableof stimulating an antitumor response are not found in highfrequencies in human breast cancer, prostate cancer, ovariancancer, or renal cell carcinoma [26–30]. Cancer cells producevarious factors such as VEGF [31–36] and IL-6 [31, 37]that suppress DC differentiation and maturation [38, 39]. Atthe same time, cytokines that promote DC differentiation,such as granulocyte–macrophage colony-stimulating factor(GM-CSF) and IL-4 and Th1 polarizing cytokines like IFN-γ and IL-12, are seldom found in large quantities in manyhuman cancers, and ovarian cancer in particular [40, 41].Thus, this skewed cytokine profile promoted by the tumorimpairs the effective priming of an immunostimulatory DCphenotype and promotes the transition of DC precursor cellsrecruited to the tumor microenvironment into a suppressivepopulation. Importantly, in several cancer systems, DCs inboth the tumor microenvironment and peripheral blood canrevert to an immunostimulatory phenotype in vitro and canprime tumor-specific T cell responses [40, 42]. Nonetheless,the modulation of APCs does not appear to be strong enoughto overcome the tolerogenic environment of many tumors.In fact, in ovarian cancer, patients receiving multiple roundsof fully matured myeloid DCs were not able to regain Tcell function after their in vivo association with suppressivetumor-associated plasmacytoid DCs [40].

3. T Cells Exert Spontaneous Immune Pressureagainst Cancer Progression

In contrast, despite the heterogeneous nature of the CD3+

T cell compartment, the presence of T cells in the variousmalignancies generally correlates with improved clinicaloutcomes to the point that CD3+ T cells are consideredthe only immune population capable of exerting antitumoreffects against established tumors [43, 44].

The evidence of immune cell infiltrates and their abilityto mount antitumor responses in various tumor systems haveled investigators to target tumors through modulation of

the immune response. Immune-based therapies are deliveredeither through active immunotherapy, in which vaccinessuch as peptides, tumor antigens, nucleic acids, engineeredtumor cells, or tumor-pulsed DCs are used to activate hostantitumor immune cells to react against the tumor, or passiveimmunity wherein antibodies or antitumor lymphocytesare transferred into tumor-bearing hosts to directly inducetumor cell destruction [45]. Passive immunotherapy hasrevealed high success rates in certain implications, however,as most protocols direct responses against a single anti-gen/epitope, and tumors often modulate their expression ofparticular antigens, there is often a high degree of inefficacy.Active immunotherapy in both mouse and human tumorsystems have resulted in potent antitumor responses andregression, and is beneficial in the fact that rather thanrestricting responses to a single epitope /antigen, polyclonalresponses can readily be induced.

While both forms of immunotherapy have demonstratedpositive results, they each have drawbacks. The ideal systemwould entail passive therapeutics that can immediately starteliminating the tumor while inducing an active endogenousresponse to continue the tumor eradication. Under idealcircumstances, transferred T cells could migrate to the tumorsite and directly lyse tumor cells while releasing endogenousimmune cells from the tumor-induced immunosuppression.However, the tumor environment is usually so immunosup-pressive that it is difficult to appropriately release these brakemechanisms on antitumor responses.

4. Adoptive Cell Transfer Therapy Inducesthe Rejection of Advanced Tumors

Adoptive cell transfer therapy (ACT), the ex vivo activation,expansion, and subsequent administration of tumor-reactiveT cells, is a vastly successful therapy against certain cancers.In fact, ACT is currently the most effective therapy againstmetastatic melanoma, with objective regressions reportedin 50% of patients [46–49]. Adoptive T cell therapieshave focused on the use of CD8+ T cells, as they haverelatively long clonal expansion times, can specifically targettumors, and are easily subjected to genetic manipulations.Lymphodepletion has been used to enhance the persistenceof transferred T cells in vivo. By eliminating suppressivepopulations, removing cytokine sinks-endogenous cells thatcompete with the transferred cells for cytokines that promotetheir activation and function, and through augmentingthe function and availability of APCs, lymphodepletion isthought to enhance the antitumor response. In fact, inmelanoma, ACT was only effective after prior lymphode-pletion of patients, and this combination produced distinctand reproducible responses in roughly 50% of melanomapatients being treated with ACT.

ACT has also displayed remarkable success in humanclinical trials against Epstein-Barr virus- (EBV-) related dis-orders, immunoblastic lymphoma, and also Non-Hodgkin’sdisease [45, 50–55]. Yet, although these findings are opti-mistic for the future of adoptive immunotherapy, thesesystems are markedly different in that they are virally inducedtumor systems, and the T cells are directed against foreign,


rather than self, antigens. In most malignancies, beingnonviral, T cell antigenic targets are often self-antigens. Thisfurther complicates the ability to produce large numbers oftumor-reactive T cells since, not only do they usually occurin only low frequencies [56], but also most T cells thatrobustly respond to self antigens have either been eliminatedduring thymic development or rendered nonfunctional bylocal tolerizing mechanisms [57–59]. In fact, T cell adoptivetherapies have not resulted in impressive clinical benefits yetagainst the most lethal-epithelial-tumors [60–62]. Therefore,the expansion protocols for transferred T cells need tomaximize both the quality and quantity of tumor-reactive Tcells produced. As such, much work has gone into identifyingstrategies to optimize the ex vivo expansion of tumor-reactive T cells for ACT.

5. Ex Vivo Generation of Tumor-ReactiveT Cells for Adoptive Transfer

The main sources of modulation of the conditions for T cellexpansion include the antigen source, the cytokine environ-ment, and the source and effector stage of the T cells beforeexpansion [57]. Various adoptive cell transfer regimens entailthe nonspecific, polyclonal expansion of T cells throughmitogenic stimulation as with phytohaemagglutinin, orantibodies to CD3/CD28. These nonspecific manipulationshave achieved significant response rates against hepatocel-lular carcinoma, myeloma, non-Hodgkin’s lymphoma, andHodgkin’s disease [45, 63–66]. This would indicate thatnot only do the direct tumorlytic effects of CD8+ T cellscontribute to the beneficial responses of ACT, but also thatsecreted factors may also play a role.

The expansion of T cells against tumors of nonviralorigin pose different challenges, including the low frequencyof CTLs against self antigens [45]. In humans, the lowfrequency of precursor populations of tumor-reactive T cellshas been circumvented by prior vaccination of patientswith helper peptide-based vaccines. Vaccination of breastand ovarian cancer patients with HER-2/neu peptide-basedvaccines supplemented with GM-CCSF adjuvant treatmentover a six-month period increased the precursor frequencyof HER-2/neu-specific T cells that were capable of secretingIFN-γ in response to tumour and directly lyse HER-2/neuexpressing tumors [67]. Similarly, vaccination of breast can-cer patients with MUC-1 helper peptide vaccines producedCTLs reactive against MUC-1 expressing tumors [68]. Recentmouse models of melanoma have shown that these time-consuming and often cumbersome vaccination strategiesmay be bypassed by appropriate in vitro programming of thetransferred T cells. These protocols have an additional benefitover helper peptide vaccination in that they can facilitate theexpansion of polyclonal lymphocyte cultures.

An innovative approach used to circumvent the low fre-quency of tumor-reactive T cells in cancer patients has beenthe genetic manipulation of autologous T cells to expresseither T cell receptors (TCR) targeted to tumor-associatedantigens (TAA), or with chimeric receptors encompassinga B cell receptor to a particular antigen complexed with

the TCR signaling domain (so-called T-bodies) [48, 69].In a phase I clinical trial, administration of allogeneic Tcells that recognize the HLA-A2-restricted peptide MART-1 induced a partial response in one patient, while remainingpatients did not yield any overall response [70]. In anotherstudy, autologous peripheral blood lymphocytes retrovirallytransduced with the MART-1 TCR induced complete regres-sion in two patients with metastatic melanoma, but hadno effect on the remaining 13 patients in the cohort [48].While this technique provides a method for bypassing thecustomary low numbers of tumor-specific T cells capable ofbeing harvested from cancer patients, it poses the problem oflimiting the potential antitumor response to a single epitopewhich, if downregulated by the tumor, would render theprocedure useless.

Alternatively, T cells may be modified with genes toinhibit the induction of apoptosis or senescence. The anti-apoptotic genes BCL-2 and BCL-xL have been introducedinto T cells resulting in the extended survival of such cellseven under conditions that would usually promote apoptosis[71, 72]. Recently, it was demonstrated that both mouse andhuman tumor-reactive T cells could be effectively transducedwith siRNA to downmodulate their expression of PD-1.As anticipated, this prevented inhibitory signaling throughthe PD-1/PD-L1 inhibitory pair and instead generated Tcells with enhanced proliferation and immune function asdetermined through IL-2 and IFN-γ secretion [73]. Thistechnique provides access to a new realm of ACT, wherein Tcells can be engineered to specifically avoid the debilitatingTME without the need to induce systemic methods fordisrupting immunosuppression.

Other manipulations include the introduction ofautocrine growth signals into T cells before transfer toenhance their in vivo proliferation. This was first attemptedthrough the overexpression of IL-2 on T cells, which hadno effect on the tumorlytic capacity of these transferred Tcells. In contrast, unlike IL-2, IL-15 does not promote theexpansion of Tregs and when overexpressed in human Tcells prolonged the expression of antiapoptotic genes thusthe persistence of tumor specific cells and enhanced theirantitumor responses [74, 75].

6. Relative Efficacy of Individual T CellDifferentiation Subsets

The therapeutic effects of ACT have been commonlyattributed to the in vivo expansion and antitumor activityof transferred lymphocytes. Correspondingly, major effortshave been focused on promoting long-term persistenceof adoptively transferred T cells [76]. Various in vitromanipulations have been evaluated to enhance the in vivopersistence of transferred T cells. Thus, the T cells havebeen cultured in the presence of, or coadministered with, thecytokines IL-2, IL-7, IL-12, IL-15, and IL-21 [77, 78]. Whileeach has been found to have its own specific benefits anddemonstrated to enhance the therapeutic effects of adoptiveimmunotherapy in mouse models of melanoma, their effectshave been related to the promotion of a specific effector


phase of the transferred T cell. This has led to several studiesevaluating the efficacy of particular T cell differentiationsubsets in adoptive T cell therapy against cancer.

Although effector memory T cells (TEM) are superior tocentral memory T cells (TCM) at inducing in vitro cytotoxi-city of transformed cell cultures, they have poor replicativecapacity in vivo, and TCM exert superior therapeutic benefitsto TEM cells [76, 79–82]. Central memory T cells, being theleast differentiated of the antigen experienced populationof T cells, and being thought to have the capacity for self-renewal, and for retaining the option to differentiate intoa vast repertoire of T cell populations, was for some timeconsidered the ideal starting population of cells for expan-sion protocols [83]. These TCM cells can undergo robustexpansion in response to secondary exposure to antigen andsecrete copious levels of IL-2, in stark contrast to TEM cells.It was later found that compared with more differentiatedeffector lymphocytes, or memory T cells, early effectors havea higher capacity for in vivo expansion, which is associatedwith enhanced therapeutic effects against melanoma [84].Thus, fewer early effector T cells specific to the gp100 antigenwere necessary to induce regression of melanoma in mice.More recently still, we demonstrated for the first time thatnaıve or briefly activated T cells can induce potent antitumorresponses in adoptive T cell transfer experiments, which wassubsequently confirmed in an independent study [24, 85,86]. These effects coincided with the in vivo differentiationof these precursor cells into cytotoxic cells and the inductionof endogenous immune responses, which were found to benecessary for the therapeutic effects.

7. Tumorlytic Activity of CD4+ versusCD8+ T Cells

Most investigations on the antitumor effects of T cells havecentered around CD8+ T cells due to their high expression invarious malignancies, ease of isolation and in vitro manip-ulation, and their keen ability to directly lyse tumor cells[87]. Several reports had demonstrated the efficacy of CD8+

T cells in inducing potent antitumor responses, although itis widely accepted that their ability to clear tumors requiresfurther manipulation of the T cell directly, for example,through genetic elimination of inhibitory surface receptors[88], addition of specific TCRs or of the hosts throughirradiation [89], or other immune interventions. Still, thecontribution of CD4+ T cells to adoptive immunotherapy,particularly against epithelial cancers, remains controversial.However, it is widely accepted that a great deal of thefailures that arise from the use of CD8+ T cells stem fromthe absence of CD4+ T cell help necessary for maintainingtheir in vivo functionality [90–93]. Despite the immenseamount of data supporting the positive contribution ofcertain subsets of CD4+ T cells in enhancing the efficacyof function and persistence of CD8+ T cells, manipulationsutilizing CD4+ T cells have been very limited. CD4+ Tcells present as a particularly difficult population of cells towork with as they do not proliferate as effectively in vitroas do CD8+ T cells [94], and very little progress has been

made in the identification of class II restricted peptides [87].Furthermore, most tumors do not express MHC-II and aretherefore not directly recognized by CD4+ T cells.

Due to the great degree of homogeneity within theCD4+ T cell compartment and thus the wide spectrum ofopposing effects potentially inducible by these cells, as wellas the deficit in knowledge of MHC-II (CD4) restrictedepitopes [87], the role of CD4+ T cells in antitumorimmunity remains an investigative area that has been largelyneglected. Furthermore, the majority of studies into thispopulation have focused on the adverse effects of regulatoryCD4+ T cells, thus creating a negative reputation for thesecells in tumor immunology. Still, there have been severalreports demonstrating the beneficial role of CD4+ T cellsin antitumor immunity, providing rationale for undertakingfurther investigations in this area.

Evidence from various studies show that in the absence ofCD8+ T cells, CD4+ T cells were still capable of eliminatingboth haematologic and solid tumors [95–97]. Using trans-genic T cells specific to different H-Y antigens, Perez-Diezand colleagues were able to demonstrate in 6 different tumormodels that CD4+ T cells were more effective than CD8+ Tcells (or a mixed population of both CD4+ and CD8+ T cells)at rejecting tumors even in the absence of MHC-II expressionon the tumor cells [98]. There exists the possibility that thedifferences in antigen epitopes and TCR avidities may beresponsible for these observed effects. Importantly, however,the authors found that antigen presentation by host cells wasrequired at the effector phase for this tumor rejection byprimed CD4+ T cells and speculate that this may be throughthe activation of local macrophages and other cells but nevervalidated this.

Recently, two articles confirmed the positive contribu-tion of CD4+ T cells in adoptive immunotherapy againstmelanoma, both describing a direct tumorlytic effect of thesetransferred T cells [85, 99]. In both cases, small numbers ofnaıve CD4+ T cells specific to the Trp1 melanoma antigenwere transferred into irradiated recipient mice bearing estab-lished B16 melanoma. Interestingly, these cells expandedrobustly and importantly differentiated into cytotoxic CD4+

T cells that directly eliminated B16 melanomas [85, 99].These tumors do not express MHC-II, but it was furthershown that the secretion of IFN-γ by the transferred CD4 ledto the upregulation of MHC-II on these tumors making themdirect targets of the transferred T cells. In one context, fur-ther immune intervention by antibody-mediated blockade ofthe coinhibitory receptor CTL-associated antigen 4 (CTLA-4) on T cells augmented the antitumor activity throughenhancing the expansion of the transferred T cells, increasingIFN-γ levels, thus cytotoxicity, and reducing the number ofTregs present.

8. Alternative Mechanisms of EnhancedAntitumor Immunity Mediated byCD4+ T Cells

While the studies referred to above underscore the directtumorlytic potential of CD4+ T cells, it is generally accepted


that most human tumors do not express MHC-II andare therefore insensitive to CD4+ T cell-mediated cytoly-sis. However, other than directly lysing tumors, CD4+ Tcells have been demonstrated to contribute to antitumorresponses through the provision of cytokine support, themaintenance and survival of CD8+ T cells and through theexpression of CD40L [100–104]. Indeed, we demonstratedthat adoptively transferred CD4+ T cells, through CD40L-CD40 interactions, license tumor-associated DCs to primeendogenous antitumor CD8+ T cells [24, 86]. Thus, DCsthat in the tumor microenvironment contributed to thepromotion of immunosuppressive conditions, when giventhe appropriate stimuli, including CD40 signaling throughCD40L- expression on transferred T cells, were capableof priming antitumor responses. This concurs with theirability to uptake tumor antigens while retaining an immaturephenotype such that the mere provision of this additionalstimulus was capable of reversing their phenotype. Thisinduction of endogenous responses had greater ramifica-tions, as we demonstrated that these host immune responsesremained active for prolonged periods and protected naıvemice from challenge with the same tumor [24, 86].

In addition to directly promoting CD8+ T cell func-tionality, CD4+ T cells have been shown to secrete variouscytokines that activate host antigen presenting cells, andtheir coadministration with CD8+ T cells revealed enhancedtherapeutic benefits coupled with the induction of a robustcentral memory response [105–107]. Thus, Hunder et al.provided evidence with a single case of effective adoptiveT cell therapy utilizing NY-ESO-1-specific CD4+ T cellscultured with IL-7 and IL-2 for the treatment of a patientwith metastatic melanoma who had not received priorlymphodepletion or vaccination therapy [49].

Moreover, the infiltration of immune populations inovarian cancer is modulated by chemokines, which thereforeinfluence the clinical outcome. Elucidation of factors thatcontribute to the infiltration of immune cells into the ovariancancer microenvironment (but not breast cancer) [108–110],revealed that tumors with significant T cell infiltrates hadelevated levels of various chemokines, including CCL5, theproduction of which was found to be restricted to the lym-phocyte population rather than the tumor cells [111, 112].Our studies demonstrate that CD4+ T cells expanded againsttumor antigen secrete high levels of CCL5, thus promotingthe recruitment of CCR5 expressing T cells and DCs tothe tumor site [24, 86]. The chemokine receptor CCR5 isexpressed on memory/effector like T cells and is associatedwith Th1 type responses. Our findings have been mirroredby a report from Dobrzanski et al. that demonstrates thatthe adoptive transfer of MUC1 specific CD4+ T cells increaseendogenous T cell activity and the survival of patients withresidual recurrent epithelial ovarian cancer, and that theseeffects corresponded with increased expression of CCR5 andassociated ligands on tumor responsive T cells [113].

Collectively, these data indicate that CD4+ T cells con-tribute positively to the induction of antitumor responsesachieved through adoptive T cell transfer regimens in ovariancancer, and likely in other tumors. We found that CD4+

T cells could independently delay tumor progression but a

mixed population of CD4+ and CD8+ T cells induced greaterantitumor efficacy against our aggressive model of ovariancancer. Thus, we now appreciate the fact that it is the qualityrather than quantity of adoptively transferred T cells that ismore relevant for achieving positive clinical outcomes, andthat the appropriate host conditioning strategies must beemployed to retain their functionality and maximize theirtherapeutic efficacy.

It should be noted, however, that preliminary resultsfrom ongoing trials in patients with metastatic melanomasuggest that the inclusion of antitumor CD4+ T cells in theadoptively transferred T cell population results in poorerclinical responses, which are associated with the expansionof the regulatory T cell compartment [114]. It is thereforelikely that the antitumor effectiveness of CD4+ T cells coulddepend on the type of cancer or the host conditioningstrategy applied to support the adoptively transferred lym-phocytes. For instance, high doses of IL-2 are administered topatients receiving antitumor T cells, but not always to tumor-bearing mice in these published reports. The preferentialeffect of IL-2 on regulatory T cells contained among theCD4+ T cells could at least partially explain the discrepanciesbetween mouse systems and these clinical results, and help todesign improved approaches.

9. Immunosuppressive TumorMicroenvironmental Networks Abrogatethe Activity of Adoptively TransferredTumor-Reactive T Cells againstAggressive Epithelial Tumors

Adoptive T cell therapy, while highly successful for manynonepithelial cancers, has not yet been effective in the mostfrequent and aggressive epithelial cancers, likely due to thepeculiarities of their respective microenvironments. In ovar-ian cancer, for instance, adoptively transferred autologousT cells directed at the α-folate receptor disappeared rapidly(often within a month) in association with increasing levelsof an undetermined inhibitory factor [115]. It appears thatmany of the immunotherapies attempted against advancedepithelial cancers have the capacity to induce the productionof potent CTLs, yet this has not proven sufficient to translateto improved survival in all cases, likely as a result of tolero-genic factors within the tumor microenvironment. Recentreports of induction of antitumor immune responses uponcombination of CTLA-4 blockade along with vaccination inovarian cancer patients [116, 117] highlight the relevanceof overcoming immunosuppression, particularly in conjunc-tion with other immune strategies to produce antitumorimmunity. Therefore, it has become abundantly clear fromthe wealth of experimental data in this field that due tothe diversity of mechanisms employed by tumors to evadeimmune destruction, the appropriate immunotherapeuticregime may not simply target an individual aspect, but mayneed to incorporate strategies that address multiple immunepathways.

Several reports propose various methods for enhancingthe in vivo survival of adoptively transferred lymphocytes.


One such method is through the sublethal irradiation oftumor-bearing hosts to create space to accommodate theexpansion of the transferred T cells. We found that evenunder the context of irradiation and depleting regulatorymyeloid cells from tumor locations, our transferred T cellsdid not persist for long periods, although the combination ofirradiation and immunosuppressive myeloid cell depletionenhanced the therapeutic benefit observed when T cellswere transferred into tumor-bearing mice. As an individualintervention, elimination of immunosuppressive myeloidcells in tumor-bearing mice disrupted tumor vasculature,produced an immunogenic boost, and thereby delayedtumor progression [15]. Accordingly, the elimination ofthis immunosuppressive population of cells bolstered thein vivo expansion and therapeutic effectiveness of adoptiveimmunotherapy in our ovarian cancer models, but not thepersistence of transferred lymphocytes [24, 86].

While irradiation did not enhance the survival of thetransferred T cells, it likely enhanced the immunogenicitythrough inducing the death of some tumor cells, and thusreleasing tumor antigen that could trigger host immuneresponses. Furthermore, irradiation can cause upregulationof certain molecules on tumor cells, such as MHC-I or thedeath receptor Fas, that render them more immunogenic andflag them as better targets for immune elimination [118].

The persistence of transferred T cells correlates withgreater efficacy in most cancer systems, thus enhancing thesurvival of these transferred T cells is a future direction to betaken into consideration. Stimulation of CD40 and Toll-likeReceptor 3 on ovarian cancer infiltrating DCs converts themfrom immunosuppressive to immunostimulatory cells andboosts T cell-mediated antitumor immune responses [22].Such pretreatment of tumor bearing hosts before ACT mayextend the survival of transferred T cells. Ongoing studies inour laboratory should define the potential of this approach.

Notwithstanding, the impact of standard treatmentmodalities should not be disregarded and immune therapiesshould probably be administered in conjunction with,rather than, in place of such. Surgical debulking maystill be a necessary procedure for the removal of largetumor masses, while, as we and others have demonstrated,chemotherapy/radiation therapy may bolster the effects ofimmunotherapies. Finally, immune-based therapies may addto the antitumor armament by eradicating residual diseaseand activating endogenous antitumor responses that persistideally in the memory compartment to prevent metastaticlesions and to control recurrences.

Such trimodal approaches (surgery plus chemother-apy/radiation plus immunotherapy) probably represent thefuture in the battle against epithelial cancers. Immunothera-peutic interventions, since largely hypothetical, are tested inpatients with late stage, very advanced disease, or recurrentdisease that is often refractory to standard therapies, inwhich case the efficacy of any intervention is highly unlikelyand mostly improbable. Trials in patients whose disease hasnot progressed as far may prove to reveal more favorableclinical outcomes, and, through the elicitation of protectiveendogenous immune responses, may prevent recurrence andincrease the rate of survival of endothelial cancer patients.

Drastic measures need to be taken to defeat the grim effectsof the most devastating cancers.

10. Effect of ACT on Endogenous OngoingAntitumor Immunity

The prevailing concept surrounding ACT is that successfulACT requires the persistence of the transferred T cells,which are considered the ultimate mediators of the antitu-mor response. Importantly, the contribution of endogenousresponses to the efficacy of immune-based therapies hasbeen a largely neglected area. As stated above, however,our studies in ovarian cancer models show that suchendogenous responses are not only important, but crucialto the elimination of established tumors and the inductionof persistent memory responses [24]. As described above, wefound that our T cells briefly primed against tumor antigensdo not persist for very long (as in human ovarian cancer)but instead elicit the awakening of host immune populationsthat induce sustained antitumor responses [24, 86]. Existing(although obviously suboptimal) antitumor responses weresignificantly boosted in mice receiving adoptively transferredtumor-reactive T cells. Most importantly, endogenous T cell-mediated responses were long-lived and more persistent thanthe activity of transferred lymphocytes. Thus, adoptivelytransferred T cells stimulate the awakening of host immuneresponses and host cells after ACT developed the ability torecognize and react to tumor antigens. The transferred Tcells required perforin for maximal effectiveness suggestingthat these transferred CTLs induce immunogenic tumordeath triggering the release of tumor antigen that may primeDC activation. CD4+ T cells provided further costimulatorymolecules to complete the activation of these DCs indicatingthat the adoptively transferred CD4+ and CD8+ T cellscooperate to induce their antitumor effects.

These results imply that while persistence and directantitumor activity of adoptively transferred T cells is crucialfor their therapeutic potential, and how they impact existingimmune responses may be another variable to optimizein a clinical context. Unleashing endogenous antitumorimmunity may also result from host-conditioning strategiesand synergize with ACT. Thus, interventions aimed to trans-form tumor microenvironmental cells from an immuno-suppressive to an immunostimulatory phenotype (such asCD40+TLR agonists) may be ideal to boost the expansion,persistence, and therapeutic activity of both adoptivelytransferred and endogenous tumor-reactive lymphocytes.

11. Concluding Remarks

Despite a great deal of effort being dedicated to the develop-ment of new therapies, there has been minimal improvementin the survival rate for most cancers including epithelial ovar-ian cancer. Strategies that have proven successful in certainmalignancies have not produced similar results in epithelialcancers like ovarian cancer, highlighting the complexitiesexisting within the microenvironment of individual cancersand emphasizing the need to consider each tumor as an


independent entity. T cell therapies often fail due to thetolerogenic environment in which the T cells are placed andthat integrating techniques that reduce the immunosuppres-sive nature of the tumor microenvironment will enhance theefficacy of ACT and make it a viable treatment modality.Newly developed immunotherapies will need to address mul-tiple immune pathways and circumvent various mechanismsof immune evasion and importantly need to incorporatestrategies that contribute to the induction of endogenousresponses which we had found to be not only beneficial,but crucial to the elimination of established tumors and theinduction of persistent memory responses. It is apparent thatthe appropriate T cell polarization and differentiation willneed to be identified in individual tumor systems for theoptimal function of anticancer lymphocytes, and to break thetumor-induced paralysis of host immune responses. Further-more, while most studies have focused on the contribution ofor administration of cytotoxic CD8+ T cells, it is becomingincreasingly clear that the coadministration of appropriateCD4+ T cell subsets may be advantageous to the therapeuticeffects of ACT, particularly through the elicitation of endoge-nous antitumor responses, and their incorporation into ACTregimens should be further investigated.

References

[1] T. H. Schreiber and E. R. Podack, “A critical analysis ofthe tumour immunosurveillance controversy for 3-MCA-induced sarcomas,” British Journal of Cancer, vol. 101, no. 3,pp. 381–386, 2009.

[2] G. P. Dunn, L. J. Old, and R. D. Schreiber, “The three Es ofcancer immunoediting,” Annual Review of Immunology, vol.22, pp. 329–360, 2004.

[3] G. P. Dunn, L. J. Old, and R. D. Schreiber, “The immuno-biology of cancer immunosurveillance and immunoediting,”Immunity, vol. 21, no. 2, pp. 137–148, 2004.

[4] G. P. Dunn, A. T. Bruce, H. Ikeda, L. J. Old, and R. D.Schreiber, “Cancer immunoediting: from immunosurveil-lance to tumor escape,” Nature Immunology, vol. 3, no. 11,pp. 991–998, 2002.

[5] J. B. Swann, M. D. Vesely, A. Silva et al., “Demonstrationof inflammation-induced cancer and cancer immunoeditingduring primary tumorigenesis,” Proceedings of the NationalAcademy of Sciences of the United States of America, vol. 105,no. 2, pp. 652–656, 2008.

[6] J. B. Swann and M. J. Smyth, “Immune surveillance oftumors,” Journal of Clinical Investigation, vol. 117, no. 5, pp.1137–1146, 2007.

[7] M. J. Smyth, J. Swann, and Y. Hayakawa, “Innate tumorimmune surveillance,” Advances in Experimental Medicineand Biology, vol. 590, pp. 103–111, 2007.

[8] M. J. Smyth, G. P. Dunn, and R. D. Schreiber, “Cancerimmunosurveillance and immunoediting: the roles of immu-nity in suppressing tumor development and shaping tumorimmunogenicity,” Advances in Immunology, vol. 90, pp. 1–50,2006.

[9] D. I. Gabrilovich, V. Bronte, S.-H. Chen et al., “The termi-nology issue for myeloid-derived suppressor cells,” CancerResearch, vol. 67, no. 1, p. 425, 2007.

[10] J.R. Cubillos-Ruiz, M. Rutkowski, and J.R. Conejo-Garcia,“Blocking ovarian cancer progression by targeting tumor

microenvironmental leukocytes,” Cell Cycle, vol. 9, no. 2, pp.260–268, 2010.

[11] I. J. Fidler, J. Yoneda, C. Herrera, J. Wood, and L. Xu, “Specifickeynote: molecular determinants of angiogenesis in ovariancancer,” Gynecologic Oncology, vol. 88, no. 1, pp. S29–S37,2003.

[12] E. Barbera-Guillem, J. K. Nyhus, C. C. Wolford, C. R. Friece,and J. W. Sampsel, “Vascular endothelial growth factor secre-tion by tumor-infiltrating macrophages essentially supportstumor angiogenesis, and IgG immune complexes potentiatethe process,” Cancer Research, vol. 62, no. 23, pp. 7042–7049,2002.

[13] J. R. Conejo-Garcia, R. J. Buckanovich, F. Benencia et al.,“Vascular leukocytes contribute to tumor vascularization,”Blood, vol. 105, no. 2, pp. 679–681, 2005.

[14] J. R. Conejo-Garcia, F. Benencia, M.-C. Courreges et al.,“Tumor-infiltrating dendritic cell precursors recruited by aβ-defensin contribute to vasculogenesis under the influenceof Vegf-A,” Nature Medicine, vol. 10, no. 9, pp. 950–958, 2004.

[15] E. Huarte, J. R. Cubillos-Ruiz, Y. C. Nesbeth et al., “Depletionof dendritic cells delays ovarian cancer progression byboosting antitumor immunity,” Cancer Research, vol. 68, no.18, pp. 7684–7691, 2008.

[16] S. Ugel, F. Delpozzo, G. Desantis et al., “Therapeutic targetingof myeloid-derived suppressor cells,” Current Opinion inPharmacology, vol. 9, no. 4, pp. 470–481, 2009.

[17] S. Suzuki, S. Tone, O. Takikawa, T. Kubo, I. Kohno, andY. Minatogawa, “Expression of indoleamine 2,3-dioxygenaseand tryptophan 2,3-dioxygenase in early concepti,” Biochem-ical Journal, vol. 355, no. 2, pp. 425–429, 2001.

[18] D. H. Munn, M. Zhou, J. T. Attwood et al., “Prevention ofallogeneic fetal rejection by tryptophan catabolism,” Science,vol. 281, no. 5380, pp. 1191–1193, 1998.

[19] Y. Kudo and C. A. R. Boyd, “Human placental indoleamine2,3-dioxygenase: cellular localization and characterizationof an enzyme preventing fetal rejection,” Biochimica etBiophysica Acta, vol. 1500, no. 1, pp. 119–124, 2000.

[20] L. Dolcetti, E. Peranzoni, S. Ugel et al., “Hierarchy ofimmunosuppressive strength among myeloid-derived sup-pressor cell subsets is determined by GM-CSF,” EuropeanJournal of Immunology, vol. 40, no. 1, pp. 22–35, 2010.

[21] T. J. Curiel, S. Wei, H. Dong et al., “Blockade of B7-H1 improves myeloid dendritic cell-mediated antitumorimmunity,” Nature Medicine, vol. 9, no. 5, pp. 562–567, 2003.

[22] U. K. Scarlett, J. R. Cubillos-Ruiz, Y. C. Nesbeth et al.,“In situ stimulation of CD40 and Toll-like receptor 3transforms ovarian cancer-infiltrating dendritic cells fromimmunosuppressive to immunostimulatory cells,” CancerResearch, vol. 69, no. 18, pp. 7329–7337, 2009.

[23] S. Nagaraj, A. G. Schrum, H.-I. Cho, E. Celis, and D. I.Gabrilovich, “Mechanism of T cell tolerance induced bymyeloid-derived suppressor cells,” Journal of Immunology,vol. 184, no. 6, pp. 3106–3116, 2010.

[24] Y. Nesbeth, U. Scarlett, J. Cubillos-Ruiz et al., “CCL5-mediated endogenous antitumor immunity elicited by adop-tively transferred lymphocytes and dendritic cell depletion,”Cancer Research, vol. 69, no. 15, pp. 6331–6338, 2009.

[25] J. R. Cubillos-Ruiz, X. Engle, U. K. Scarlett et al.,“Polyethylenimine-based siRNA nanocomplexes reprogramtumor-associated dendritic cells via TLR5 to elicit therapeu-tic antitumor immunity,” Journal of Clinical Investigation, vol.119, no. 8, pp. 2231–2244, 2009.

[26] W. Zoul, V. Machelon, A. Coulomb-L’Hermin et al.,“Stromal-derived factor-1 in human tumors recruits and


alters the function of plasmacytoid precursor dendritic cells,”Nature Medicine, vol. 7, no. 12, pp. 1339–1346, 2001.

[27] D. Bell, P. Chomarat, D. Broyles et al., “In breast carcinomatissue, immature dendritic cells reside within the tumor,whereas mature dendritic cells are located in peritumoralareas,” Journal of Experimental Medicine, vol. 190, no. 10, pp.1417–1425, 1999.

[28] M. Iwamoto, H. Shinohara, A. Miyamoto et al., “Prognosticvalue of tumor-infiltrating dendritic cells expressing CD83 inhuman breast carcinomas,” International Journal of Cancer,vol. 104, no. 1, pp. 92–97, 2003.

[29] A. Troy, P. Davidson, C. Atkinson, and D. Hart, “Phenotypiccharacterisation of the dendritic cell infiltrate in prostatecancer,” Journal of Urology, vol. 160, no. 1, pp. 214–219, 1998.

[30] A. J. Troy, K. L. Summers, P. J. T. Davidson, C. H. Atkinson,and D. N. J. Hart, “Minimal recruitment and activation ofdendritic cells within renal cell carcinoma,” Clinical CancerResearch, vol. 4, no. 3, pp. 585–593, 1998.

[31] I. Kryczek, A. Lange, P. Mottram et al., “CXCL12 andvascular endothelial growth factor synergistically induceneonaniogenisis in human ovarian cancers,” Cancer Research,vol. 65, no. 2, pp. 465–472, 2005.

[32] A. Kadambi, C. M. Carreira, C.-O. Yun et al., “Vascularendothelial growth factor (VEGF)-C differentially affectstumor vascular function and leukocyte recruitment: role ofVEGF-receptor 2 and host VEGF-A,” Cancer Research, vol.61, no. 6, pp. 2404–2408, 2001.

[33] T. J. Curiel, P. Cheng, P. Mottram et al., “Dendritic cellsubsets differentially regulate angiogenesis in human ovariancancer,” Cancer Research, vol. 64, no. 16, pp. 5535–5538, 2004.

[34] P. Carmeliet and R. K. Jain, “Angiogenesis in cancer and otherdiseases,” Nature, vol. 407, no. 6801, pp. 249–257, 2000.

[35] P. Carmeliet, “VEGF as a key mediator of angiogenesis incancer,” Oncology, vol. 69, no. 3, pp. 4–10, 2005.

[36] P. Carmeliet, “Angiogenesis in health and disease,” NatureMedicine, vol. 9, no. 6, pp. 653–660, 2003.

[37] I. Kryczek, M. Grybos, L. Karabon, A. Klimczak, and A.Lange, “IL-6 production in ovarian carcinoma is associatedwith histiotype and biological characteristics of the tumourand influences local immunity,” British Journal of Cancer, vol.82, no. 3, pp. 621–628, 2000.

[38] D. I. Gabrilovich, S. Nadaf, J. Corak, J. A. Berzofsky, and D. P.Carbone, “Dendritic cells in antitumor immune responses II.Dendritic cells grown from bone marrow precursors, but notmature DC from tumor-bearing mice, are effective antigencarriers in the therapy of established tumors,” CellularImmunology, vol. 170, no. 1, pp. 111–119, 1996.

[39] D. I. Gabrilovich, H. L. Chen, K. R. Girgis et al., “Productionof vascular endothelial growth factor by human tumorsinhibits the functional maturation of dendritic cells,” NatureMedicine, vol. 2, no. 10, pp. 1096–1103, 1996.

[40] W. Zou, “Immunosuppressive networks in the tumour envi-ronment and their therapeutic relevance,” Nature ReviewsCancer, vol. 5, no. 4, pp. 263–274, 2005.

[41] S. Wei, I. Kryczek, R. P. Edwards et al., “Interleukin-2administration alters the CD4+FOXP3+ T-cell pool andtumor trafficking in patients with ovarian carcinoma,”Cancer Research, vol. 67, no. 15, pp. 7487–7494, 2007.

[42] M. Salio, M. Cella, W. Vermi et al., “Plasmacytoid den-dritic cells prime IFN-γ-secreting melanoma-specific CD8lymphocytes and are found in primary melanoma lesions,”European Journal of Immunology, vol. 33, no. 4, pp. 1052–1062, 2003.

[43] M. Dougan and G. Dranoff, “The immune response totumors,” Current Protocols in Immunology, no. 85, pp.20.11.1–20.11.4, 2009.

[44] M. Dougan and G. Dranoff, “Immune therapy for cancer,”Annual Review of Immunology, vol. 27, pp. 83–117, 2009.

[45] S.-K. Tey, C. M. Bollard, and H. E. Heslop, “Adoptive T-cell transfer in cancer immunotherapy,” Immunology and CellBiology, vol. 84, no. 3, pp. 281–289, 2006.

[46] S. A. Rosenberg, N. P. Restifo, J. C. Yang, R. A. Morgan,and M. E. Dudley, “Adoptive cell transfer: a clinical pathto effective cancer immunotherapy,” Nature Reviews Cancer,vol. 8, no. 4, pp. 299–308, 2008.

[47] M. E. Dudley, J. R. Wunderlich, P. F. Robbins et al.,“Cancer regression and autoimmunity in patients after clonalrepopulation with antitumor lymphocytes,” Science, vol. 298,no. 5594, pp. 850–854, 2002.

[48] R. A. Morgan, M. E. Dudley, J. R. Wunderlich et al., “Cancerregression in patients after transfer of genetically engineeredlymphocytes,” Science, vol. 314, no. 5796, pp. 126–129, 2006.

[49] N. N. Hunder, H. Wallen, J. Cao et al., “Treatment ofmetastatic melanoma with autologous CD4+ T cells againstNY-ESO-1,” The New England Journal of Medicine, vol. 358,no. 25, pp. 2698–2703, 2008.

[50] C. Rossig, C. M. Bollard, J. G. Nuchtern, C. M. Rooney,and M. K. Brenner, “Epstein-Barr virus-specific human Tlymphocytes expressing antitumor chimeric T-cell receptors:potential for improved immunotherapy,” Blood, vol. 99, no.6, pp. 2009–2016, 2002.

[51] C. U. Louis, K. Straath, C. M. Bollard et al., “Enhancing thein vivo expansion of adoptively transferred EBV-specific CTLwith lymphodepleting CD45 monoclonal antibodies in NPCpatients,” Blood, vol. 113, no. 11, pp. 2442–2450, 2009.

[52] A. A. Kennedy-Nasser, C. M. Bollard, and C. M. Rooney,“Adoptive immunotherapy for Hodgkin’s lymphoma,” Inter-national Journal of Hematology, vol. 83, no. 5, pp. 385–390,2006.

[53] J. I. Cohen, C. M. Bollard, R. Khanna, and S. Pittaluga,“Current understanding of the role of Epstein-Barr virusin lymphomagenesis and therapeutic approaches to EBV-associated lymphomas,” Leukemia and Lymphoma, vol. 49,no. 1, pp. 27–34, 2008.

[54] C. M. Bollard, B. Savoldo, C. M. Rooney, and H. E. Heslop,“Adoptive T-cell therapy for EBV-associated post-transplantlymphoproliferative disease,” Acta Haematologica, vol. 110,no. 2-3, pp. 139–148, 2003.

[55] C. M. Bollard, S. Gottschalk, A. M. Leen et al., “Completeresponses of relapsed lymphoma following genetic modifi-cation of tumor-antigen presenting cells and T-lymphocytetransfer,” Blood, vol. 110, no. 8, pp. 2838–2845, 2007.

[56] M. L. Disis, K. L. Knutson, K. Schiffman, K. Rinn, andD. G. McNeel, “Pre-existent immunity to the HER-2/neuoncogenic protein in patients with HER-2/neu overexpress-ing breast and ovarian cancer,” Breast Cancer Research andTreatment, vol. 62, no. 3, pp. 245–252, 2000.

[57] K. L. Knutson, W. Wagner, and M. L. Disis, “Adoptive T celltherapy of solid cancers,” Cancer Immunology, Immunother-apy, vol. 55, no. 1, pp. 96–103, 2006.

[58] J. Sprent and H. Kishimoto, “The thymus and centraltolerance,” Philosophical Transactions of the Royal Society B,vol. 356, no. 1409, pp. 609–616, 2001.

[59] J. Sprent and H. Kishimoto, “The thymus and negativeselection,” Immunological Reviews, vol. 185, pp. 126–135,2002.


[60] M. E. Dudley and S. A. Rosenberg, “Adoptive-cell-transfertherapy for the treatment of patients with cancer,” NatureReviews Cancer, vol. 3, no. 9, pp. 666–675, 2003.

[61] P. Straten and J. C. Becker, “Adoptive cell transfer in thetreatment of metastatic melanoma,” Journal of InvestigativeDermatology, vol. 129, no. 12, pp. 2743–2745, 2009.

[62] M. E. Dudley, J. C. Yang, R. Sherry et al., “Adoptive celltherapy for patients with metastatic melanoma: evaluationof intensive myeloablative chemoradiation preparative regi-mens,” Journal of Clinical Oncology, vol. 26, no. 32, pp. 5233–5239, 2008.

[63] T. Takayama, T. Sekine, M. Makuuchi et al., “Adoptiveimmunotherapy to lower postsurgical recurrence rates ofhepatocellular carcinoma: a randomised trial,” The Lancet,vol. 356, no. 9232, pp. 802–807, 2000.

[64] L. F. Porrata, M. R. Litzow, A. Tefferi et al., “Early lymphocyterecovery is a predictive factor for prolonged survival afterautologous hematopoietic stem cell transplantation for acutemyelogenous leukemia,” Leukemia, vol. 16, no. 7, pp. 1311–1318, 2002.

[65] L. F. Porrata, D. J. Inwards, I. N. Micallef, S. M. Ansell, S.M. Geyer, and S. N. Markovic, “Early lymphocyte recoverypost-autologous haematopoietic stem cell transplantation isassociated with better survival in Hodgkin’s disease,” BritishJournal of Haematology, vol. 117, no. 3, pp. 629–633, 2002.

[66] L. F. Porrata, D. J. Inwards, S. M. Ansell et al., “Early lympho-cyte recovery predicts superior survival after autologous stemcell transplantation in non-Hodgkin lymphoma: a prospec-tive study,” Biology of Blood and Marrow Transplantation, vol.14, no. 7, pp. 807–816, 2008.

[67] K. L. Knutson, K. Schiffman, and M. L. Disis, “Immunizationwith a HER-2/neu helper peptide vaccine generates HER-2/neu CD8 T-cell immunity in cancer patients,” Journal ofClinical Investigation, vol. 107, no. 4, pp. 477–484, 2001.

[68] M. A. Reddish, G. D. MacLean, R. Rao Koganty et al.,“Anti-MUC1 class I restricted CTLs in metastatic breastcancer patients immunized with a synthetic MUC1 peptide,”International Journal of Cancer, vol. 76, no. 6, pp. 817–823,1998.

[69] R. A. Morgan, M. E. Dudley, Y. Y. L. Yu et al., “High efficiencyTCR gene transfer into primary human lymphocytes affordsavid recognition of melanoma tumor antigen glycoprotein100 and does not alter the recognition of autologousmelanoma antigens,” Journal of Immunology, vol. 171, no. 6,pp. 3287–3295, 2003.

[70] L. Duval, H. Schmidt, K. Kaltoft et al., “Adoptive transfer ofallogeneic cytotoxic T lymphocytes equipped with a HLA-A2 restricted MART-1 T-cell receptor: a phase I trial inmetastatic melanoma,” Clinical Cancer Research, vol. 12, no.4, pp. 1229–1236, 2006.

[71] D. Eaton, D. E. Gilham, A. O’Neill, and R. E. Hawkins,“Retroviral transduction of human peripheral blood lym-phocytes with bcl-XL promotes in vitro lymphocyte survivalin pro-apoptotic conditions,” Gene Therapy, vol. 9, no. 8, pp.527–535, 2002.

[72] J. Charo, S. E. Finkelstein, N. Grewal, N. P. Restifo, P.F. Robbins, and S. A. Rosenberg, “Bcl-2 overexpressionenhances tumor-specific T-cell survival,” Cancer Research,vol. 65, no. 5, pp. 2001–2008, 2005.

[73] L. Borkner, A. Kaiser, W. van de Kasteele et al., “RNA inter-ference targeting programmed death receptor-1 improvesimmune functions of tumor-specific T cells,” CancerImmunology, Immunotherapy, vol. 59, no. 8, pp. 1173–1183,2010.

[74] K. Liu and S. A. Rosenberg, “Transduction of an IL-2gene into human melanoma-reactive lymphocytes results intheir continued growth in the absence of exogenous IL-2and maintenance of specific antitumor activity,” Journal ofImmunology, vol. 167, no. 11, pp. 6356–6365, 2001.

[75] C. Hsu, M. S. Hughes, Z. Zheng, R. B. Bray, S. A. Rosenberg,and R. A. Morgan, “Primary human T lymphocytes engi-neered with a codon-optimized IL-15 gene resist cytokinewithdrawal-induced apoptosis and persist long-term in theabsence of exogenous cytokine,” Journal of Immunology, vol.175, no. 11, pp. 7226–7234, 2005.

[76] C. A. Klebanoff, L. Gattinoni, and N. P. Restifo, “CD8+ T-cell memory in tumor immunology and immunotherapy,”Immunological Reviews, vol. 211, pp. 214–224, 2006.

[77] C. S. Hinrichs, R. Spolski, C. M. Paulos et al., “IL-2 and IL-21confer opposing differentiation programs to CD8+ T cells foradoptive immunotherapy,” Blood, vol. 111, no. 11, pp. 5326–5333, 2008.

[78] S. Kaneko, S. Mastaglio, A. Bondanza et al., “IL-7 and IL-15 allow the generation of suicide gene modified alloreac-tive self-renewing central memory human T lymphocytes,”Blood, vol. 113, no. 5, pp. 1006–1015, 2009.

[79] C. A. Klebanoff, L. Gattinoni, P. Torabi-Parizi et al., “Centralmemory self/tumor-reactive CD8+ T cells confer superiorantitumor immunity compared with effector memory Tcells,” Proceedings of the National Academy of Sciences of theUnited States of America, vol. 102, no. 27, pp. 9571–9576,2005.

[80] C. S. Hinrichs, Z. A. Borman, L. Cassard et al., “Adoptivelytransferred effector cells derived from naıve rather thancentral memory CD8+ T cells mediate superior antitumorimmunity,” Proceedings of the National Academy of Sciencesof the United States of America, vol. 106, no. 41, pp. 17469–17474, 2009.

[81] L. Gattinoni, X.-S. Zhong, D. C. Palmer et al., “Wnt signalingarrests effector T cell differentiation and generates CD8+

memory stem cells,” Nature Medicine, vol. 15, no. 7, pp. 808–813, 2009.

[82] D. L. Barber, E. J. Wherry, and R. Ahmed, “Cutting edge:rapid in vivo killing by memory CD8 T cells,” Journal ofImmunology, vol. 171, no. 1, pp. 27–31, 2003.

[83] C. H. June, “Principles of adoptive T cell cancer therapy,”Journal of Clinical Investigation, vol. 117, no. 5, pp. 1204–1212, 2007.

[84] L. Gattinoni, C. A. Klebanoff, D. C. Palmer et al., “Acquisitionof full effector function in vitro paradoxically impairs thein vivo antitumor efficacy of adoptively transferred CD8+

T cells,” Journal of Clinical Investigation, vol. 115, no. 6, pp.1616–1626, 2005.

[85] Y. Xie, A. Akpinarli, C. Maris et al., “Naive tumor-specificCD4+ T cells differentiated in vivo eradicate establishedmelanoma,” Journal of Experimental Medicine, vol. 207, no.3, pp. 651–667, 2010.

[86] Y. C. Nesbeth, D. G. Martinez, S. Toraya et al., “CD4+ Tcells elicit host immune responses to MHC class II-negativeovarian cancer through CCL5 secretion and CD40-mediatedlicensing of dendritic cells,” Journal of immunology, vol. 184,no. 10, pp. 5654–5662, 2010.

[87] P. Muranski and N. P. Restifo, “Adoptive immunotherapy ofcancer using CD4+ T cells,” Current Opinion in Immunology,vol. 21, no. 2, pp. 200–208, 2009.

[88] C. Blank, I. Brown, A. C. Peterson et al., “PD-L1/B7H-1 inhibits the effector phase of tumor rejection by T cell


receptor (TCR) transgenic CD8+ T cells,” Cancer Research,vol. 64, no. 3, pp. 1140–1145, 2004.

[89] L. Gattinoni, S. E. Finkelstein, C. A. Klebanoff et al.,“Removal of homeostatic cytokine sinks by lymphodepletionenhances the efficacy of adoptively transferred tumor-specificCD8+ T cells,” Journal of Experimental Medicine, vol. 202, no.7, pp. 907–912, 2005.

[90] D. J. Shedlock and H. Shen, “Requirement for CD4 T cell helpin generating functional CD8 T cell memory,” Science, vol.300, no. 5617, pp. 337–339, 2003.

[91] M. J. Bevan, “Helping the CD8+ T-cell response,” NatureReviews Immunology, vol. 4, no. 8, pp. 595–602, 2004.

[92] J. C. Sun, M. A. Williams, and M. J. Bevan, “CD4+ T cells arerequired for the maintenance, not programming, of memoryCD8+ T cells after acute infection,” Nature Immunology, vol.5, no. 9, pp. 927–933, 2004.

[93] J. C. Sun and M. J. Bevan, “Defective CD8 T cell memoryfollowing acute infection without CD4 T cell help,” Science,vol. 300, no. 5617, pp. 339–342, 2003.

[94] K. E. Foulds, L. A. Zenewicz, D. J. Shedlock, J. Jiang, A. E.Troy, and H. Shen, “Cutting edge: CD4 and CD8 T cells areintrinsically different in their proliferative responses,” Journalof Immunology, vol. 168, no. 4, pp. 1528–1532, 2002.

[95] H. Fujiwara, M. Fukuzawa, and T. Yoshioka, “The role oftumor-specific Lyt-1+2- T cells in eradicating tumor cells invivo. 1. Lyt-1+2- T cells do not necessarily require recruit-ment of host’s cytotoxic T cell precursors for implementationof in vivo immunity,” Journal of Immunology, vol. 133, no. 3,pp. 1671–1676, 1984.

[96] P. D. Greenberg, D. E. Kern, and M. A. Cheever, “Therapyof disseminated murine leukemia with cyclophosphamideand immune Lyt-1+,2- T cells. Tumor eradication doesnot require participation of cytotoxic T cells,” Journal ofExperimental Medicine, vol. 161, no. 5, pp. 1122–1134, 1985.

[97] D. Mumberg, P. A. Monach, S. Wanderling et al., “CD4+

T cells eliminate MHC class II-negative cancer cells in vivoby indirect effects of IFN-γ,” Proceedings of the NationalAcademy of Sciences of the United States of America, vol. 96,no. 15, pp. 8633–8638, 1999.

[98] A. Perez-Diez, N. T. Joncker, K. Choi et al., “CD4 cells can bemore efficient at tumor rejection than CD8 cells,” Blood, vol.109, no. 12, pp. 5346–5354, 2007.

[99] S. A. Quezada, T. R. Simpson, K. S. Peggs et al., “Tumor-reactive CD4+ T cells develop cytotoxic activity and eradicatelarge established melanoma after transfer into lymphopenichosts,” Journal of Experimental Medicine, vol. 207, no. 3, pp.637–650, 2010.

[100] R. E. M. Toes, S. P. Schoenberger, E. I. H. Van DerVoort, R. Offringa, and C. J. M. Melief, “CD40-CD40Ligandinteractions and their role in cytotoxic T lymphocyte primingand anti-tumor immunity,” Seminars in Immunology, vol. 10,no. 6, pp. 443–448, 1998.

[101] S. P. Schoenberger, R. E. M. Toes, E. I. H. Van Dervoort,R. Offringa, and C. J. M. Melief, “T-cell help for cytotoxicT lymphocytes is mediated by CD40-CD4OL interactions,”Nature, vol. 393, no. 6684, pp. 480–483, 1998.

[102] S. P. Schoenberger, L. E. Jonges, R. J. D. Mooijaart etal., “Efficient direct priming of tumor-specific cytotoxic Tlymphocyte in vivo by an engineered APC,” Cancer Research,vol. 58, no. 14, pp. 3094–3100, 1998.

[103] E. M. Janssen, E. E. Lemmens, T. Wolfe, U. Christen, M.G. Von Herrath, and S. P. Schoenberger, “CD4+ T cells arerequired for secondary expansion and memory in CD8+ Tlymphocytes,” Nature, vol. 421, no. 6925, pp. 852–856, 2003.

[104] K. Chan, D. J. Lee, A. Schubert et al., “The roles of MHC classII, CD40, and B7 costimulation in CTL induction by plasmidDNA,” Journal of Immunology, vol. 166, no. 5, pp. 3061–3066,2001.

[105] C. Yee, J. A. Thompson, D. Byrd et al., “Adoptive T celltherapy using antigen-specific CD8+ T cell clones for thetreatment of patients with metastatic melanoma: in vivopersistence, migration, and antitumor effect of transferred Tcells,” Proceedings of the National Academy of Sciences of theUnited States of America, vol. 99, no. 25, pp. 16168–16173,2002.

[106] W. Y. Ho, C. Yee, and P. D. Greenberg, “Adoptive therapy withCD8+ T cells: it may get by with a little help from its friends,”Journal of Clinical Investigation, vol. 110, no. 10, pp. 1415–1417, 2002.

[107] A. P. Rapoport, E. A. Stadtmauer, N. Aqui et al., “Restorationof immunity in lymphopenic individuals with cancer byvaccination and adoptive T-cell transfer,” Nature Medicine,vol. 11, no. 11, pp. 1230–1237, 2005.

[108] E. Azenshtein, G. Luboshits, S. Shina et al., “The CCchemokine RANTES in breast carcinoma progression: regu-lation of expression and potential mechanisms of promalig-nant activity,” Cancer Research, vol. 62, no. 4, pp. 1093–1102,2002.

[109] G. Soria, N. Yaal-Hahoshen, E. Azenshtein et al., “Con-comitant expression of the chemokines RANTES and MCP-1 in human breast cancer: a basis for tumor-promotinginteractions,” Cytokine, vol. 44, no. 1, pp. 191–200, 2008.

[110] H. Saji, M. Koike, T. Yamori et al., “Significant correlationof monocyte chemoattractant protein-1 expression withneovascularization and progression of breast carcinoma,”Cancer, vol. 92, no. 5, pp. 1085–1091, 2001.

[111] R. P. M. Negus, G. W. H. Stamp, J. Hadley, and F. R.Balkwill, “Quantitative assessment of the leukocyte infiltratein ovarian cancer and its relationship to the expression of C-C chemokines,” American Journal of Pathology, vol. 150, no.5, pp. 1723–1734, 1997.

[112] L. Zhang, J. R. Conejo-Garcia, D. Katsaros et al., “Intratu-moral T cells, recurrence, and survival in epithelial ovariancancer,” The New England Journal of Medicine, vol. 348, no. 3,pp. 203–213, 2003.

[113] M. J. Dobrzanski, K. A. Rewers-Felkins, I. S. Quinlin et al.,“Autologous MUC1-specific Th1 effector cell immunother-apy induces differential levels of systemic TReg cell sub-populations that result in increased ovarian cancer patientsurvival,” Clinical Immunology, vol. 133, no. 3, pp. 333–352,2009.

[114] X. Yao, M. Ahmadzadeh, S. A. Rosenberg, and P. F. Robbins,“Reconstitution of peripheral CD4+FOXP3+ regulatory Tcells in cancer patients receiving adoptive immunotherapyis related to tbe clinical response to therapy,” in AmericanAssociation of Immunologists 97th Annual Meeting, Baltimore,Md, USA, 2010.

[115] M. H. Kershaw, J. A. Westwood, L. L. Parker et al., “A phaseI study on adoptive immunotherapy using gene-modified Tcells for ovarian cancer,” Clinical Cancer Research, vol. 12, no.20, part 1, pp. 6106–6115, 2006.

[116] F. S. Hodi, M. C. Mihm, R. J. Soiffer et al., “Biologic activityof cytotoxic T lymphocyte-associated antigen 4 antibodyblockade in previously vaccinated metastatic melanoma andovarian carcinoma patients,” Proceedings of the NationalAcademy of Sciences of the United States of America, vol. 100,no. 8, pp. 4712–4717, 2003.


[117] F. S. Hodi, M. Butler, D. A. Oble et al., “Immunologicand clinical effects of antibody blockade of cytotoxic Tlymphocyte-associated antigen 4 in previously vaccinatedcancer patients,” Proceedings of the National Academy ofSciences of the United States of America, vol. 105, no. 8, pp.3005–3010, 2008.

[118] R. E. Roses, M. Xu, G. K. Koski, and B. J. Czerniecki, “Radia-tion therapy and Toll-like receptor signaling: implications forthe treatment of cancer,” Oncogene, vol. 27, no. 2, pp. 200–207, 2008.


Review Article

B7-H3 and Its Role in Antitumor Immunity

Martin Loos, Dennis M. Hedderich, Helmut Friess, and Jorg Kleeff

Chirurgische Klinik und Poliklinik, Klinikum Rechts der Isar, Technische Universitat Munchen, Ismaninger Straße 22,81675 Munchen, Germany

Correspondence should be addressed to Jorg Kleeff, [email protected]

Received 29 June 2010; Revised 16 September 2010; Accepted 19 October 2010

Academic Editor: C. D. Pauza

Copyright © 2010 Martin Loos et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

B7-H3 is one of the most recently identified members of the B7/CD28 superfamily of costimulatory molecules serving as anaccessory modulator of T-cell response. Recently, B7-H3 expression has been reported in several human cancers indicating anadditional function of B7-H3 as a regulator of antitumor immunity. However, its precise physiologic role is still elusive, becauseboth stimulatory and inhibitory capacities have been demonstrated. This paper summarizes the available data on B7-H3 in theregulation of T-cell response focusing on its potential role in antitumor immunity.

1. Introduction

T lymphocytes of the adaptive immune system are able torecognize and specifically respond to an incredible varietyof foreign and native antigens. To ensure an appropriate T-cell response, which is essential to eradicate pathogens andto maintain self-tolerance, T-cell activation is finely tuned bytwo independent signaling pathways. The first signal requiresrecognition of the antigen-bearing major histocompatibilitycomplex (MHC) on the surface of antigen-presenting cells(APCs) by the corresponding antigen-specific T-cell receptor(TCR) on T-cells. The second signal, which is antigenindependent, is delivered by costimulatory molecules ofthe B7/CD28 family. B7-1/B7-2:CD28/CTLA-4 signalingrepresents the best characterized costimulatory pathway [1].Engagement of B7-1 on APCs with CD28 on T-cells enhancesT-cell proliferation and IL-2 production. In the absence ofthis simultaneous costimulatory signal, ligation of the TCRby an antigenic peptide results in T-cell dysfunction, intoler-ance or anergy. Apart from stimulatory signals that augmentand sustain T-cell responses, costimulatory pathways alsodeliver inhibitory signals that downregulate or terminate T-cell responses [2]. Binding of CTLA-4 to B7-1 and/or B7-2inhibits IL-2 synthesis and progression through the cell cycleleading to the termination of T-cell response. Within thepast two decades, new costimulatory ligands and receptorshave been identified, including B7-H1 (programmed death-1

ligand-1), B7-DC (programmed death-1 ligand-2), PD-1 (programmed death-1), ICOS (inducible costimulator),ICOSL (ICOS-ligand), BTLA (B and T lymphocyte attenu-ator), B7-H3, and B7-H4 [3].

Recently, these previously identified B7 homologues havebeen implicated as potential regulators of antitumor immu-nity. For example, aberrant B7-H1 expression by cancercells has been associated with adverse pathologic featuresand poor outcome in different human malignancies and hastherefore been postulated as a potential mechanism by whichmalignant tumors may evade host immune response [4–8].Taking advantage of manipulation of costimulatory signalingby cancer cells is comprehensible as T-cells play an importantrole in antitumor immunity. Under normal conditions, APCsthat scavenge tumor cell debris and migrate to lymphoidtissues can interact with CD4+ and CD8+ T-cells to induceactivation of T-cells capable of recognizing tumor-specific ortumor-associated antigens. Thus, downregulation of tumor-specific T-cell responses by abusing inhibitory signalingpathways with induction of T-cell anergy or apoptosisthrough aberrant tumor B7-H1 expression may representa possible immune escape mechanism. Hence, immune-based therapies which eliminate inhibitory T-cell signalingmay represent a potent new approach for the treatment ofhuman malignancies. Indeed, several phase I/II trials usinghumanized monoclonal antibodies (mAbs) to block CTLA-4signaling have shown promising results in different human


cancers [9–12]. These studies provide evidence that treat-ment with anti-CTLA-4 mAbs is generally well tolerated andcapable of inducing objective tumor responses in patientswith prostate cancer, renal cell carcinoma, melanoma, andlymphoma [13–17].

B7-H3 is another recently identified costimulatorymolecule that has been implicated as a potential regulatorof antitumor response. However, its role in the regulationof T-cell response and in antitumor immunity remainscontroversial. This paper summarizes the existing data onthe immunological function of B7-H3 and focuses on thepotential role of B7-H3 in antitumor immunity.

2. B7-H3

2.1. Structure and Expression Pattern. B7-H3, identified in2001, is a type I transmembrane protein that shares 20%–27% amino acid identity with other B7 family members [18].Among the B7 family members, B7-H3 is the most conservedone with ∼88% amino acid identity between mice andhumans. While murine B7-H3 consists of a single extracel-lular variable-type immunoglobulin (Ig)V-IgC domain anda signature intracellular domain (2Ig B7-H3), human B7-H3possesses an additional isoform, the so-called 4Ig B7-H3 thatcontains a nearly exact tandem duplication of the IgV-IgCdomain [19, 20]. The 4Ig transcript is the dominant form inhuman tissues. So far, only one potential receptor of murineB7-H3 called triggering receptor expressed on myeloid cells(TREM-) like transcript 2 (TLT-2) has been identified.TLT-2 belongs to the TREM receptor family [21]. Thesereceptors function as modulators of cellular responses andplay important roles in both innate and adaptive immunities[22]. TLT-2 protein expression has been shown on CD8+

T-cells constitutively and is induced on activated CD4+ T-cells. Hashiguchi et al. recently found that binding of murineB7-H3 to TLT-2 especially on CD8+ T-cells enhances T-celleffector functions such as proliferation, cytokine production,and cytotoxicity. Blockade of the TLT-2:B7-H3 pathway bymAbs against B7-H3 or TLT-2 effectively inhibited both theinduction and effector phases of the contact hypersensitivityresponses. Although these in vitro data investigating theinteraction of murine B7-H3 with TLT-2 suggest that TLT-2 might function as a receptor for B7-H3, Leitner andcolleagues did not find evidence for such an interaction inboth mice and humans [23]. As an accessory costimulatorymolecule, B7-H3 protein is not constitutively expressedon T-cells, natural killer (NK) cells, and APCs, but itsexpression can be induced on these cell types. In contrastto B7-1 and B7-2 whose expressions are mainly limitedto immune cells such as APCs, B7-H3 protein is foundon osteoblasts, fibroblasts, fibroblast-like synoviocytes, andepithelial cells as well as in human liver, lung, bladder, testis,prostate, breast, placenta, and lymphoid organs. This broadexpression pattern suggests more diverse immunological andprobably nonimmunological functions of B7-H3, especiallyin peripheral tissues. Recently, B7-H3 expression has alsobeen found in a variety of different human cancers, includingprostate cancer, clear cell renal cell carcinoma (ccRCC),

non-small-cell lung cancer (NSCLC), pancreatic cancer,gastric cancer, ovarian cancer, colorectal cancer (CRC) andurothelial cell carcinoma [24–31]. Although these findingssuggest a possible involvement of B7-H3 in the regulation ofantitumor immunity, its exact role remains far from clear,because both stimulatory and inhibitory properties havebeen identified and both beneficial as well as adverse effectsof B7-H3 expression in cancers have been reported.

2.2. Functional Studies. Initial work on the functional prop-erties of B7-H3 showed a stimulatory effect of human B7-H3on T-cells. In vitro, B7-H3 was able to increase proliferationof both CD4+ and CD8+ T-cells, enhance the induction ofcytotoxic T lymphocytes (CTLs), and selectively stimulateinterferon-gamma (IFN-γ) production in the presence ofanti-CD3 Abs to mimic the TCR signal [18]. By contrast,inclusion of antisense B7-H3 oligonucleotides decreasedthe expression of B7-H3 on dendritic cells (DCs) andinhibited IFN-γ production by DC-stimulated allogeneic T-cells. Further functional data in support of a stimulatoryeffect of B7-H3 come from several in vivo studies. Car-diac allografts in treated B7-H3–/– mice showed markedlydecreased productions of key cytokine, chemokine, andchemokine receptor mRNA transcripts as compared to wild-type controls [32]. Moreover, the incidence of chronicrejection in two different cardiac allograft models wasalso inhibited in B7-H3–/– mice as compared to wild-type recipients. In a mouse model of allergic asthma,administration of anti-B7-H3 mAbs significantly reducedairway hyperactivity and resulted in decreased productionof Th2 cytokines (interleukin-4 (IL-4), IL-5, and IL-13) ascompared with control IgG-treated mice [33]. Furthermore,transfection of B7-H3 into mouse P815 tumor cells that wereinoculated into syngeneic DBA/2 mice resulted in completetumor regression of about one-half of the tumors andamplification of tumor-specific CD8+ CTL response [34].The authors concluded that B7-H3 transfection enhancedthe immunogenicity of the inoculated tumor cells. Similarly,injection of a mouse B7-H3 pcDNA3 expression plasmidinto EL-4 lymphomas led to complete regression of 50%of tumors or otherwise significantly slowed tumor growth[35]. B7-H3-driven antitumor immunity was mediatedby CD8+ T-cells and NK cells. In an orthotopic murinecolon cancer model, treatment by intratumoral injection ofan adenovirus expressing mouse B7-H3 (Ad-B7-H3-GFP)resulted in a reduction of tumor size compared to controlanimals [36]. In addition, the occurrence of secondarymetastasis was significantly reduced. Ad-B7-H3-GFP-treatedanimals showed significantly higher frequencies of tumor-specific IFN-γ producing CD8+ T-cells. Based on thesemouse cancer models, tumor-associated B7-H3 seems topreferentially regulate CD4-independent induction of CD8+

CTL responses. Further lines of evidence for a stimulatoryeffect of B7-H3 come from a murine hepatocellular carci-noma model. Intratumoral injection of B7-H3-expressingplasmids followed by vasostatin-expressing plasmid injection24 hours later resulted in a complete eradication of subcu-taneous H22 tumors [37]. Interestingly, neither B7-H3 nor


vasostatin monotherapy was effective. In contrast to thesefindings,which strongly suggest a stimulatory effect of B7-H3on T-cell responses and antitumor immunity, other groupshave proposed opposite functions for B7-H3. In mice, B7-H3protein inhibited T-cell activation and effector cytokine pro-duction [38]. Furthermore, an antagonistic mAb to B7-H3enhanced T-cell proliferation in vitro and led to exacerbatedexperimental autoimmune encephalomyelitis (EAE) in vivo[39]. In B7-H3–/– mice, Th1-mediated hypersensitivity andonset of EAE were promoted, and treatment with a blockinganti-B7-H3 mAb exacerbated EAE [38]. In a murine modelof experimental allergic conjunctivitis (EC), administrationof anti-B7-H3 mAbs during the induction phase augmentedthe severity of Th2-mediated EC [40]. In a different study,DC-associated B7-H3 induced by CD4+CD25+ regulatoryT-cells (Tregs) impaired T-cell stimulatory function in vivo[41]. Accordingly, B7-H3–/– mice developed EAE earlier aswild-type littermates. Moreover, B7-H3–/– mice developedmore severe airway inflammation under conditions in whichT helper cells differentiated toward Th1 rather than Th2. In adifferent in vitro study, B7-H3 inhibited T-cell proliferationof both CD4+ and CD8+ T-cells mediated by Ab to T-cellreceptor or allogeneic APCs [38].

2.3. Retrospective Analyses. In accordance to its inconsistentimmunologic function regarding the regulation of T-cellresponses which was demonstrated by several in vitro and invivo studies, the role of B7-H3 in human cancer remains farfrom clear.

Data in support of a possible stimulatory functionof B7-H3 in T-cell and antitumor responses come fromretrospective analyses in different human cancers. In gastriccancer, 58.8% of gastric cancer cells in a series of 102 patientshave been shown to express B7-H3 in the cell membraneand cytoplasm [26]. Tumor B7-H3 expression positivelycorrelated with survival time, infiltration depth, and tissuetype. In pancreatic cancer, high tumor B7-H3 expressionby cancer cells in 68 examined patients was significantlyassociated with prolonged patient survival after surgicalresection and significantly correlated with the number oftumor-infiltrating CD8+ T-cells [25].

However, several studies in other human cancers cor-relating tumor B7-H3 expression with clinicopathologicalfeatures do not concur with these findings. Tumor B7-H3expression in 70 patients with NSCLC inversely correlatedwith the number of tumor-infiltrating lymphocytes (TILs)and significantly correlated with lymph node metastasis [24].In a separate study, the level of circulating soluble B7-H3 (sB7-H3) in patients with NSCLC was associated withhigher tumor stage, tumor size, nodal metastasis, and distantmetastasis [42]. In ccRCC, 17.4% of tumor cells and 95.1%of tumor vasculature in 743 examined patients expressedB7-H3. B7-H3 expression in either tumor cells or tumorvasculature was found to significantly associate with anincreased risk of death from ccRCC [28]. In another clinicalstudy, B7-H3 was found to be uniformly and aberrantlyexpressed in adenocarcinomas of the prostate (n = 338).Marked B7-H3 intensity, which was found in approximately

20% of the examined specimens, was associated with a>4-fold increased risk of cancer progression after surgery[31]. Similar results were shown in another study thatinvestigated B7-H3 expression in 823 patients with prostatecancer. Tumor B7-H3 expression was found in 93% ofpatients treated with radical prostatectomy [43]. Strong B7-H3 expression in the resected specimens correlated withdisease spread and poor outcome. In CRC, strong B7-H3expression could be observed in 54.3% of 102 CRC patients.Higher B7-H3 expression positively correlated with a moreadvanced tumor grade and negatively correlated with theintensity of TILs [29]. Most recently, B7-H3 was found tobe expressed in 93% of 103 examined ovarian borderlinetumors and carcinomas. B7-H3 was also expressed in theendothelium of tumor-associated vasculature in 44% ofpatients. Carcinomas with B7-H3-positive tumor vasculaturewere associated with a significantly shorter survival time anda higher incidence of recurrence [27]. All relevant studiesregarding the clinical significance of B7-H3 in human cancerare summarized in Table 1 [24–31, 43–47].

2.4. Reasons for the Contrasting Immunomodulatory Effectsof B7-H3. Based on the latter experimental and clinicaldata, the role of B7-H3 in human cancer remains unclear.Several explanations for the seemingly conflicting data exist,including the possible existence of additional receptorsfor B7-H3. So far, only TLT-2 has been identified as apotential receptor for B7-H3 that seems to enhance T-celleffector function in mice [21]. However, B7-H3 might useanother receptor besides TLT-2 for its inhibitory function.Furthermore, it has to be considered that most of the existingfunctional studies have been performed in mouse modelsand that data regarding the immunomodulatory effects ofB7-H3 from these murine studies might not be transferableinto humans. Evidence for differences between mouse andhuman B7-H3 includes the existence of varying isoforms.In mice, the predominant isoform of B7-H3 consists of aclassical single extracellular IgV-IgC domain (2Ig B7-H3)whereas the predominant isoform in humans consists of adual IgV-IgC domain (4Ig B7-H3). It is also conceivablethat other additional isoforms may exist. Another possibleexplanation for the inconclusive data on the functional roleof B7-H3 in antitumor response comprises the level of B7-H3 expression on cancer cells that may be of relevance forthe induction of different immunological functions. Basedon its expression level, B7-H3 may interact with differentaffinities for several existing receptors and may thereforeexert different functions. A similar functional discrepancyregarding the expression levels of a costimulatory moleculein cancer has been shown for B7-1 [48]. Because of inhibitoryeffects on immune response, low B7-1 expression on murinecolon carcinoma cell lines has been implicated as a possibleimmune-escape mechanism for tumor cells, presumablythrough binding of the inhibitory T-cell receptor CTLA-4.By contrast, artificially enforced expression of B7-1 on thesetumor cells resulted in a strong increase of immunogenicity.The authors of this study concluded that the different effectof B7-1 may be explained by the noticeable higher affinity of


Table 1: Relevant clinical studies investigating the relationship between B7-H3 expression in human cancer tissues with clinicopathologicalfeatures.

Author Journal YearType ofmalignancy

Number ofpatients

Positivetumor-associated

Correlation with clinicopathologic features

B7-H3expression

Favorable Adverse

Sun et al.[29]

CancerImmunolImmunother

2010Colorectalcancer

102 87.3% —Higher tumor B7-H3 correlatedwith a more advanced tumor grade

Zang et al.[27]

Mod Pathol 2010Ovariancarcinoma

10393% of tumorcells

—

Significant shorter survival time andhigher incidence of recurrence forpatients with positive B7-H3 tumorvasculature

44% of tumorvasculature

Parker etal. [46]

Int J RadiatOncol Biol Phys

2010Recurrentprostate cancer

148 100% —

Increased risk of biochemicalrecurrence for patients withmoderate and marked B7-H3staining

Loos et al.[25]

BMC Cancer 2009Pancreaticcancer

68 88.2%

high tumorB7-H3expression wasassociated withsignificantlybetterpostoperativeprognosis

—

Boorjianet al. [44]

Urology 2009

Renal angiomy-olipoma/pulmonarylymphangio-leiomyomatosis

110/7100% and2.7%

— —

Yamatoet al. [47]

Br J Cancer 2009Pancreaticcancer

59 93.2% —

Strong tumor B7-H3 expression wassignificantly associated with lymphnode metastasis and advancedpathological stage

Crispenet al. [28]

Clin Cancer Res 2008 ccRCC 74317% of tumorcells —

Either tumor cell or diffusetumor vasculature B7-H3expression was significantlyassociated with an increased riskof death from ccRCC

95% of tumorvasculature

Boorjianet al. [30]

Clin Cancer Res 2008Urothelial cellcarcinoma

318 70.7% — —

Greorioet al. [45]

Histopathology 2008 Neuroblastoma 53 74% —High tumor B7-H3 expression wasassociated with a worse event-freesurvival

Zang et al.[43]

PNAS 2007 Prostate cancer 823 93% —

Patients with strong tumor B7-H3expression were at significantlyincreased risk of clinical cancerrecurrence and cancer-specific death

Roth et al.[31]

Cancer Res 2007 Prostate cancer 338 100% —

(1) Increasing levels of tumorB7-H3 intensity correlated withworsening clinicopathologicfeatures, including tumor volume,extraprostatic extension, higherGleason score, seminal vesicleinvolvement, surgical margins

(2) Marked tumor B7-H3 intensitywas significantly associated withcancer progression


Table 1: Continued.

Author Journal YearType ofmalignancy

Number ofpatients

Positivetumor-associated

Correlation with clinicopathologic features

B7-H3expression

Favorable Adverse

Sun et al.[24]

Lung Cancer 2006 NSCLC 70 37.1% —High tumor B7-H3 expression wassignificantly more common in caseswith lymph node metastasis

Wu et al.[26]

World JGastroenterol

2006 Gastric cancer 102 58.8%

Positive tumorB7-H3expression wassignificantlyassociated withbetterpostoperativesurvival

—

B7-1 for CTLA-4 which has been shown to be 100- to 1,000-fold higher than for CD28. Apart from the possible existenceof additional B7-H3 receptors, the definition of positive B7-H3 expression and B7-H3 expression levels in the availablestudies was not standardized. In the study analyzing the roleof B7-H3 in human gastric cancer, specimens were scoredas B7-H3-expressing tumors when more than 20% of tumorcells stained positive for B7-H3 [26]. Sun et al. defined lowtumor-B7-H3 expression in NSCLC when less than 10% oftumors expressed B7-H3 [24]. In this study, 37.1% of theexamined specimens expressed B7-H3. In ccRCC, tumorswith less than 10% of cells stained positive were scoredas having negative B7-H3 expression. 17% of specimensrevealed positive tumor cell B7-H3 expression [28]. Zanget al. did not define the difference between high or lowtumor B7-H3 expression in prostate cancer at all [43]. Innone of these studies, B7-H3 staining intensity was nottaken into account. In our recent study, the most detailedscoring system was used for expression analysis of B7-H3 inpancreatic cancer. Scores were given separately for the stainedarea and for the intensity of staining [25]. In addition to thedifferent definitions of positive B7-H3 expression and B7-H3 expression levels that were used in previous retrospectivestudies, a possible influence of soluble forms of B7-H3 wasnot examined. Based on a study analyzing sB7-H3 levelsin NSCLC that showed that the level of circulating sB7-H3was associated with higher tumor stage, tumor size, nodalmetastasis, and distant metastasis, one could speculate thatsB7-H3 may also contribute to the modulation of immuneresponse [42]. Another possible reason may include theexpression of aberrant forms of B7-H3 on tumor cells whichcannot be differentiated by the existing antibodies.Yi andChen recently worried that many so-called “neutralizingantibodies” may not be just blocking antibodies but haveother effects such as triggering the B7-H3 signal [49].Furthermore, genetic polymorphisms in B7-H3 may modifyT-cell responses in human cancers. Recent studies haveshown that polymorphisms in the inhibitory moleculeCTLA-4 alter cancer susceptibility through modification of

T-cell response [50]. Finally, B7-H3 may also affect otherimmune cells than T-cells. In neuroblastoma, 4Ig-B7-H3molecules expressed at the tumor cell surface have beenshown to exert a protective role from NK-mediated lysisby interacting with a still undefined inhibitory receptorexpressed on NK cells [19].

2.5. Therapeutic Potential of B7-H3. Immune-based thera-pies which additionally stimulate T-cell activation or elim-inate inhibitory T-cell signaling in order to enforce tumor-reactive T-cell responses represent a potent new approachfor the treatment of human malignancies. Blockade of theinhibitory receptor CTLA-4 by mAbs has been tested as asingle agent or in combinations in patients with advancedcancer, including breast cancer, lymphoma, melanoma,ovarian cancer, prostate cancer, and ccRCC [9, 15, 17, 51–54]. Most trials have not only shown that anti-CTLA-4mAb treatment is safe but also provide evidence for itsantitumor effects. In unresectable advanced melanoma forinstance, durable tumor responses and disease control rateshave been observed. PD-1 is another inhibitory receptorexpressed on activated T-cells that may suppress antitu-mor immunity. Therefore, single-agent anti-PD-1 blockadehas been tested in a Phase I trial of 39 patients withadvanced metastatic melanoma, CRC, NSCLC, castrate-resistant prostate cancer, and ccRCC. Anti-PD-1 treatmentresulted in one durable complete response and two partialresponses [55].

Given its immunomodulatory capacities, B7-H3 mayalso represent a new target in cancer treatment. In contrast toCTLA-4 and PD-1, however, one has to take into account thatB7-H3 is more broadly expressed, especially in peripheralhealthy tissues. Therefore, blockade of B7-H3 by mAbs ortreatment with B7-H3 (i.e., by gene transfer) may be associ-ated with severe adverse effects. Furthermore, the functionalrole of B7-H3 in antitumor immunity is not completelyunderstood, and controversies regarding its stimulatory andinhibitory capacities remain to be elucidated.


3. Conclusions

Within the past decade, new insights into immunomod-ulatory capacities of costimulatory signaling in antitumorresponse have opened the door for new potent approaches incancer therapy. Recent clinical Phase I/II trials have providedsolid evidence that treatment with anti-CTLA-4 mAbs iscapable of inducing objective antitumor responses. B7-H3,a recently identified member of the B7/CD28 superfamilyof costimulatory molecules, has been shown to play animportant role in immune regulation. Although data on theprecise role of B7-H3 in the regulation of T-cell responsesand especially in antitumor immunity has yet to be eluci-dated, B7-H3 represents a promising new target for immune-based antitumor therapies. The previous identification ofthe costimulatory receptor TLT-2 is the first significant steptoward resolving the available conflicting data.

However, further work particularly concerned with theidentification of inhibitory receptors for B7-H3 is ongoing.

References

[1] A. K. Abbas, “The control of T cell activation vs. tolerance,”Autoimmunity Reviews, vol. 2, no. 3, pp. 115–118, 2003.

[2] L. Chen, “Co-inhibitory molecules of the B7-CD28 family inthe control of T-cell immunity,” Nature Reviews Immunology,vol. 4, no. 5, pp. 336–347, 2004.

[3] R. J. Greenwald, G. J. Freeman, and A. H. Sharpe, “The B7family revisited,” Annual Review of Immunology, vol. 23, pp.515–548, 2005.

[4] M. Loos, N. A. Giese, J. Kleeff et al., “Clinical significance andregulation of the costimulatory molecule B7-H1 in pancreaticcancer,” Cancer Letters, vol. 268, no. 1, pp. 98–109, 2008.

[5] J. Konishi, K. Yamazaki, M. Azuma, I. Kinoshita, H. Dosaka-Akita, and M. Nishimura, “B7-H1 expression on non-smallcell lung cancer cells and its relationship with tumor-infiltrating lymphocytes and their PD-1 expression,” ClinicalCancer Research, vol. 10, no. 15, pp. 5094–5100, 2004.

[6] Y. Liu, B. Zeng, Z. Zhang, Y. Zhang, and R. Yang, “B7-H1 onmyeloid-derived suppressor cells in immune suppression bya mouse model of ovarian cancer,” Clinical Immunology, vol.129, no. 3, pp. 471–481, 2008.

[7] H. Ghebeh, S. Mohammed, A. Al-Omair et al., “The B7-H1 (PD-L1) T lymphocyte-inhibitory molecule is expressedin breast cancer patients with infiltrating ductal carcinoma:correlation with important high-risk prognostic factors,”Neoplasia, vol. 8, no. 3, pp. 190–198, 2006.

[8] R. H. Thompson, M. D. Gillett, J. C. Cheville et al., “Cos-timulatory B7-H1 in renal cell carcinoma patients: indicatorof tumor aggressiveness and potential therapeutic target,”Proceedings of the National Academy of Sciences of the UnitedStates of America, vol. 101, no. 49, pp. 17174–17179, 2004.

[9] E. J. Small, N. S. Tchekmedyian, B. I. Rini, L. Fong, I. Lowy,and J. P. Allison, “A pilot trial of CTLA-4 blockade with humananti-CTLA-4 in patients with hormone-refractory prostatecancer,” Clinical Cancer Research, vol. 13, no. 6, pp. 1810–1815,2007.

[10] E. D. Kwon, B. A. Foster, A. A. Hurwitz et al., “Elimination ofresidual metastatic prostate cancer after surgery and adjunc-tive cytotoxic T lymphocyte-associated antigen 4 (CTLA-4) blockade immunotherapy,” Proceedings of the National

Academy of Sciences of the United States of America, vol. 96, no.26, pp. 15074–15079, 1999.

[11] M. R. Theoret, P. M. Arlen, M. Pazdur, W. L. Dahut, J. Schlom,and J. L. Gulley, “Phase I trial of an enhanced prostate-specific antigen-based vaccine and anti-CTLA-4 antibodyin patients with metastatic androgen-independent prostatecancer,” Clinical Genitourinary Cancer, vol. 5, no. 5, pp. 347–350, 2007.

[12] C. I. Liakou, A. Kamat, D. N. Tang et al., “CTLA-4 blockadeincreases IFNγ-producing CD4+ICOS hi cells to shift the ratioof effector to regulatory T cells in cancer patients,” Proceedingsof the National Academy of Sciences of the United States ofAmerica, vol. 105, no. 39, pp. 14987–14992, 2008.

[13] R. H. Thompson, J. P. Allison, and E. D. Kwon, “Anti-cytotoxicT lymphocyte antigen-4 (CTLA-4) immunotherapy for thetreatment of prostate cancer,” Urologic Oncology, vol. 24, no.5, pp. 442–447, 2006.

[14] J. A. Blansfield, K. E. Beck, K. Tran et al., “CytotoxicT-lymphocyte-associated antigen-4 blockage can induceautoimmune hypophysitis in patients with metastaticmelanoma and renal cancer,” Journal of Immunotherapy, vol.28, no. 6, pp. 593–598, 2005.

[15] J. Yuan, S. Gnjatic, H. Li et al., “CTLA-4 blockadeenhances polyfunctional NY-ESO-1 specific T cell responses inmetastatic melanoma patients with clinical benefit,” Proceed-ings of the National Academy of Sciences of the United States ofAmerica, vol. 105, no. 51, pp. 20410–20415, 2008.

[16] D. O’Mahony, J. C. Morris, C. Quinn et al., “A pilot study ofCTLA-4 blockade after cancer vaccine failure in patients withadvanced malignancy,” Clinical Cancer Research, vol. 13, no. 3,pp. 958–964, 2007.

[17] S. M. Ansell, S. A. Hurvitz, P. A. Koenig et al., “Phase Istudy of ipilimumab, an anti-CTLA-4 monoclonal antibody,in patients with relapsed and refractory B-cell non-Hodgkinlymphoma,” Clinical Cancer Research, vol. 15, no. 20, pp.6446–6453, 2009.

[18] A. I. Chapoval, J. Ni, J. S. Lau et al., “B7-H3: a costimulatorymolecule for T cell activation and IFN-γ production,” NatureImmunology, vol. 2, no. 3, pp. 269–274, 2001.

[19] R. Castriconi, A. Dondero, R. Augugliaro et al., “Identificationof 4Ig-B7-H3 as a neuroblastoma-associated molecule thatexerts a protective role from an NK cell-mediated lysis,”Proceedings of the National Academy of Sciences of the UnitedStates of America, vol. 101, no. 34, pp. 12640–12645, 2004.

[20] P. Steinberger, O. Majdic, S. V. Derdak et al., “Molecularcharacterization of human 4Ig-B7-H3, a member of the B7family with four Ig-like domains,” Journal of Immunology, vol.172, no. 4, pp. 2352–2359, 2004.

[21] M. Hashiguchi, H. Kobori, P. Ritprajak, Y. Kamimura, H.Kozono, and M. Azuma, “Triggering receptor expressed onmyeloid cell-like transcript 2 (TLT-2) is a counter-receptorfor B7-H3 and enhances T cell responses,” Proceedings of theNational Academy of Sciences of the United States of America,vol. 105, no. 30, pp. 10495–10500, 2008.

[22] J. Klesney-Tait, I. R. Turnbull, and M. Colonna, “The TREMreceptor family and signal integration,” Nature Immunology,vol. 7, no. 12, pp. 1266–1273, 2006.

[23] J. Leitner, C. Klauser, W. F. Pickl et al., “B7-H3 is a potentinhibitor of human T-cell activation: no evidence for B7-H3and TREML2 interaction,” European Journal of Immunology,vol. 39, no. 7, pp. 1754–1764, 2009.

[24] Y. Sun, Y. Wang, J. Zhao et al., “B7-H3 and B7-H4 expressionin non-small-cell lung cancer,” Lung Cancer, vol. 53, no. 2, pp.143–151, 2006.


[25] M. Loos, D. M. Hedderich, M. Ottenhausen et al., “Expressionof the costimulatory molecule B7-H3 is associated withprolonged survival in human pancreatic cancer,” BMC Cancer,vol. 9, article 463, 2009.

[26] X.-P. Wu, J.-T. Jiang, M. Tan et al., “Relationship between co-stimulatory molecue B7-H3 expression and gastric carcinomahistology and prognosis,” World Journal of Gastroenterology,vol. 12, no. 3, pp. 457–459, 2006.

[27] X. Zang, P. S. Sullivan, R. A. Soslow et al., “Tumor associatedendothelial expression of B7-H3 predicts survival in ovariancarcinomas,” Modern Pathology, vol. 23, no. 8, pp. 1104–1112,2010.

[28] P. L. Crispen, Y. Sheinin, T. J. Roth et al., “Tumor cell andtumor vasculature expression of B7-H3 predict survival inclear cell renal cell carcinoma,” Clinical Cancer Research, vol.14, no. 16, pp. 5150–5157, 2008.

[29] J. Sun, L. -J. Chen, G. -B. Zhang et al., “Clinical significanceand regulation of the costimulatory molecule B7-H3 in humancolorectal carcinoma,” Cancer Immunology, Immunotherapy,vol. 59, no. 8, pp. 1163–1171, 2010.

[30] S. A. Boorjian, Y. Sheinin, P. L. Crispen et al., “T-Cellcoregulatory molecule expression in urothelial cell carcinoma:clinicopathologic correlations and association with survival,”Clinical Cancer Research, vol. 14, no. 15, pp. 4800–4808, 2008.

[31] T. J. Roth, Y. Sheinin, C. M. Lohse et al., “B7-H3 ligandexpression by prostate cancer: a novel marker of prognosis andpotential target for therapy,” Cancer Research, vol. 67, no. 16,pp. 7893–7900, 2007.

[32] L. Wang, C. C. Fraser, K. Kikly et al., “B7-H3 promotesacute and chronic allograft rejection,” European Journal ofImmunology, vol. 35, no. 2, pp. 428–438, 2005.

[33] O. Nagashima, N. Harada, Y. Usui et al., “B7-H3 contributesto the development of pathogenic Th2 cells in a murine modelof asthma,” Journal of Immunology, vol. 181, no. 6, pp. 4062–4071, 2008.

[34] L. Luo, A. I. Chapoval, D. B. Flies et al., “B7-H3 enhancestumor immunity in vivo by costimulating rapid clonalexpansion of antigen-specific CD8+ cytolytic T cells,” Journalof Immunology, vol. 173, no. 9, pp. 5445–5450, 2004.

[35] X. Sun, M. Vale, E. Leung, J. R. Kanwar, R. Gupta, and G.W. Krissansen, “Mouse B7-H3 induces antitumor immunity,”Gene Therapy, vol. 10, no. 20, pp. 1728–1734, 2003.

[36] C. M. Lupu, C. Eisenbach, A. D. Lupu et al., “Adenoviral B7-H3 therapy induces tumor specific immune responses andreduces secondary metastasis in a murine model of coloncancer,” Oncology reports, vol. 18, no. 3, pp. 745–748, 2007.

[37] L. Ma, L. Luo, H. Qiao et al., “Complete eradication of hepato-cellular carcinomas by combined vasostatin gene therapy andB7H3-mediated immunotherapy,” Journal of Hepatology, vol.46, no. 1, pp. 98–106, 2007.

[38] W.-K. Suh, B. U. Gajewska, H. Okada et al., “The B7 familymember B7-H3 preferentially down-regulates T helper type 1-mediated immune responses,” Nature Immunology, vol. 4, no.9, pp. 899–906, 2003.

[39] D. V. R. Prasad, T. Nguyen, Z. Li et al., “Murine B7-H3 is anegative regulator of T cells,” Journal of Immunology, vol. 173,no. 4, pp. 2500–2506, 2004.

[40] A. Fukushima, T. Sumi, K. Fukuda et al., “B7-H3 regulates thedevelopment of experimental allergic conjunctivitis in mice,”Immunology Letters, vol. 113, no. 1, pp. 52–57, 2007.

[41] K. Mahnke, S. Ring, T. S. Johnson et al., “Induction ofimmunosuppressive functions of dendritic cells in vivo byCD4+CD25+ regulatory T cells: role of B7-H3 expression and

antigen presentation,” European Journal of Immunology, vol.37, no. 8, pp. 2117–2126, 2007.

[42] G. Zhang, Y. Xu, X. Lu et al., “Diagnosis value of serum B7-H3expression in non-small cell lung cancer,” Lung Cancer, vol. 66,no. 2, pp. 245–249, 2009.

[43] X. Zang, R. H. Thompson, H. A. Al-Ahmadie et al., “B7-H3and B7x are highly expressed in human prostate cancer andassociated with disease spread and poor outcome,” Proceedingsof the National Academy of Sciences of the United States ofAmerica, vol. 104, no. 49, pp. 19458–19463, 2007.

[44] S. A. Boorjian, Y. Sheinin, P. L. Crispen, C. M. Lohse,B. C. Leibovich, and E. D. Kwon, “T-cell co-regulatorymolecule expression in renal angiomyolipoma and pulmonarylymphangioleiomyomatosis,” Urology, vol. 74, no. 6, pp. 1359–1364, 2009.

[45] A. Gregorio, M. V. Corrias, R. Castriconi et al., “Small roundblue cell tumours: diagnostic and prognostic usefulness of theexpression of B7-H3 surface molecule,” Histopathology, vol.53, no. 1, pp. 73–80, 2008.

[46] A. S. Parker, M. G. Heckman, Y. Sheinin et al., “Evaluation ofB7-H3 expression as a biomarker of biochemical recurrenceafter salvage radiation therapy for recurrent prostate cancer,”International Journal of Radiation Oncology, Biology, Physics. Inpress.

[47] I. Yamato, M. Sho, T. Nomi et al., “Clinical importance of B7-H3 expression in human pancreatic cancer,” British Journal ofCancer, vol. 101, no. 10, pp. 1709–1716, 2009.

[48] I. Tirapu, E. Huarte, C. Guiducci et al., “Low surfaceexpression of B7-1 (CD80) is an immunoescape mechanismof colon carcinoma,” Cancer Research, vol. 66, no. 4, pp. 2442–2450, 2006.

[49] K. H. Yi and L. Chen, “Fine tuning the immune responsethrough B7-H3 and B7-H4,” Immunological Reviews, vol. 229,no. 1, pp. 145–151, 2009.

[50] T. Sun, Z. Hu, H. Shen, and D. Lin, “Genetic polymorphismsin cytotoxic T-lymphocyte antigen 4 and cancer: the dialecticalnature of subtle human immune dysregulation,” CancerResearch, vol. 69, no. 15, pp. 6011–6014, 2009.

[51] M. Z. Dewan, A. E. Galloway, N. Kawashima et al., “Fraction-ated but not single-dose radiotherapy induces an immune-mediated abscopal effect when combined with anti-CTLA-4antibody,” Clinical Cancer Research, vol. 15, no. 17, pp. 5379–5388, 2009.

[52] A. V. Maker, G. Q. Phan, P. Attia et al., “Tumor regressionand autoimmunity in patients treated with cytotoxic Tlymphocyte-associated antigen 4 blockade and interleukin 2:a phase I/II study,” Annals of Surgical Oncology, vol. 12, no. 12,pp. 1005–1016, 2005.

[53] F. S. Hodi, M. C. Mihm, R. J. Soiffer et al., “Biologic activityof cytotoxic T lymphocyte-associated antigen 4 antibodyblockade in previously vaccinated metastatic melanoma andovarian carcinoma patients,” Proceedings of the NationalAcademy of Sciences of the United States of America, vol. 100,no. 8, pp. 4712–4717, 2003.

[54] H. J. J. van der Vliet, H. B. Koon, S. C. Yue et al., “Effects ofthe administration of high-dose interleukin-2 on immunoreg-ulatory cell subsets in patients with advanced melanoma andrenal cell cancer,” Clinical Cancer Research, vol. 13, no. 7, pp.2100–2108, 2007.

[55] J. R. Brahmer, C. G. Drake, I. Wollner et al., “Phase I studyof single-agent anti-programmed death-1 (MDX-1106) inrefractory solid tumors: safety, clinical activity, pharmaco-dynamics, and immunologic correlates,” Journal of ClinicalOncology, vol. 28, no. 19, pp. 3167–3175, 2010.


Research Article

Changes of Immunological Profiles in Patients withChronic Myeloid Leukemia in the Course of Treatment

Zuzana Humlova,1, 2 Hana Klamova,3 Ivana Janatkova,2 Karin Malıckova,2 Petra Kralıkova,4

Ivan Sterzl,1 Zdenek Roth,5 Eva Hamsıkova,6 and Vladimır Vonka6

1 Department of Immunology and Microbiology, 1st Medical Faculty, Charles University,and the General Teaching Hospital in Prague, Karlovo namestı 32, 121 11 Prague 2, Czech Republic

2 Department of Clinical Biochemistry and Laboratory Medicine, 1st Medical Faculty,Charles University, Karlovo namestı 32, 121 11 Prague 2, Czech Republic

3 Clinical Department, Institute of Hematology and Blood Transfusion, U Nemocnice 1, 128 20 Prague 2, Czech Republic4 Department of Immunology, 2nd Medical Faculty, Charles University, V Uvalu 84, 150 06 Prague 5, Czech Republic5 Department of Biostatistics, National Institute of Health, Srobarova 48, 100 00 Prague 10, Czech Republic6 Department of Experimental Virology, Institute of Hematology and Blood Transfusion, U Nemocnice 1,128 20 Prague 2, Czech Republic

Correspondence should be addressed to Zuzana Humlova, [email protected]

Received 28 June 2010; Revised 15 September 2010; Accepted 20 October 2010

Academic Editor: Stuart Berzins

Copyright © 2010 Zuzana Humlova et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

In the previous paper of ours we compared, prior to start any treatment, a number of immunological parameters in 24 chronicmyeloid leukemia patients with the same number of healthy subjects matched by age and sex. We found significant differencesin the levels of immunoglobulins, the C4 component of complement, the C-reactive protein, interleukin 6, the composition oflymphocyte population and the production of some cytokines by stimulated CD3+ cells. Eleven of these patients were followedlongitudinally. After treatment with hydroxyurea, interferon alpha, imatinib mesylate and dasatinib, or various combinationsthereof, hematological remission was achieved in all patients and complete cytogenetic remission in nine of them. There wasa nearly general tendency towards normalization of the abnormalities observed in the patients at their enrollment.

1. Introduction

The treatment of chronic myeloid leukemia (CML) nowoffers several options from which to choose. Hydroxyurea(HU) was introduced in the late 1960s and for decadesremained the mainstay of palliation in CML. However,HU does not induce cytogenetic remissions in a significantpercentage of patients nor does it markedly change thenatural history of the disease. The adverse effects includegastrointestinal problems and cutaneous defects as leg ulcers[1], hyperpigmentation of the skin and nails, a lichen planus-like eruption, lupus erythematosus, and dermatomyositis-like eruption [2]. The first observational reports on a cytore-ductive effect of interferon α (IFNα) in CML patients dateback to 1980s, when IFNα treatment was introduced at

the M.D. Anderson Cancer Center, Houston, Texas [3, 4].IFNα induces durable major and even complete cytogeneticremissions (CCR) persisting for months, sometimes even foryears [5]. IFNα not only mediates antileukemic responses viainduction of T-cell immunity [6, 7], but it also promoteshumoral immunity against CML antigens [8]. Some param-eters of innate immunity, which apparently plays a role inanticancer immunity, are also favorably influenced by IFNα[9, 10]. This might elucidate the efficacy of IFNα treatment invivo by orchestrating a network of immune cells rather thanby the activation of individual populations. Other mecha-nisms involved in modulating the course of the disease byIFNα are connected with its antiproliferative effect. However,long-term treatment with IFNα can also produce or exac-erbate immune-mediated complications [11, 12], such as


cutaneous vasculitis, hemolytic anemia, thyroid gland disor-ders, immune-mediated thrombocytopenia, nephrotoxicity,pemphigus foliaceus, rheumatoid arthritis, systemic lupuserythematosus, and even heart dysfunction based probablyon immune mechanisms [11]. A revolution into therapy ofCML has been brought by the introduction of the so-calledtargeted drugs. The first of these disease-tailored productshas been imatinib mesylate (IM) which blocks the ATP-binding pocket on the BCR-ABL tyrosine-kinase and thusprevents the activation of this enzyme which plays the keyrole in the pathogenesis of CML [13]. IM has been reportedto have induced CCR in 74% of the newly diagnosed patientsand is also active in patients previously treated with INFα[14]. According to a recent update, a five-year survival hasbeen achieved in nearly 90% of CML patients [15]. However,in a portion of patients, resistance to the drug developsmostly due to the mutations in the enzyme catalytic domain[16] or as a consequence of the amplification of the bcr-abl fusion gene [17]. To deal with the problem, a newgeneration of targeted drugs is being introduced and someof its representatives are already in clinical use, for example,dasatinib [18] or nilotinib [19].

Still, neither of these drugs can cure the disease mostprobably due to their failure to hit the quiescent cancerstem cells. When the treatment is interrupted, the diseaserelapses. Many oncohematologists believe that the problemof curing CML might be unriddled by supplementingthe chemotherapy with immunotherapeutic approaches. Amathematical model has been constructed suggesting thatimmunotherapeutic intervention tailored to the clinicalcondition and the underlying immune status of the patientmay result in the cure of CML [20].

Although the role of immune reactions in the courseof CML has been demonstrated beyond reasonable doubt,the first vaccine trials reported in the past 10 years havenot been particularly successful (for review see [21]). Weare of the opinion that to achieve the immunization goal itwill be necessary to augment our present knowledge on theimmunology of CML patients and that very likely this willlead to appreciable progress in the future immunotherapeu-tic undertakings.

It was the purpose of the present study to constructimmunological profiles of CML patients by testing severalparameters of their innate immunity early after diagnosis,that is, prior to the start of any therapy and then tofollow the influence of different therapeutic regimens onthese parameters and the association of their changes withthe clinical condition. In a previous paper of ours [22],representing the first part of the present study, we reportedthe findings obtained in 24 CML patients before the startof any therapy and in the same number of matched healthysubjects. We found a number of deviations from the norm inthe immune reactivity of CML patients and significant dif-ferences between the patients’ and control groups. The maindifferences encountered in the patients were represented byincreased levels of IgA (P < .02), the C4 component ofcomplement (P < .05), C-reactive protein (CRP) (P < .02),and interleukin 6 (IL-6) (P < .0005). Furthermore, a highly

significantly decreased production of interleukin-2 (IL-2)(P < .0001) and tumor necrosis factor α (TNFα) (P < .001)in stimulated CD3+ lymphocytes and a decreased phagocy-tosis of killed E. coli by polymorphonuclears (P < .0001)were observed. In spite of the frequency of these aberrations,no consistent pattern which might be characteristic for CMLwas revealed. In the subsequent follow-up we unfortunatelymet considerable difficulties. First of all, we lost the majorityof patients (13 of 24) for various reasons. Several of themwere transplanted, a few moved out of Prague or even left thecountry, and some simply lost their interest in participatingin the study. Another complication resulted from the dra-matic progress in the therapy of CML due to the introductionof a new generation of highly effective and relatively verywell-tolerated new drugs (see above). From ethical reasons,it was necessary to substitute these new drugs for the olderones. It followed that most of the patients were treated withmore than one drug.

2. Materials and Methods

2.1. Patients. The basic data on 11 patients enrolled at thetime of diagnosis (i.e., before starting any therapy) whoremained in the study throughout are shown in Table 1.The group consisted of 5 males and 6 females, theirmedian age was 48 years, and their age range from 33 to62 years. The follow-up lasted for 44 to 58 months. Allpatients were treated either with imatinib mesylate (IM) ordasatinib (DS), but not in all of them were these drugs usedas the starting therapy. In six patients, administration oftyrosine-kinase inhibitors (TKI) was preceded by interferonα (IFNα) treatment and in five patients the administrationof IFNα followed the initial treatment with hydroxyurea(HU). In two patients, HU was given prior to IM treatment.Hematological remission was achieved in all patients andcomplete cytogenetic remission (CCR) was achieved in nineof them in the course of the observation period. This wasassociated with the decrease of lymphocyte count in nearlyall the patients. The only exception was patient no. 6, withleukopenia prior to the start of the therapy, in whom theremission was associated with an increase of lymphocytecount up to the norm. At the end of the observation period,thrombocyte count was below the level of 450 × 109/L inall patients. In eight of the patients the therapy was notassociated with any complication. Patient no. 1, who hadoriginally been treated with HU and who had developedsymptoms of immunodeficiency (neutropenia, aphthousstomatitis, repeated infections of the upper and lower respi-ratory tract) in the course of the observation period, under-went a supportive therapy with growth factors (Neupogen300 μG/week) and immunoglobulins (Pasteurised HumanImmunoglobulin Grifols 16% 5 mL i.m. once a week for3 weeks). Moreover, erythematous eruptions diagnosed aserythema nodosum vasculitis developed on her feet. As anadditional therapy, she received routine anti-inflammatorydrugs and local corticosteroids. In this patient CCR was notachieved in spite of the IM dose having been raised to 600 mgper day.


In patients nos. 6 and 7, laboratory tests revealed autoan-tibodies against thyroidal peroxidase and thyreoglobulinwithout clinical signs of hyper- or hypothyroidism. Afterswitching to IM or DS, the laboratory findings normalized(see the Results section).

2.2. Drugs Used. Patients were treated as specified in Table 1.LITALIR (HU—Hydroxycarbamidum 500 mg, Bristol-MyersSquibb, Ltd, Prague, Czech Republic), ROFERON-A (INF-α-Interferon α 2a 18 MIU/0.6 mL inj. sol., Roche, Ltd, Prague,Czech Republic), Glivec (imatinib-mesylate Glivec, 478 mg,Novartis Europharm Ltd., Horsham, West Sussex, GreatBritain), and SPRYCEL (Dasatinib 100 mg, Bristol-MyersSquibb Pharma Eeig, Uxbridge Business Park, SandersonRoad, Uxbridge, UK) were used. The dosage of the drugsadministered is indicated in Table 1.

2.3. Blood Samplings. Before sampling, written InformedConsent was obtained from all patients and the study wasapproved by the Ethical Committees of the institutionsconcerned. In addition to samples taken for routine hemato-logical and biochemical testing, materials for immunologicalassays were obtained from each subject. The first sampleswere taken prior to the start of any therapy. Subsequentsamples were usually collected before the change in therapy.The intervals varied (with one exception, patients no. 8)from 6 to 30 months. The blood samples were distributedinto tubes purchased from BD Vacutainer Systems, BelliverIndustrial Estate, Plymouth, UK, as follows: (i) 7 mL of bloodwere taken into Z tubes, coagulated, and the serum wasused for humoral immunity tests; (ii) 2 mL of blood weremixed with EDTA for immunophenotypic analysis; (iii) 2 mLof the whole blood were mixed with sodium heparine forthe measuring of intracellular cytokines. The materials weretested (see below) immediately after arrival at the laboratory,except the viral antibodies, CRP and IL-6 (see below) forwhich all sera were tested simultaneously. Portions of thematerials were preserved for further tests. Sera were storedat −20◦C and leukocyte suspensions were stored in liquidnitrogen for possible future tests.

2.4. Immunoglobulins. The levels of the total IgG, of the IgGsubclasses, of IgA, and IgM were measured by nephelometryas described previously [22]. The standard laboratory refer-ential ranges are 6.9–14.0 g/L for IgG, 4.9 –11.4 g /L for IgG1,1.5–6.4 g/L for IgG2, 0.2–1.1 g/L for IgG3, 0.08–1.4 g/L forIgG4, 0.7–3.7 g/L for IgA, and 0.34–2.4 g/L for IgM.

2.5. Autoantibodies. Antibodies against thyroidal peroxidase(ATPOAb), thyreoglobulin (ATGAb), nucleus (ANAb), mito-chondria (AMAb), smooth muscles (ASMAb), cytoplasmof neutrophils (ANCAb), and endomysium (AE-GAb, AE-AAb; G refers to IgG and A refers to IgA antibody,resp.) were detected by indirect immunofluorescence asdescribed previously [22]. We have also tested the presenceof antidesmosomal antibodies (ADESAb). Only the presenceor absence of antibodies was monitored, not their titers.

2.6. Complement. The levels of the C3 and C4 componentsof complement were measured by nephelometry as describedpreviously [22]. The standard laboratory referential rangesare 0.75–1.4 g/L for C3 and 0.10–0.34 g/L for C4.

2.7. C-Reactive Protein. The levels of the C-reactive protein(CRP) were measured by nephelometry as described previ-ously [22]. The standard laboratory reference range is 0.00–5.0 mg/L.

2.8. Interleukin-6. For the determination of interleukin 6(IL-6), the quantitative sandwich enzyme immunoassaytechniques was used as described previously [22]. Thestandard laboratory referential range is 3.13–12.5 μg/L.

2.9. Subpopulations of Lymphocytes. Immunophenotypicanalysis of lymphocytes was performed using monoclonalantibodies directed against the following human surfaceantigens: CD3, CD4, CD8, CD19, CD16, and CD56 by flowcytometry as described previously [22]. The sum of cellsstained with CD4, CD8, CD19, and CD16-CD56 antibodieswas considered 100%. The standard laboratory referentialranges are: 59%–84% for CD3+, 25%–59% for CD4+, 19%–48% for CD8+, 6%–22% for CD19+, 6%–30% for CD16+,CD56+ cells.

2.10. Intracellular Cytokine Production by Stimulated CD3+Cells. Intracellular production of interleukin-2 (IL-2) inter-leukin-4 (IL-4), tumor necrosis alpha (TNFα), and inter-feron gamma (INFγ) in CD3+ cells stimulated by the mix-ture of brefeldinA and phorbol-12-myristate-13-acetate wasmonitored by flow cytometry as described previously [22].The figures shown in the Results section indicate thepercentages of CD3+ producing the individual cytokines.

2.11. Antibodies against Herpesviruses and Human Papillo-maviruses. Antibodies were determined against the follow-ing herpesviruses: the herpes simplex virus type 1 and type 2(HSV 1 and 2, IgG and IgM), the varicella-zoster virus (VZV,IgG and IgM), the human cytomegalovirus (CMV, IgG andIgM) and the EB virus (EBV, IgG and IgM against the viruscapsid antigen [VCA], IgG against the virus nuclear antigen[EBNA1] and against the early antigens [EA R+D]). Thefollowing commercial kits were used: ETI-HSVK-G-1/2 andETI-HSVK-M-1/2 (DiaSorin S.p.A., Italy) for the detectionof HSV 1/2 IgG and IgM antibody, respectively; ETI-CYTOK-G Plus and ETI-CYTOK-M REV Plus (DiaSorinS.p.A., Italy ) for the detection of CMV IgG and IgMantibody, respectively; VZV IgG and VZV IgM (Nova Tec,Immunodiagnostica, GmbH Germany) for the detection ofVZV IgG and IgM antibody, respectively; ETI-EBV VCA-Gand ETI-EBV-VCA-M-Rev (DiaSorin S.p.A., Italy) for thedetection of EBV VCA IgG and IgM antibody, respectively;ETI-EBNA-G (DiaSorin S.p.A., Italy) for the detection ofEBV EBNA IgG antibody; and ETI-EA-G (DiaSorin S.p.A.,Italy) for the detection of EBV EA IgG antibody. In addition,antibodies to six types of human papillomaviruses (HPV),namely, types 6, 11, 16, 18, 31, and 33 were determined


Table 1: Patients followed.

No. Gender Age1 Treatment Hematologic

Mo2 Drug Dosage/d WBC× 109/L Lympho× 109/L Hb (gr/L) Trc× 109/L HR3 CCR3

(1) F 55

0 dg, HU 2000 mg 302.6 4.60 89 413

19 No

8 IFN 5 MIU 16.2 1.45 130 188

16 IFN 5 MIU 69.0 3.43 131 442

22 IM 400 mg 10.5 2.86 124 399

27 IM 400 mg 7.4 1.93 120 505

39 IM 600 mg 4.0 1.11 120 175

58 DS 100 mg 2.1 0.89 119 139

(2) F 62

0 dg, HU 2500 mg 120.0 3.54 132 226

6 19

6 IFN 3 MIU 4.7 1.49 151 163

14 IFN 3 MIU 5.1 1.20 131 352

15 IFN 5 MIU 5.8 0.91 140 356

25 IM 400 mg 3.2 1.12 138 157

43 IM 400 mg 7.0 1.38 130 430

56 IM 400 mg 9.3 1.61 130 648

(3) M 40

0 dg, HU 3000 mg 21.9 2.07 168 327

8 26

7 IFN 3 MIU 8.1 1.08 163 262

15 IFN 5 MIU 3.1 0.85 160 125

26 IM 400 mg 4.2 1.10 146 246

43 IM 400 mg 4.5 1.37 147 209

58 IM 400 mg 3.9 1.01 158 210

(4) M 60

0 dg, HU 2000 mg 198.5 5.06 110 550

6 36

6 IFN 3 MIU 3.4 0.86 132 115

12 IFN 5 MIU 3.4 1.21 137 152

42 IM 400 mg 5.5 1.52 131 301

56 IM 400 mg 3.5 1.56 136 311

(5) M 33

0 dg, HU 2500 mg 99.7 5.91 142 169

2 40

6 HU 2000 mg 9.9 3.56 145 151

9 IFN 3 MIU 47.8 3.12 137 102

40 IM 400 mg 4.3 1.69 142 202

53 IM 400 mg 5.5 2.01 140 258

58 IM 400 mg 6.6 1.88 136 266

(6) F 35

0 dg, IFN 3 MIU 107.5 0.48 117 276

2 35

10 IFN 3 MIU 4.4 0.76 101 182

12 IM 400 mg 2.8 1.12 102 145

36 DS 100 mg 2.9 1.61 89 233

48 DS 100 mg 2.8 1.83 89 331

(7) M 54

0 dg, HU 2000 mg 16.4 4.86 136 595

11 38

7 HU 2000 mg 4.5 5.71 137 532

20 HU 1000 mg 5.5 3.21 134 209

34 IM 400 mg 6.9 1.23 134 206

47 IM 400 mg 6.8 2.19 132 236


Table 1: Continued.

No. Gender Age1 Treatment Hematologic

Mo2 Drug Dosage/d WBC× 109/L Lympho× 109/L Hb (gr/L) Trc× 109/L HR3 CCR3

(8) F 43

0 dg, IM 400 mg 26.9 4.05 144 464

2 17

2 IM 400 mg 8.3 1.66 129 228

9 IM 400 mg 6.0 1.76 123 185

17 IM 400 mg 7.5 1.26 126 207

31 IM 300 mg 7.6 1.73 120 239

45 IM 300 mg 8.3 1.45 125 315

(9) M 48

0 dg, IM 400 mg 439.2 6.16 81 184

7 No30 IM 400 mg 4.4 1.96 150 117

44 IM 400 mg 5.4 1.41 149 112

(10) F 38

0 dg, IM 400 mg 328.6 1.81 88 614

5 277 IM 400 mg 3.0 1.32 121 200

28 IM 400 mg 3.3 1.10 94 212

36 IM 400 mg 4.9 0.98 103 236

(11) F 61

0 dg,HU 2000 mg 16.8 5.31 109 1895

4 2529 DS 100 mg 5.8 2.02 100 519

52 DS 100 mg 5.4 2.75 104 2491Age at enrollment; 2Mo: month; WBC: white blood cell count; Hb: hemoglobin; and Trc, thrombocyte count; HU: hydroxyurea; IFN: interferon-alpha 2a;

IM: imatinib mesylate; DS: dasatinib; MIU, Millions International Units. 3The figure indicates month after diagnosis at which HR or CCR was first observed;“no” means that CCR was not achieved.Notes: Patient no. 1 developed vasculitis (erythema nodosum), neutropenia which was subsequently treated with growth factors and immunoglobulins.Patients 6 and 7 had laboratory tests positive for autoimmune thyroiditis without any clinical manifestations.Patient no. 11 decided to undergo homeopathic therapy and self-therapy with HU and was out of the evidence for some time. At the indicated interval, DStherapy was started because of pathological findings both in peripheral blood and bone marrow.

using ELISA using virus-like particles as antigen as describedpreviously [23]. Sera were diluted according to correspond-ing manufacturer recommendation, in case of anti-HPVantibodies 1 : 25. All samples originating from the samepatient were tested simultaneously on one microplate.

2.12. Statistical Methods. In individual patients, Spear-man’s correlation coefficient was used for quantifying themonotony of trend in time of the measured items. Thestatistical difference P for the deviation from zero wascalculated. When analyzing the hematologic findings for thewhole group of patients, the data transformed to logarithmswere analyzed by covariance analysis testing for differencesamong patients and the common linear regression onto thetime of treatment. The linear function was used as a basiccomponent of the experienced time trend. For comparison ofthe relative percentage distribution of different lymphocytepopulations in the first and last samples available, χ2 testwas used. For evaluation of the production of cytokines bystimulated CD3+ cells, the mean ratio of the last and firstsamples available was tested by One Sample Student t-testfor difference from 1. P-values have not been adjusted formultiple comparisons.

3. Results

3.1. Immunoglobulins Levels. The results of measuring theimmunoglobulins levels are summarized in Table 2. Nearly

generally, there was a tendency to a decrease of their levels.The decrease of total IgG for the whole group was just onthe brink of statistical significance (P = .05). In two patients(nos. 2 and 7) the IgG levels dropped below the norm.On the other hand, a marked increase of IgGs was observedin one patient (No. 6, treated gradually with IFNα, IMand DS) who had had the lowest level of total IgG andpathologically low levels of IgG1, IgG2, IgG3 and IgG4 inher pretreatment serum. The treatment resulted in theirrestoration up to the norm. It may be of interest that in thispatient the hemoglobin level dropped in the course of theobservation period (see Table 1). The most frequent changeswere in the IgM levels. Their slight or moderate decrease wasobserved in 9 patients and usually it was most marked aftertreatment with IM. The decrease of IgM levels for the wholegroup was significant (P = .011). Changes in IgA levels wereseen less frequently. Its levels dropped significantly with timein only one patient (no. 2) and increased in other three (nos.6, 9 and 11). The first one of these was the already mentionedpatient no. 6 with an increase in all subclasses of IgG. Itis noteworthy that throughout the observation period thechanges in this particular patient were not associated withany marked variation in the IgM level.

3.2. Presence of Autoantibodies. The presence of autoanti-bodies is shown in Table 3. Again, no consistent pattern isapparent. Autoantibodies, which had been present in thepretreatment sera of only three patients (twice against TPO


Table 2: Imunoglobulins.

No. Therapy IgG IgG1 IgG2 IgG3 IgG4 IgA IgM

(1)

dg, HU 10.60 6.73 4.29 0.686 0.526 2.16 1.65

IFN 11.10 6.14 3.74 0.767 0.364 1.95 1.50

IFN 9.30 5.97 3.22 0.530 0.250 1.93 1.56

IM 12.30 8.90 3.99 0.709 0.119 2.74 1.91

IM 9.40 6.82 3.34 0.575 0.157 2.17 1.61

IM 11.30 6.14 4.21 0.687 0.193 2.92 1.54

DS 11.40 6.70 3.81 0.907 0.253 2.73 1.29

(2)

dg, HU 10.00 6.72 3.34 0.418 0.460 2.31 0.76

IFN 8.58 5.00 3.34 0.353 0.339 2.12 0.72

IFN 8.53 5.64 3.00 0.358 0.397 1.68 0.45

IFN 9.39 4.50 3.13 0.320 0.247 1.50 0.40

IM 8.49 4.41 2.75 0.292 0.259 1.59 0.33

IM 6.78 3.98 2.18 0.312 0.179 1.35 0.20

IM 6.41∗∗ 3.91∗∗ 2.24∗∗ 2.92∗∗ 1.52∗∗ 1.23∗∗ 0.23∗∗

(3)

dg, HU 10.00 6.41 3.03 0.225 0.079 1.79 1.44

IFN 9.19 5.83 2.74 0.196 0.082 1.66 0.83

IFN 10.80 6.33 3.28 0.239 0.058 1.74 0.87

IM 9.08 5.71 3.23 0.190 0.091 1.66 0.38

IM 8.72 5.04 2.68 0.248 0.083 1.73 0.54

IM 9.41 5.18∗ 3.35 0.213 0.065 1.78 0.47

(4)

dg, HU 11.60 7.79 3.94 0.666 0.833 1.86 0.94

IFN 13.00 8.44 4.05 0.686 0.303 1.62 1.18

IFN 13.60 10.40 3.34 0.637 0.270 1.80 0.98

IM 10.90 6.51 3.10 0.541 0.231 1.62 0.58

IM 10.90 6.98 3.34 0.613 0.209∗∗ 1.84 0.66

(5)

dg, HU 10.40 6.29 3.04 0.248 0.133 1.59 1.76

HU 10.20 7.40 3.03 0.158 0.066 1.98 1.20

IFN 9.35 6.20 2.97 0.195 0.142 1.61 1.31

IM 9.21 5.01 2.82 0.236 0.097 1.83 1.13

IM 9.81 5.33 3.16 0.226 0.082 1.93 1.12

IM 9.11∗ 6.77 2.62 0.605 0.081 1.95 1.06∗∗

(6)

dg, IFN 8.05 5.63 1.27 0.135 0.073 1.17 1.10

IFN 11.30 n.t.a n.t. n.t. n.t. 1.39 1.29

IM 12.00 n.t. n.t. n.t. n.t. 1.58 1.41

DS 17.90 12.70 1.75 0.611 0.152 2.07 0.79

DS 13.6∗ 11.20 2.17∗∗ 0.416 0.188∗∗ 2.26∗∗ 1.18

(7)

dg, HU 8.92 4.60 4.01 0.388 0.610 1.22 0.95

HU 8.79 4.45 3.95 0.337 0.797 1.30 0.76

HU 8.76 3.81 4.11 0.266 0.702 1.34 0.51

IM 6.08 2.89 2.72 0.262 0.507 0.99 0.32

IM 6.63∗ 31.4∗ 2.90 0.303 0.425 0.93 0.36∗

(8)

dg, IM 12.60 9.51 3.53 0.080 n.t. 1.49 0.97

IM 11.60 8.81 2.73 0.083 0.320 1.75 0.79

IM 12.20 7.43 2.74 0.087 0.276 1.84 0.57

IM 12.60 7.64 3.21 0.087 0.188 1.67 0.63

IM 12.60 9.30 3.26 0.094 0.432 1.94 0.74

IM 11.70 7.41 2.95 0.072 0.273 1.93 0.72


Table 2: Continued.

No. Therapy IgG IgG1 IgG2 IgG3 IgG4 IgA IgM

(9)

dg, IM 10.60 7.58 3.41 0.523 0.697 1.31 1.11

IM 9.43 5.36 3.41 0.480 0.790 1.47 1.11

IM 9.53 5.90 3.95 0.479∗∗ 0.738 1.81∗∗ 1.32

(10)

dg, IM 13.50 8.46 4.93 0.231 2.180 1.04 1.61

IM 14.20 7.63 5.63 0.211 1.970 1.28 1.60

IM 12.10 7.49 4.49 0.262 1.200 1.04 1.39

IM 12.20 6.63∗∗ 4.51 0.232 1.340 1.15 1.41

(11)

dg, HU 12.00 6.21 5.66 0.231 0.146 0.88 0.67

DS 11.50 5.97 5.63 0.271 0.085 0.91 0.45

DS 12.40 8.12 3.50∗∗ 0.677∗∗ 0.080∗∗ 1.02∗∗ 0.32∗∗an.t.: not tested.∗P < .05, ∗∗P < .01P for trend with time.

and once against SM), were detected in the course of theobservation period in a total of eight patients. In most ofthem, their appearance was transitory. In two patients (nos.1 and 6, both of them were females), autoantibodies weredetected against three antigens: in two patients (nos. 4 and7) against two antigens and in four patients (nos. 2, 3,9 and 10) against one antigen only. The most frequentlydetected autoantibodies were reactive with TPO, TG andnucleus (in all instances in three patients). In two patientsantibodies reactive with SM were detected. In 1 of these(patient no. 1), they were present in the pretreatment serumsample, disappeared after starting the therapy with HU andwere not detected later on when HU was gradually replacedby IFNα and IM. A similar phenomenon was seen in patientno. 7, in whom the initial reactivity with TPO and TGdisappeared in the course of the therapy. Thus no consistentpattern was evident and, because of the multiplicity of drugsemployed for individual patients, no clear dependence onthe therapy used was observed. It may be of interest that theantibodies against mitochondria (AMAb) and endomysium(E-AAb and E-GAb) were never detected.

There seems to be some, but not quite a consistentcorrelation with the mode of therapy and the developmentof autoantibodies. Autoantibodies were detected in five ofsix patients treated with IFNα and in four out of fivepatients treated with HU (all of them had later on beentreated with IFNα) but in only two out of four patientsexclusively treated by TKI (either IM or DS). Furthermore,in four patients autoantibodies detected after HU and IFNαtreatment disappeared when these drugs were replaced byIM.

3.3. Complement C3 and C4 Components. In the pretreat-ment sera normal levels of C3 were observed in all but onepatient (no. 11), in whom the level was below the norm.As indicated in Table 4, little variation of C3 levels wasobserved in the course of the observation period; still, inthree patient’s a transitory decrease of its level below thestandard laboratory range was observed. In all instances(including patient no. 11), their levels returned to the norm

by the time hematological remission was achieved. In two ofthree patients with the increased levels of C4 in pre-treatmentsera (nos. 4, and 5), their levels dropped in the course oftreatment.

3.4. C-Reactive Protein and IL-6. The results are alsopresented in Table 4. Prior to the start of the therapy,increased levels of CRP were detected in six patients. Judgingby the levels detected in the last samples collected, thelevels dropped to the norm in all of them. However, it isnoteworthy that in two patients (nos. 3 and 8), in spite of thehematological remission having been achieved, their levelsincreased, and in four other patients (nos. 1, 2, 5 and 7)a transitory increase of CRP was detected in the course ofthe observation period. In patient no. 7 the increased levelof CRP corresponded with an acute infection of the upperrespiratory tract. In the other four patients (nos. 1, 3, 5 and7) no clinical complications at the time of CRP increase wereobserved or reported by the patients. As concerns IL-6 levels,there was a nearly general decrease in association with thehematological remission. The decrease for the whole groupwas highly significant (P = .001). Again, as in the previousstudy [22], no clear correlation between IL-6 and CRP levelswas apparent.

3.5. Lymphocyte Subpopulations. As indicated in Table 5,in spite of the drop of lymphocytes (shown in Table 1),there was little variation in the percentage of CD3+ cellsin the course of the observation period, including patientno. 6, in whom lymphopenia was detected in the pre-treatment sample (see Table 1), and patient no. 10 withCD3+ lymphocyte percentage slightly below the norm. Theonly exception was patient no. 11, initially treated with HUwho after a rather long interval without any treatment wastreated with DS. In this patient a marked drop of CD3+cells was detected. It is clear, however, that this decrease wasrelative, reflecting a substantial increase of NK cells. Thechanges in the percentages of CD4+ and CD8+ were slightor moderate in nearly all patients and no consistent pattern


Table 3: Autoantibodies.

No. Therapy TPOAb TGAb ANAb ANCAb AMAb ENDOAb A ENDOAb G SMAb DESMAb

(1)

dg, HU na n n n n n n pb n

IFN n n n p n n n n n

IFN n n n n n n n n p

IM n n n p n n n n p

IM n n n p n n n n p

IM n n n n n n n n p

DS n n n n n n n n p

(2)

dg, HU n n n n n n n n n

IFN n n n n n n n n n


IFN n n p n n n n n n

IM n n n n n n n n n



(3)



IFN n n n n n n n p n


IM n n n n n n n p n


(4)



IFN n n n n n n n n p

IM p n n n n n n n n


(5)


HU n n n n n n n n n





(6)

dg, IFN p n n n n n n n n

IFN p p n n n n n n n

IM p n p n n n n n n

DS p n n n n n n n n

DS p n n n n n n n n

(7)

dg, HU p p n n n n n n n

HU p p n n n n n n n

HU n.t.c n.t. n n n n n n n

IM p n n n n n n n n


(8)

dg, IM n n n n n n n n n




IM n.t. n.t. n n n n n n n



Table 3: Continued.

No. Therapy TPOAb TGAb ANAb ANCAb AMAb ENDOAb A ENDOAb G SMAb DESMAb

(9)



IM n n n p n n n n n

(10)




IM n n p n n n n n n

(11)


DS n n n n n n n n n

DS n n n n n n n n n

TPOAb: antibodies against thyroidal peroxidase; TGAb: antibodies against thyreoglobulin; ANAb: antinuclear antibodies; ANCAb: antibodies against thecytoplasm of neutrophils; AMAb: antimitochondrial antibodies; ENDOAb A, ENDOAb G: antibodies of the IgA and IgG classes against endomysium; SMAB:antibodies against smooth muscless; DESMAb: antidesmosomal antibodies.aNegative for the respective antibodies.bPositive for the respective antibody.cn.t.:not tested.

was evident. Still, in four of seven patients, in whose pre-treatment samples the percentage of CD8 cells was below thenorm, their percentages did not reach the lower limit of thereferential range. In four patients (nos. 4, 5, 6 and 10) theremission was associated with an increase in CD19+ cells.When the distribution of lymphocyte subpopulations in thefirst and last sample was compared, significant differences(P < .05 to< .001) were encountered in six patients (nos. 1, 3,5, 6, 9 and 11). However, no consistent pattern was apparentand thus the real significance of these findings is doubtful.

3.6. Intracellular Production of Cytokines. The results arepresented in Table 6. Two observations may be of interest.Possibly the most important one was an increase of the CD3+cells, which produce the cytokines, in most of the patients,including those, in whom the production of these cytokinesprior to the start of the therapy had been pathologically low(e.g., in the case of IL-2, patients nos. 3 and 8). Of the tenpatients, which could be evaluated, the percentage of IL-2-producing cells increased in seven, IL-4-producing cells alsoin seven, TNFα-producing cells in six and INFγ producingcells in seven. In some of the patients the level of cytokineproduction remained essentially unchanged (e.g., patient no.8, all four cytokines) and in one patient there was a dropof IL-2 and IL-4 production (patient no. 1) and in anotherpatient the drop of INFγ production (patient no. 2).

The other noteworthy observation, closely associatedwith the first one, was an increase in cells producing morethan one cytokine after stimulation, as reflected by anincrease of the percentage sum of reactive cells. The increasewas expressed in terms of the Cytokine Production Index(CPI) given by the ratio between the sum of cytokine-producing cells as detected in the last sample and the sumof cytokine producing cells in the first sample available (i.e.,in all but one patient before the start of therapy). A marked

increase in CPI was detected in seven (nos. 3, 4, 5, 6, 7, 9and 10) out of 10 patients which could be evaluated. Thedifference for the group was highly significant (P = .006).

3.7. Antibodies Against Herpesviruses and Papillomaviruses.To clarify whether treatment of the CML patients with HU,IFNα or TKI was associated with activation of latent and/orpersistent virus infections, we tested sera from the patientsfor presence of antibodies against four human herpesvirusesand six human papillomaviruses. The results are summarizedin Tables 7 and 8. For the sake of simplicity, only the resultsof testing sera taken prior to the start of any treatment andat the end of the observation period are presented. It isevident that there was only very little difference between thetwo sets of sera. In one patient (no. 11) treated with DS,cytomegalovirus infection was reactivated as revealed by avery marked increase in IgG antibody and the appearance ofIgM antibody (results not shown).

4. Discussion

In the preceding paper [22], we showed that CML patientsbefore treatment differed from matched healthy subjects ina number of immunological parameters. The major aimof the present investigation was to find out, whether theseaberrations persisted, decreased, or disappeared in the courseof treatment, in particular whether these changes correlatedwith the induction of hematological and/or cytogeneticremission, whether and how they were influenced by thedrugs used for treatment, and, in the long run, what was theirprognostic value, if any. Our efforts were seriously hamperedby two circumstances, that is, by the loss of 13 out of 24patients originally enrolled and, for ethical reasons, by theimpossibility to maintain the original treatment regimen


Table 4: Complement components, CRP and IL-6.

No. Therapy C3 C4 CRP IL-6

(1)

dg, HU 0.94 0.40 5.2 5.10

IFN 0.67 0.37 3.0 2.53

IFN 0.83 0.42 24.4 5.42

IM 0.71 0.30 3.5 1.83

IM 0.53 0.21 3.5 1.53

IM 0.72 0.30 3.3 1.67

DS 1.26 0.52 3.2 2.72

(2)

dg, HU 1.24 0.30 12.8 9.54

IFN 1.22 0.28 3.0 3.47

IFN 1.22 0.28 3.5 3.87

IFN 1.15 0.25 3.5 4.92

IM 1.19 0.26 3.2 3.21

IM 1.27 0.27 5.4 2.45

IM 1.20 0.24∗ 3.2 2.82∗

(3)

dg, HU 1.19 0.26 3.0 3.9

IFN 1.14 0.26 3.0 1.8

IFN 0.94 0.24 3.5 1.5

IM 1.15 0.27 3.3 2.16

IM 1.20 0.26 3.2 1.14

IM 1.40 0.30 32.9 3.03

(4)

dg, HU 1.35 0.38 11.6 4.40

IFN 0.90 0.21 3.0 1.54

IFN 1.11 0.23 3.5 1.97

IM 1.20 0.27 3.2 1.45

IM 1.06 0.25 3.2 1.50

(5)

dg, HU 1.03 1.23 3.0 3.30

HU 0.97 0.21 3.5 3.67

IFN 0.80 0.20 3.5 2.16

IM 0.91 0.19 3.2 2.49

IM 1.14 0.23 8.8 2.30

IM 1.00 0.25 3.4 1.99

(6)

dg, IFN 0.98 0.21 3.0 2.80

IFN 0.69 0.19 n.t.a 1.82

IM 0.78 0.17 3.5 1.87

DS 1.31 0.30 3.2 2.31

DS 1.14 0.24 3.2 0.00

(7)

dg, HU 0.91 0.16 3.5 5.08

HU 0.68 0.15 3.5 4.24

HU 0.91 0.18 3.3 5.29

IM 0.91 0.19 21.3 1.82

IM 0.98 0.17 3.2 3.00

(8)

dg, IM 0.93 0.19 3.5 3.80

IM 0.94 0.19 3.8 2.63

IM 0.89 0.17 3.2 2.31

IM 0.89 0.18 3.3 1.53

IM 1.07 0.26 9.4 0.27

IM 1.11 0.33 6.1 1.70∗


Table 4: Continued.

No. Therapy C3 C4 CRP IL-6

(9)

dg, IM 0.86 0.29 23.0 14.20

IM 0.91 0.24 3.2 1.21

IM 1.05∗∗ 0.27 3.2 1.00∗

(10)

dg, IM 0.79 0.30 5.9 16.71

IM 0.86 0.24 3.3 4.97

IM 0.81 0.17 3.2 n.t.

IM 0.76 0.17 3.2 1.00∗∗

(11)

dg, HU 0.65 0.36 6.8 5.20

DS 0.59 0.25 3.2 2.23

DS 0.80 0.35 3.4 3.50an.t.:not tested.∗P < .05, ∗∗P < .01P for trend with time.

with either HU or INFα. Because of the changes in the treat-ment modalities, the original set of patients split into twogroups. The first one consisted of eight patients originallytreated with HU and/or INFα and subsequently with TKI,while the remaining three patients treated exclusively withTKI constituted the second group.

Of these two shortcomings, the diminution of our setof patients from 24 to 11 was certainly the more impor-tant one. This reduction increased the difficulties alreadyinherent in the group studied, that is, its inhomogeneity(age, sex, different treatments), as regards evaluation of thedata obtained. Still, because in all patients hematologicalremission and in nearly all of them even CCR was achieved,this bringing an element of homogeneity into the studygroup, some conclusions can be drawn. The most importantone is the nearly general association of the remission witha normalization of the aberrant immunity parameters, bothhumoral a cellular. One can assume that this normalizationwas due to an alleviation of the CML-associated processes.The continuing followup of the patients should revealwhether any relapse of the disease would be associated withdeviations from the norm and with the reappearance of thesame or appearance of some other aberrations.

As regards the levels of immunoglobulins, there was atendency towards their decrease, predominantly IgM. Thegradual decrease of IgM levels was highly significant. Thesefindings correspond with the results reported by Steegmannet al. [24], who described considerable reduction of levels ofserum immunoglobulins, including IgM, in patients previ-ously exposed to IFNα and then treated with IM; accordingto their findings, the reduction of immunoglobulins wasespecially marked in patients expressing a pronounced cyto-genetic response. A marked selective effect of IM treatmenton serum IgM level in one patient has been reported byNagasawa and Mizutan [25]. We are unable to provide anyreasonable explanation for our observation. We can onlyspeculate that the decrease observed was due to qualitativealteration of B cells by the therapy employed. Otherwise,no consistent pattern was apparent. We only rarely detected

decrease in IgA levels, which we found significantly increasedin the pretreatment CML patients when compared with theirmatched healthy controls [22].

When monitoring the presence of autoantibodies, wedetected them in five out of six patients treated withINFα. These antibodies were directed against cytoplasm ofneutrophils (ANCAb), smooth muscles (ASMAb), nucleus(ANAb), thyreoglobulin (ATG), or thyroidal peroxidase(ATPO). This is not particularly surprising, since INFαexhibits allo- and autostimulation activity on antigen pre-senting cells. However, its activity in vivo may dependon the simultaneous regulation of network of immunecells rather than on the activation of individual popula-tions [10]. INFα, but not IM, was reported to cause anincreased transcription of proteinase 3 in CD14-positivemonocytes, this suggesting another possible mechanism,by which IFNα may promote self-antigen presentation [7].In our study the appearance of autoantibodies was in mostinstances of transitory character and that their disappear-ance quite frequently followed the replacement of INFαby IM. However, autoantibodies also developed in twoout of five patients untreated with INFα. It is thereforedifficult to claim, on the basis of the present results, thatINFα was markedly more active than TKI in inducingautoantibodies. It should be added that, in our patients, thedevelopment of antibodies did not manifest itself in anyclinically recognizable disease, with the possible exceptionof the simultaneous presence of ANCAb and vasculitis inpatient no. 1. This possible association and the subsequentcure of the vasculitis in parallel with the disappearance ofthe antibody after IM treatment seems to be in line witha recent report [26].

As concerns the C3 and C4 components of complementthere was little variation during the observation period.Again, there was a tendency to normalization. This seemsto be in agreement with other findings in patients withhematological malignancies including CML [27]; however,the changes we observed were rather small and their signif-icance is questionable. It seems clear that the follow-up of


Table 5: Lymphocytes subpopulations.

No. Therapy CD3a CD4 CD8 CD19 CD3-CD16,56+

(1)

dg, HU 72 54 18 14 14

IFN 74 53 21 4 13

IFN 84 52 32 10 6

IM 76 48 28 11 12

IM 76 49 24 10 12

IM 73 47 26 13 10

DS 79 51 27 5 14

(2)

dg, HU 86 61 25 5 7

IFN 84 52 32 3 13

IFN 75 50 25 4 21

IFN 82 70 12 8 9

IM 69 54 15 11 20

IM 78 61 15 7 14

IM 80 62 15 7 13

(3)

dg, HU 76 39 37 12 12

IFN 79 42 36 8 13

IFN 73 44 29 13 13

IM 69 42 27 17 13

IM 72 48 24 15 13

IM 63 44 19 12 25

(4)

dg, HU 74 60 14 6 19

IFN 86 59 27 6 8

IFN 68 39 28 10 21

IM 75 52 23 13 12

IM 78 53 25 12 10

(5)

dg, HU 62 42 18 6 32

HU 66 43 23 20 14

IFN 62 40 22 24 14

IM 65 42 23 15 20

IM 66 41 20 13 20

IM 67 43 24 15 17

(6)

dg, IFN 83 40 43 10 6

IFN 77 53 23 11 11

IM 76 52 23 11 12

DS 76 31 43 11 12

DS 65 29 34 23 12

(7)

dg, HU 75 51 24 10 15

HU 81 54 27 7 12

HU 74 48 21 11 14

IM 73 52 20 12 14

IM 73 53 19 10 16

(8)

dg, IM 76 54 22 9 14

IM 72 55 17 13 16

IM 76 54 22 8 15

IM 77 55 19 10 11

IM 74 54 17 9 17

IM 74 57 17 12 14


Table 5: Continued.

No. Therapy CD3a CD4 CD8 CD19 CD3-CD16,56+

(9)

dg, IM 85 29 53 3 12

IM 76 34 38 7 15

IM 74 37 34 9 15

(10)

dg, IM 67 52 14 14 16

IM 61 40 21 20 18

IM 55 37 16 18 25

IM 54 37 14 23 23

(11)

dg,HU 67 57 9 20 12

DS 59 49 9 16 23

DS 39 32 7 10 50aThe figures indicate the percentages of cells positive for the respective surface CDs.

C3 and C4 components of complement was not particularlyrewarding in this study.

The results obtained when testing the CRP and IL-6 levels are of greater interest. Both tended to decreasefollowing the therapy. The decrease was quite common inthe case of IL-6, where it was highly significant. Thesedata suggest that a reduction of IL-6 levels might be quitea reliable marker of successful therapy. Moreover, becauseof its biological effects, that is, inhibition of p53-inducedapoptosis [28] and suppression of phosphorylation of theretinoblastoma protein [29], its drop might contributeto the favorable course of the disease. Another favorableeffect of IL-6 decrease might be associated with its rolein STAT-3 activation. High levels of STAT-3 can preventdendritic cell maturation and subsequent presentation of theantigens [30]. Thus, increased levels of IL-6 might exhibitan immunosuppressive effect. In a way, the present datacorrespond with the earlier observations indicating that IL-6levels are raised in parallel with the progression of the diseaseinto blastic phase [31, 32]. An association of remission witha decrease of CRP levels was less consistent. Although CRPlevels dropped in all patients with increased levels detectedin their sera taken prior to the start of the therapy, in fourother patients a CRP increase was observed in the course ofthe observation period; in two of them it was only transitory.Because of the nature of this acute phase reactant [33–35],it is rather difficult to interpret the latter findings. In one ofthe patient the transitional rise of CRP was associated witha respiratory disease. In the other patients some undetectedmicroinflammation processes might have been involved.

Possibly the most important among our findings are thechanges of cell-mediated immunity parameters associatedwith the achievement of remission. They were representedby a markedly increased capability of stimulated CD3+ cellsto produce cytokines. Since at the time of remission, thesum of percentages of CD3+ cells producing any of thecytokines tested exceeded 100 in most of the patients, it ispossible to conclude that the changes observed were mainlydue to an increase in cells producing more than one cytokine.Thus, a restoration of the Tcell activities associated with the

suppression of the disease was observed. Similar results werereported by Reuben et al. [36]. and Guarini et al. [37]. inpatients, in whom hematologic remission was achieved byINFα treatment. The mechanisms responsible for the presentfindings are not quite clear. In this respect, two previous invitro studies may be of interest and provide a lead for furtherstudies. Pawelec et al. [38]. have reported the production ofthe immunosuppressive IL-10 cytokine, a potent inhibitor oftype 1 cytokines, by CML cells cultivated in vitro and havedemonstrated that its neutralization by monoclonal antibodyconsiderably enhanced the proliferation of lymphocytes inmixed lymphocyte/tumor cell cultures [38]. In anotherstudy, Kiani et al. [39]. Showed that CD4+ cells from CMLpatients, which had been separated from the leukemic cells,after stimulation produced type 1 cytokines in amountscomparable to those seen in normal subjects. Taken together,these two sets of data suggest that the products of CMLcells may themselves be responsible for the reduced immuno-competence of the CD3+ cells. One point deserves a specialcomment. All ten patients, in whom the respective data wereavailable, had been treated, at least for some time, withIM. This drug has been reported to suppress CD3+ cellactivation in vitro [40, 41] as well as other parameters of Tcellimmunity in vivo [42–44]. Our findings in the present studydo not seem to be in agreement with those observations. Onecan therefore hypothesize that the alleviation of the CML-associated processes, which was induced by the successfultherapy, was a more important factor than the puta-tive immunosuppressive effect of IM. Furthermore, it hasrecently been reported that IM is suppressing the activationand proliferation of CD4+CD25+ regulatory cells (Treg cells)which are producing strong immunosuppressive factors likeIL-10, transforming growth factor β (TGFβ) and granzymeB [45]. It is thus very well possible that in the patientsstudied the IM-induced suppression of Treg contributed tothe present observation. Unfortunately, we did not monitorthe levels of Treg cells in the present study. It should alsobe recalled that a significant increase of INFγ-producingTcells following IM treatment has been reported byAswald et al. [46].


Table 6: Intracellular cytokines production in activated CD3+ lymphocytes.

No. Therapy IL-2 TNFalfa INF-γ IL-4 CPI

(1)

dg, HU 56 44 12 4

IFN 42 49 18 3

IFN 31 35 23 5

IM 53 60 30 2

IM 61 48 36 3

IM n.t.a n.t. n.t. n.t.

DS 32 60 23∗ 1 1.000

(2)

dg, HU 31 53 62 2

IFN 10 23 29 5

IFN 69 66 30 4

IFN 76 65 32 5

IM 67 73 42 6

IM n.t. n.t. n.t. n.t.

IM 55 45 35 20∗ 1.047

(3)

dg, HU 1 54 41 6

IFN 24 46 36 3

IFN 48 66 50 3

IM 23 51 25 1


IM 49 57 51 20 1.735

(4)

dg, HU 17 20 25 6

IFN 31 46 29 8

IFN 38 78 38 6


IM 56∗∗ 65 44∗∗ 11 2.588

(5)

dg, HU 22 21 14 4

HU 50 55 25 4

IFN 39 50 30 2

IM 37 49 24 8

IM 52 58 30 2

IM 38 41 30 5 1.869

(6)

dg, IFN 11 17 22 2

IFN 36 51 32 3

IM 48 55 35 5

DS 20 59 45 3

DS 18 44 37∗ 19 2.269

(7)

dg, HU 51 46 34 3

HU 52 42 30 2

HU 52 68 55 5


IM 52 53 34 25 1.223

(8)

dg, IM 42 44 23 2

IM 38 44 19 2

IM 16 25 13 5



IM 41 37 23 3 0.937

(9)dg, IM 2 34 45 1


IM 31 74 57 13 2.134


Table 6: Continued.

No. Therapy IL-2 TNFalfa INF-γ IL-4 CPI

(10)

dg, IM n.t. n.t. n.t. n.t.

IM 27 32 14 1

IM 38 54 24 3

IM 50∗∗ 51 30∗∗ 6∗∗ 1.827

(11)dg, HU n.t. n.t. n.t. n.t.

DS 59 62 23 1

DS n.t. n.t. n.t. n.t. n.t.

CPI:Cytokine Production Index—a ratio between the sum of cytokine producing cells ls as detected in the last sample and the sum of cytokine producingcells in the first sample available.an.t.:not tested.∗P < .05, ∗∗P < .01. P for trend with time.

Table 7: Presence and geometric mean titres of antibodies against human herpesviruses in CML patients prior to the start of the therapyand after achieiving hematological and/or complete cytogenetic remission.

Antibody presence

Sample No. HSV1/2 CMV VZV EBV

IgG IgM IgG IgM IgG IgM VCAIgG VCAIgM EAIgG EBNA IgM

First 11 11 2 8 0 11 2 11 2 3 10

Last 11 11 0 8 1 11 2 11 2 3 10

GMT

First 11 8.12 0.65 1.83 0.47 2.83 0.74 7.41 0.33 0.57 4.28

Last 11 8.33 0.57 2.01 0.47 2.61 0.54 8.20 0.34 0.65 4.79

Table 8: Presence and geometric mean titres of antibodies against human papillomaviruses in CML patients prior to the start of the therapyand after achieving hematological and/or complete cytogenetic remission.

Antibody presence

Sample No. HPV6 HPV11 HPV16 HPV18 HPV31 HPV33

First 11 6 5 3 3 3 1

Last 11 5 3 3 3 2 1

GMT

First 11 1.27 1.01 0.68 1.00 0.79 0.50

Last 11 1.08 0.88 0.65 0.75 0.61 0.38

Our attempts to find a reflection of these changes in theantibody titers against herpesviruses and papillomaviruses,two virus families known to be activated under immunosup-pression, failed completely. As indicated in a previous paper[22], there were no significant differences in the prevalence ofantibodies against these viruses between the untreated CMLpatients and matched normal control subjects. In the presentstudy the antibody patterns in sera taken before the startof therapy and after achieving remission were comparable,this strongly suggesting that in the course of the observationperiod the activation of latent infections did not occur orcould not be revealed by the antibody tests used.

In spite of inconsistency of the treatment regimensdescribed above, some differences in the changes of theimmunological parameters studied, which might have beenassociated with the drugs used, were recorded. Two obser-vations are noteworthy. First, the autoantibodies were more

frequently seen in patients who had been treated with HUand INFα than in those treated exclusively with TKI, and theytended to disappear after substituting TKI for HU and INFα.Still, the lack of consistency and the small number of patientstested preclude any conclusions. Second, increase of NK cellswas detected more frequently in those exclusively treatedwith TKI than in patients who had undergone combinedtherapy.

5. Conclusions

The present results indicate that the altered parameters ofinnate immunity found in the CML patients prior to thestart of any therapy tended to normalize in the course oftherapy leading to remission. Other data suggested, but didnot prove, that some of the parameters monitored might beinfluenced by the treatment modality used.


Acknowledgments

Supported by grants from the Internal Granting Agencyof the Ministry of Health, Czech Republic (IGA MZCRNC/6957-3, and NR/9075-3) and by Research Project No.000 273 3601, Institute of Hematology and Blood Transfu-sion. The authors express their indebtedness to H. Tlaskalovafor critical reading of the manuscript.

References

[1] M. Suehiro, S. Kishimoto, T. Wakabayashi et al., “Hydroxyureadermopathy with a dermatomyositis-like eruption and a largeleg ulcer,” British Journal of Dermatology, vol. 139, no. 4, pp.748–749, 1998.

[2] M. J. Dacey and J. P. Callen, “Hydroxyurea-induced dermat-omyositis-like eruption,” Journal of the American Academy ofDermatology, vol. 48, no. 3, pp. 439–441, 2003.

[3] M. Talpaz, K. B. McCredie, G. M. Mavligit, and J. U. Gutter-man, “Leukocyte interferon-induced myeloid cytoreductionin chronic myelogenous leukemia,” Blood, vol. 62, no. 3, pp.689–692, 1983.

[4] M. Talpaz, K. McCredie, H. Kantarjian, J. Trujillo, M. Keating,and J. Gutterman, “Chronic myelogenous leukaemia: haema-tological remissions with alpha interferon,” British Journal ofHaematology, vol. 64, no. 1, pp. 87–95, 1986.

[5] F. Bonifazi, A. De Vivo, G. Rosti et al., “Chronic myeloidleukemia and interferon-α: a study of complete cytogeneticresponders,” Blood, vol. 98, no. 10, pp. 3074–3081, 2001.

[6] J. J. Molldrem, P. P. Lee, C. Wang et al., “Evidence that specificT lymphocytes may participate in the elimination of chronicmyelogenous leukemia,” Nature Medicine, vol. 6, no. 9, pp.1018–1023, 2000.

[7] A. Burchert, S. Wolfl, M. Schmidt et al., “Interferon-α, butnot the ABL-kinase inhibitor imatinib (STI571), inducesexpression of myeloblastin and a specific T-cell response inchronic myeloid leukemia,” Blood, vol. 101, no. 1, pp. 259–264,2003.

[8] C. J. Wu, X.-F. Yang, S. McLaughlin et al., “Detection of apotent humoral response associated with immune-inducedremission of chronic myelogenous leukemia,” Journal ofClinical Investigation, vol. 106, no. 5, pp. 705–714, 2000.

[9] D. Chakrabarti, B. Hultgren, and T. A. Stewart, “IFN-α inducesautoimmune T cells through the induction of intracellularadhesion molecule-1 and B7.2,” Journal of Immunology, vol.157, no. 2, pp. 522–528, 1996.

[10] D. Tosi, R. Valenti, A. Cova et al., “Role of cross-talk betweenIFN-α-induced monocyte-derived dendritic cells and NK cellsin priming CD8+ T cell responses against human tumorantigens,” Journal of Immunology, vol. 172, no. 9, pp. 5363–5370, 2004.

[11] S. Sacchi, H. Kantarjian, S. O’Brien, P. R. Cohen, S. Pierce,and M. Talpaz, “Immune-mediated and unusual complica-tions during interferon alfa therapy in chronic myelogenousleukemia,” Journal of Clinical Oncology, vol. 13, no. 9, pp.2401–2407, 1995.

[12] E. Tothova, A. Kafkova, N. Stecova, M. Fricova, T. Guman,and E. Svorcova, “Immune-mediated complications duringinterferon alpha therapy in chronic myelogenous leukemia,”Neoplasma, vol. 49, no. 2, pp. 91–94, 2002.

[13] M. W. N. Deininger, J. M. Goldman, and J. V. Melo, “Themolecular biology of chronic myeloid leukemia,” Blood, vol.96, no. 10, pp. 3343–3356, 2000.

[14] H. Kantarjian, C. Sawyers, A. Hochhaus et al., “Hematologicand cytogenetic responses to imatinib mesylate in chronicmyelogenous leukemia,” New England Journal of Medicine, vol.346, no. 9, pp. 645–652, 2002.

[15] B. J. Druker, M. Talpaz, D. J. Resta et al., “Efficacy and safety ofa specific inhibitor of the BCR-ABL tyrosine kinase in chronicmyeloid leukemia,” New England Journal of Medicine, vol. 344,no. 14, pp. 1031–1037, 2001.

[16] V. Nardi, M. Azam, and G. Q. Daley, “Mechanisms andimplications of imatinib resistance mutations in BCR-ABL,”Current Opinion in Hematology, vol. 11, no. 1, pp. 35–43, 2004.

[17] M. E. Gorre, M. Mohammed, K. Ellwood et al., “Clinicalresistance to STI-571 cancer therapy caused by BCR-ABL genemutation or amplification,” Science, vol. 293, no. 5531, pp.876–880, 2001.

[18] N. P. Shah, C. Tran, F. Y. Lee, P. Chen, D. Norris, and C. L.Sawyers, “Overriding imatinib resistance with a novel ABLkinase inhibitor,” Science, vol. 305, no. 5682, pp. 399–401,2004.

[19] E. Weisberg, P. Manley, J. Mestan, S. Cowan-Jacob, A. Ray,and J. D. Griffin, “AMN107 (nilotinib): a novel and selectiveinhibitor of BCR-ABL,” British Journal of Cancer, vol. 94, no.12, pp. 1765–1769, 2006.

[20] P. S. Kim, P. P. Lee, and D. Levy, “Dynamics and potentialimpact of the immune response to chronic myelogenousleukemia,” PLoS Computational Biology, vol. 4, no. 6, articleno. e1000095, 2008.

[21] V. Vonka, “Immunotherapy of chronic myeloid leukemia:present state and future prospects,” Immunotherapy, vol. 2, no.2, pp. 227–241, 2010.

[22] Z. Humlova, H. Klamova, I. Janatkova et al., “Immunologicprofiles of patients with chronic myeloid leukemia. I. Statebefore the start of treatment,” Folia Biologica (Praha), vol. 52,no. 3, pp. 47–58, 2006.

[23] E. Hamsıkova, V. Ludvıkova, M. Smahel, M. Sapp, and V.Vonka, “Prevalence of antibodies to human Papillomavirusesin the general population of the Czech Republic,” InternationalJournal of Cancer, vol. 77, no. 5, pp. 689–694, 1998.

[24] J. L. Steegmann, G. Moreno, C. Alaez et al., “Chronicmyeloid leukemia patients resistant to or intolerant ofinterferon α and subsequently treated with imatinib showreduced immunoglobulin levels and hypogammaglobuline-mia,” Haematologica, vol. 88, no. 7, pp. 762–768, 2003.

[25] M. Nagasawa and S. Mizutani, “Selective effect of imatinib onserum IgM in a patient with CML,” International Journal ofHematology, vol. 80, no. 4, pp. 381–382, 2004.

[26] A.-I. Kalsch, M. Soboletzki, W. H. Schmitt et al., “Imatinibmesylate, a new kid on the block for the treatment of anti-neutrophil cytoplasmic autoantibodies-associated vasculitis?”Clinical and Experimental Immunology, vol. 151, no. 3, pp.391–398, 2008.

[27] D. Q. Minh, E. Czink, A. Mod, G. Fust, and S. R. Hol-lan, “Serial complement measurements in patients withleukaemia,” Clinical and Laboratory Haematology, vol. 5, no.1, pp. 23–34, 1983.

[28] E. Yonish-Rouach, D. Resnitzky, J. Lotem, L. Sachs, A. Kimchi,and M. Oren, “Wild-type p53 induces apoptosis of myeloidleukaemic cells that is inhibited by interleukin-6,” Nature, vol.352, no. 6333, pp. 345–347, 1991.

[29] D. Resnitzky, N. Tiefenbrun, H. Berissi, and A. Kimchi,“Interferons and interleukin 6 suppress phosphorylation ofthe retinoblastoma protein in growth-sensitive hematopoietic


cells,” Proceedings of the National Academy of Sciences of theUnited States of America, vol. 89, no. 1, pp. 402–406, 1992.

[30] S.-J. Park, T. Nakagawa, H. Kitamura et al., “IL-6 regulates invivo dendritic cell differentiation through STAT3 activation,”Journal of Immunology, vol. 173, no. 6, pp. 3844–3854, 2004.

[31] M. Anand, S. K. Chodda, P. M. Parikh, and J. S. Nadkarni,“Abnormal levels of proinflammatory cytokines in serumand monocyte cultures from patients with chronic myeloidleukaemia in different stages, and their role in prognosis,”Haematological Oncology, vol. 16, no. 4, pp. 143–154, 1998.

[32] K. E. Panteli, E. C. Hatzimichael, P. K. Bouranta et al., “Seruminterleukin (IL)-1, IL-2, sIL-2Ra, IL-6 and thrombopoietinlevels in patients with chronic myeloproliferative diseases,”British Journal of Haematology, vol. 130, no. 5, pp. 709–715,2005.

[33] T. W. Du Clos and C. Mold, “C-reactive protein. An activatorof innate immunity and a modulator of adaptive immunity,”Immunologic Research, vol. 30, no. 3, pp. 261–277, 2004.

[34] A. J. Szalai, “C-reactive protein (CRP) and autoimmunedisease: facts and conjectures,” Clinical and DevelopmentalImmunology, vol. 11, no. 3-4, pp. 221–226, 2004.

[35] R. Kleemann, P. P. Gervois, L. Verschuren, B. Staels, H. M.G. Princen, and T. Kooistra, “Fibrates down-regulate IL-1-stimulated C-reactive protein gene expression in hepatocytesby reducing nuclear p50-NFκB-C/EBP-β complex formation,”Blood, vol. 101, no. 2, pp. 545–551, 2003.

[36] J. M. Reuben, B. N. Lee, H. Johnson, H. Fritsche, H. M.Kantarjian, and M. Talpaz, “Restoration of Th1 cytokinesynthesis by T cells of patients with chronic myelogenousleukaemia in cytogenetic and haematologic remission withinterferon-alpha,,” Clinical and Cancer Research, vol. 6, no. 5,pp. 1671–1677, 2000.

[37] A. Guarini, M. Breccia, E. Montefusco et al., “Phenotypic andfunctional characterization of the host immune compartmentof chronic myeloid leukaemia patients in complete haemato-logical remission,” British Journal of Haematology, vol. 113, no.1, pp. 136–142, 2001.

[38] G. Pawelec, A. Rehbein, E. Schlotz, and P. Da Silva, “Cel-lular immune responses to autologous chronic myelogenousleukaemia cells in vitro,” Cancer Immunology Immunotherapy,vol. 42, no. 3, pp. 193–199, 1996.

[39] A. Kiani, I. Habermann, K. Schakel, A. Neubauer, L. Rogge,and G. Ehninger, “Normal intrinsic Th1/Th2 balance inpatients with chronic phase chronic myeloid leukemia nottreated with interferon-α or imatinib,” Haematologica, vol. 88,no. 7, pp. 754–761, 2003.

[40] R. Seggewiss, K. Lore, E. Greiner et al., “Imatinib inhibits T-cell receptor-mediated T-cell proliferation and activation ina dose-dependent manner,” Blood, vol. 105, no. 6, pp. 2473–2479, 2005.

[41] A. B. Dietz, L. Souan, G. J. Knutson, P. A. Bulur, M. R.Litzow, and S. Vuk-Pavlovic, “Imatinib mesylate inhibits T-cell proliferation in vitro and delayed-type hypersensitivity invivo,” Blood, vol. 104, no. 4, pp. 1094–1099, 2004.

[42] J. J. Gu, N. Zhang, Z. W. He, A. J. Koleske, and A. M.Pendergast, “Defective T cell development and function in theabsence of Abelson kinases,” Journal of Immunology, vol. 179,no. 11, pp. 7334–7343, 2007.

[43] P. Sinai, R. E. Berg, J. M. Haynie, M. J. Egorin, R. L. IlariaJr., and J. Forman, “Imatinib mesylate inhibits antigen-specificmemory CD8 T cell responses in vivo,” Journal of Immunology,vol. 178, no. 4, pp. 2028–2037, 2007.

[44] S. Mumprecht, M. Matter, V. Pavelic, and A. F. Ochsenbein,“Imatinib mesylate selectively impairs expansion of memorycytotoxic T cells without affecting the control of primary viralinfections,” Blood, vol. 108, no. 10, pp. 3406–3413, 2006.

[45] J. Chen, A. Schmitt, K. Giannopoulos et al., “Imatinib impairsthe proliferation and function of CD4+CD25+ regulatory Tcells in a dose-dependent manner,” International Journal ofOncology, vol. 31, no. 5, pp. 1133–1139, 2007.

[46] J. M. Aswald, J. H. Lipton, S. Aswald, and H. A. Messner,“Increased IFN-γ synthesis by T cells from patients onimatinib therapy for chronic myeloid leukemia,” Cytokines,Cellular and Molecular Therapy, vol. 7, no. 4, pp. 143–149,2002.


Review Article

DNA Vaccination: Using the Patient’s Immune System toOvercome Cancer

Georg Eschenburg,1 Alexander Stermann,1 Robert Preissner,2 Hellmuth-Alexander Meyer,2, 3

and Holger N. Lode4

1 Experimental Oncology Group, Department of Pediatrics, Charite-University Medicine Berlin, 13353 Berlin, Germany2 Structural Bioinformatics Group, Institute for Physiology, Charite-University Medicine Berlin, 14195 Berlin, Germany3 Department of Urology, Charite-University Medicine Berlin, 10117 Berlin, Germany4 Department of Pediatric Hematology and Oncology, University of Greifswald, 17475 Greifswald, Germany

Correspondence should be addressed to Georg Eschenburg, [email protected]

Received 30 June 2010; Revised 8 October 2010; Accepted 21 October 2010

Academic Editor: Nima Rezaei

Copyright © 2010 Georg Eschenburg et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Cancer is one of the most challenging diseases of today. Optimization of standard treatment protocols consisting of the maincolumns of chemo- and radiotherapy followed or preceded by surgical intervention is often limited by toxic side effectsand induction of concomitant malignancies and/or development of resistant mechanisms. This requires the development oftherapeutic strategies which are as effective as standard therapies but permit the patients a life without severe negative side effects.Along this line, the development of immunotherapy in general and the innovative concept of DNA vaccination in particular mayprovide a venue to achieve this goal. Using the patient’s own immune system by activation of humoral and cellular immuneresponses to target the cancer cells has shown first promising results in clinical trials and may allow reduced toxicity standardtherapy regimen in the future. The main challenge of this concept is to transfer the plethora of convincing preclinical and earlyclinical results to an effective treatment of patients.

1. Introduction

1.1. Immunotherapy and Cancer. Cancer is a leading cause ofdeath and is responsible for a magnitude of all disease-relateddeaths worldwide [1, 2]. Standard cancer therapy includesintensive chemo- and/or radiotherapy able to effectivelyeradicate cancer cells but with the disadvantage of severeside effects. Additionally, many cancers are diagnosed atan advanced tumor stage, where standard therapy has itslimitations and is only able to cure low numbers of patients.Vaccination against cancer is a promising approach to inducethe immune system to specifically target the tumor cells.

However, the successful use of cancer vaccines isdependent on several problems that need to be overcome.Advanced tumor progression often leads to immune sup-pression; patients are weakened by previous therapies andaging [3]. In mouse tumor models, there are indicationsthat young mice are better protected against a lethal tumor

challenge showing improved primary immune responsesthan older mice making the use of vaccines in patientsat an advanced age challenging because the thymus stopsproducing naıve T cells with age [4, 5].

The fact that the status of the patient’s immune systemis critical for the ability to develop an effective antitu-mor immune response is supported by the correlationbetween the amount of tumor-infiltrating lymphocytes witha favourable prognosis [6, 7]. In contrast, patients who arechronically immune suppressed as a result of therapy orother reasons have an unfavourable prognosis [8, 9].

The challenge is to amplify the patients’ own immuneresponse and translate it into a long-lasting memory withoutinduction of unmanageable autoimmunity in order toprotect against metastasis in the future.

One approach is the use of whole inactivated tumor cellsas a source of antigen based on promising results in mousemodels [10]. The biological background to this strategy is


the presence of tumor-specific or tumor-associated antigensexpressed by the cells used for vaccination and the malignanttarget cells. Many antigens involved in effective cell-basedvaccination strategies were identified, characterized, andfound to be also expressed on normal cells. The expression isoften compartmentalized in distinct tissues and is frequentlyof a significantly lower magnitude than on the cancer cells;however, this fact always bears a risk of autoimmunity.Therefore, the ideal cancer vaccine targets tumor-specificantigens expressed exclusively on tumor cells or tumor-associated antigens without harming normal cells expressingthe same antigen, a problem which is hard to solve. Animportant key to success of cancer vaccines is to break self-tolerance, since tumor-associated antigens are self-antigensoverexpressed by tumor cells. This challenge involves the useof distinct prime-boost strategies with different formulationsof the tumor-associated antigen used for vaccination. Thisincludes peptide- or protein-based antigens or delivery withviral vectors, which need to be used in combination in orderto elicit measurable immune responses [11, 12]. Promisingresults were obtained with the use of self-replicating RNAand DNA vaccines which were able to break tolerance againsttumor-associated self-antigens involving pathways of innateantiviral immunity. These vaccines enhance the immuno-genicity and production of antigen-specific antibodies andCD8+ T cells without negative side effects. Although theproduction of antigen was not increased in comparison withconventional DNA vaccines, it is likely that the efficacy ofthe self-replicating vaccines was associated with caspase-dependent apoptotic cell death of transfected cells and asubsequent uptake of these cells by dendritic cells (DCs)[13–15].

Application of xenogeneic tumor-associated antigensis another interesting strategy to overcome some of theobstacles mentioned above. The magnitude of an antitumorresponse is clearly improved by using a tumor-associatedantigen obtained from a different species sharing criticalepitopes flanked by xenogeneic protein sequences furtherstimulating the antitumor response [16–19]. The deliveryof such xenogeneic tumor-associated antigens by DNAvaccination may be a promising venue to the design of asuccessful strategy.

1.2. DNA Vaccination. Historical observations that transferof foreign DNA by different in vitro and in vivo techniquesled to expression of antigen were the basis for the generationof DNA vaccines [20]. DNA vaccines can consist of tumor-specific or tumor-associated antigens (TAAs) and additionalimmune-stimulatory factors cloned into a bacterial plasmiddownstream of an appropriate eukaryotic promoter forstrong and stable expression.

TAAs used for cancer vaccines are exogenous viral anti-gens expressed by virus-induced cancers, tumor-restrictedantigens also called neoantigens, tumor-associated differen-tiation antigens that are only expressed in specific tissues, orgenerally expressed antigens that are overexpressed in cancercells.

Cancer vaccines include MHC class I and class IIepitopes, multiple TAAs to effectively target the wholeinhomogeneous tumor population to decrease the risk ofimmune escape, and can contain TAAs that correspondto proteins involved in tumor transformation [21]. DNAminigenes are a special type of DNA vaccines harbouringonly short antigen epitopes which can efficiently induce acytotoxic T cell (CTL), B-cell or T-helper cell response. Theyare as effective as whole cDNA vaccines but without the riskof introducing a functional cDNA with possible devastatingconsequences [22–25]. This is especially important if anti-gens that are used can function as oncogenes. If whole cDNAvaccines are favoured, rearranged or mutated sequenceswere shown to be useful for full immunological activitywithout the risk of negative properties of the functionalprotein [26, 27]. DNA vaccines have many advantages ifcompared with classical vaccines; they combine the diversityof possible TAAs expressed on whole tumor cells or subunitvaccines with the efficiency of in vivo antigen synthesis andpresentation able to induce both cellular (CD4+ and CD8+

cells) and humoral immune responses [28].

The risks of DNA vaccines are limited [29]. Severalgroups demonstrated that cancer vaccines can be effectivein induction of specific immunity against cancer-associatedantigens without negative side effects like integration of plas-mid DNA into the host genomes or induction of pathogenicanti-DNA antibodies [30–38]. The results in animal modelsand initial clinical trials are promising but so far havenot resulted in a successful, standardized translation intothe clinic, emphasizing the enormous differences betweenanimal models and patients. Possible reasons may be relatedto the compromised immune system of the cancer patientsafter their chemo- and/or radiotherapy. Additionally, tumorsdevelop mechanisms to escape the immune system, such asthe loss of MHC class I molecules or antigen, so they cannotbe recognized by CTLs [39–42]. Other mechanisms are theoccurrence of regulatory T cells that negatively influencethe induction of anti-tumour responses, systemic defects inimmune cells, secretion of immunosuppressive cytokines,resistance to apoptosis, and many more [43, 44], which needto be addressed.

1.3. Effective Activation of the Patient’s Immune Systemby DNA Vaccines. DNA vaccines capable to activate thepatient’s immune system to effectively target cancer needthe activation of effector cells that are able to kill thetumor or can indirectly trigger a cascade that subsequentlylead to its eradication. Naıve T cells are a basic part ofthis complex system which are activated if they get twoindependent signals. The first signal is provided by bindingof a specific antigen-MHC class I complex to its T cellreceptor (TCR). The second costimulatory signal is diversein nature and may consist of CD40 expressed on antigen-presenting cells (APCs) or soluble factors such as cytokines(IL-2). To increase the immunogenicity of tumor cells, theywere manipulated to express costimulatory molecules and/orcytokines thereby significantly enhancing the induction of animmune response [45, 46].


The delivery system used for the application of anticancervaccines also plays an important role in increasing theamplitude of an immune response. Plasmid-based DNA vac-cines can be applied with ballistic delivery (gene gun) [47],liposomal or microsphere encapsulation [48] incorporatedin bacterial or host cell carriers [49, 50], or by electroporation[51–54]. The latter technique is used efficiently in preclinicaland clinical trials of melanoma and prostate cancer. Usingbacteria as DNA vaccine vehicles bears the advantage ofefficient stimulation of the innate immune system throughthe recognition of lipopolysaccharides (LPSs) in the bacteriaouter membrane. LPSs stimulate APCs by binding to Toll-like receptor 4 (TLR4) which subsequently supports anefficient activation of T cells, directly activates naturalkiller cells (NK cells), or leads to an increased lifetime ofantigen-specific T cells [12, 55]. Unmethylated CpG motifsincluded in DNA of bacterial origin have an additionalimmune stimulatory effect on cells of the innate immunesystem. Binding of CpGs to Toll-like receptor 9 (TLR9)expressed on DCs, NK cells, or monocytes/macrophagesleads to further maturation and activation of these cells andsubsequent secretion of proinflammatory cytokines of theTh 1 type including IL-12, TNF-α, IFN-α, and IFN-γ and theupregulation of costimulatory molecules such as CD80 andCD86 on APCs [56–58]. The dependency of TLR9 activationby CpGs is challenged by recent observations indicating thatthe DNA sugar backbone is also crucial for either activationor inhibition of TLRs by DNA. Natural DNA activates TLRswith phosphodiester (PD) backbone independent of CpGsin contrast to synthetic phosphorothioate- (PS-) modifiedDNA, which is TLR antagonistic. In the latter case, CpGmotifs can transform the antagonistic PS-modified DNAto a strong activator of TLRs restricting the dependencyof CpGs on TLR activation to this special case [59, 60].Furthermore, the detection of foreign non-CpG DNA isalso mediated by TLR-independent sensors leading to theexpression of interferon genes and induction of innateimmunity. One candidate sensor called DNA-dependentactivator of interferon regulatory factors (DAI) was shown todirectly interact with DNA in interplay with the interferonregulatory factor 3 transcription factor (IRF3) leading toa release of interferon-β in a TANK-binding kinase 1-(TBK1-) dependent manner [61–63]. Another crucial factorin the activation of the TBK1 pathway seems to be STING(stimulator of interferon genes) [64]. This novel independentpathway shows that AT-rich DNA can also serve as a templatefor RNA polymerase III leading to a transcription intodouble-stranded RNA (dsRNA). dsRNA subsequently actsas a ligand for the potential cytosolic DNA sensor RIG-I(Retinoic acid-induced gene I) and to a production of typeI interferons [65, 66].

1.4. Presentation of Antigens Encoded in DNA Vaccines.Based on the DNA vaccine delivery system and the DNAdesign of the antigen sequences, there are at least threedifferent mechanisms as DNA vaccines can be processedand presented in vivo. First, DNA vaccines can directlylead to production of the antigen by somatic cells likekeratinocytes or myocytes. These cells share a poor capability

to directly present processed antigen to immune cells byMHC class I and II molecules. Therefore, this mechanismis considered to play only a subordinate role. Second, theproduction of antigen by somatic cells may result in effectivepresentation to the immune system by a mechanism called“cross-priming”. Antigen may be released from the site ofproduction travelling to the draining lymph node, whereit is taken up by APCs processed and presented to T cells[67, 68]. Third, DNA vaccines also lead to the productionand direct presentation of antigen by professional APCs. Forthis purpose, the normal infection pathway of intracellularbacteria, for example, attenuated Salmonella thyphimurium,can be used for oral DNA vaccination. The bacteria enterthe host through the gastrointestinal tract after oral gavageand move through the M cells that cover the Peyer’s patches(lymph nodes) of the gut. From there, they enter APCslike macrophages or DCs by phagocytosis. In the APCs, thebacteria die, delivering multiple copies of the vaccine DNAthat can encode for antigen to the phagosome or cytosol[69, 70].

1.5. Processing DNA-Encoded Antigen. The activation ofCTLs which are key players in DNA vaccine-mediated tumorimmunity is induced by degradation of protein componentsinto smaller peptides and presentation of antigen by APCs[71, 72].

Processed antigens can be either presented by the endoge-nous or the exogenous pathway. CD8+ T cells, the precursorsof CTLs, are in general activated by intracellular pathogen-derived antigens of 8–10 amino acids length which arepresented by MHC class I molecules (endogenous pathway)[73, 74]. In contrast, CD4+ T-helper cells generally recognizeexogenous antigens presented as 12–15 amino acid longpeptides bound to MHC class II molecules. Upon activation,T-helper cells secrete cytokines, which is crucial for theinduction and maintenance of immunologic memory [75–78]. If antigens are transported to the cytoplasm or proteinsare produced endogenously, they are degraded by theproteasome into small peptide fragments. These peptides arethen transported by TAP1 and TAP2 into the endoplasmaticreticulum (ER) and bind to a dimer consisting of an MHCclass I molecule and β2-microglobulin. MHC class I antigencomplexes are then transported to the cell surface, where theyare presented. The trimolecular complex can be recognizedby CD8+ cytotoxic T cells. This mechanism is important forinduction and activation of antigen-specific CD8+ T cells byAPCs and for the effector function of CD8+ cytotoxic T cellsafter trimolecular complex recognition on the tumor cellresulting in subsequent target cell lysis [79–82]. Presentationof antigens via the endogenous pathway dominantly leadsto activation of Th 1 cells and CTL responses, whereas theexogenous pathway leads to the activation of Th 2 cellsand the production of antibodies [83, 84]. DNA vaccinescan increase the Th 1 immune response and the levels ofimmunoglobulins by directly inducing the expression ofinterferons, IL-12, IL-18, or TNF-α [85]. The activationof the exogenous pathway is usually insufficient to preventtumor growth in animal models of cancer but is in contrast


preferable in the case of protection against extracellularpathogens or in targeting chronic infections.

The introduction of specific sequences to the antigenlike an N-terminal ubiquitination signal can further enhancethe induced CTL response, preferentially of the Th 1 type[86]. It is possible to increase the protection against virusor tumor challenge by directing the antigen to specificcell compartments like the proteasome leading to fastdegradation and presentation of antigen to MHC class Imolecules [87]. Ubiquitination of DNA minigenes probablyleads to polyubiquitination of the cleaved peptide epitopesresulting in a more effective delivery of the minigene tothe proteasome and an increase of the frequency of CTLprecursors [24].

2. Melanoma

Malignant melanoma is a neuroectodermal solid tumoraffecting predominantly Caucasians. More than 160.000 newcases were diagnosed in 2002 with an increasing incidence.Despite favourable survival rates in the developed countriesof greater than 90% in early stages, around 40.000 deathswere caused by melanoma in 2002. In the US alone, morethan 10.000 people will probably die of skin cancer in 2010,and the prognosis of stage IV melanoma remains poor(<20% 5 y EFS) [1, 88]. Melanoma is a highly immunogeniccancer and maybe the most prominent model for DNAvaccination and the development of tumor vaccines ingeneral [89]. In patients, spontaneous complete melanomaremission is occasionally detected, a phenomenon medi-ated by endogenous CTLs and subsequent tumor rejection[90, 91]. A variety of melanocyte differentiation antigenswere identified as tumor antigens and found to be highlyexpressed in melanoma cells. Consequently, these antigenswere successfully used as targets for DNA vaccination, andtheir application in humans, dogs, and preclinical models arediscussed in the following paragraphs [92, 93].

Based on the successful use of gp100, MART-1/Melan-A, and tyrosinase, several clinical trials were conductedor are still ongoing. In an initial study, patients withmetastatic melanoma were immunized with human gp100(hgp100) expressing naked plasmids showing no clinicalor immunological responses, indicating that the deliverysystem and adequate costimulation play an important rolefor the success of this approach. For example, the use offowl poxvirus encoding for hgp100 or hgp100 peptides wascapable to break the self-tolerance against the gp100 tumorantigen [40]. Also the use of particle-mediated epidermaldelivery (PMED) of hgp100 cDNA in combination withcostimulatory granulocyte macrophage colony-stimulatingfactor (GM-CSF) into healthy skin of melanoma patientsrevealed the recruitment of DCs to the vaccination sites.Subsequently, a low but detectable antimelanoma immuneresponse was observed [94]. A role for GM-CSF DNA asan adjuvant was established in a phase I/II trial using aDNA vaccine encoding for hgp100 and tyrosinase epitopesresulting in specific CD8+ responses in 42% of the treatedmelanoma patients [95]. Another strategy effectively break

tolerance against self-antigens is the use of xenogeneicDNA vaccines. This is indicated by two recent phase Itrials using mouse gp100 (mgp100) DNA vaccines aloneor in combination with the human homologue. Melanomapatients immunized with the xenogeneic vaccines developedhgp100-specific and IFN-γ-secreting CD8+ T cells, and 30%of them showed an immune response [16, 96].

MART-1 is another melanoma antigen used in clinicalDNA vaccination trials. In an early phase I study, 12 patientswith resected melanoma received MART-1 plasmids, againwithout further adjuvant strategy. Immunological responseswere not detectable [97]. Similar poor results were observedin nineteen patients with stage IV melanoma which weretreated with a plasmid encoding T cell epitopes fromMART-1 and tyrosinase by intranodal injection [98]. Thevaccination approach induced an immune response in someof the patients but was not able to stop the progression ofthe disease. This again suggested that DNA vaccines have tobe used in combination with adjuvants, cytokines, or in thecontext of distinct prime/boost approaches to increase theimmune response for effective treatment of melanoma.

Changing the application system from naked plasmidto a viral delivery system may not be sufficient to translateimmune to clinical response demonstrated by clinical DNAvaccination trials using tyrosinase as antigen. Tyrosinase wasused as single cDNA vaccine applied in stage II melanomapatients by recombinant modified vaccinia virus Ankara(MVA). There was a strong immune response against thevirus, indicated by virus-specific CD4+ and CD8+ T cellsand antibody titres, but no tyrosinase-specific T cells orantibodies were detected [99]. In two subsequent clinicaltrials, tyrosinase was delivered by vaccinia or fowlpox viruseswhich were applied to patients with advanced metastaticmelanoma who also received systemic IL-2 [100]. Antityrosi-nase immunity was enhanced in some patients but withoutclinical benefit compared to effects expected for IL-2 alone.

Numerous DNA vaccination studies in animal modelsof melanoma demonstrated efficacy. Particular success wasreported for xenogeneic strategies, viral delivery systems,and the use of IL-2 as an adjuvant [101]. In mousemodels, efficacy of DNA vaccines encoding for all knownmelanoma-associated antigens was reported including gp100(melanocyte protein 17/Pmel-17), GRP (gastrin-releasingpeptide) [102], MAGE-1 (melanoma-associated antigen)[103], MART-1, MUC-18/MCAM [104], TRP-1 (tyrosinase-related protein-1/gp75), TRP-2, or tyrosinase. Also inhibitorof apoptosis proteins (IAPs) like ML-IAP (melanomainhibitor of apoptosis protein) [105] and survivin [106] wereused. In some approaches, less relevant antigens were usedsuch as melanoma cell lines stably transfected to express viralantigens of hepatitis B virus or HPV (human papillomavirus)or human oncogenes like Mucin 1 (MUC-1) [107–109]. Forvaccination studies in mice, usually the melanoma cell lineB16 is used syngeneic to C57BL/6 mice leading to melanomagrowth and metastasis serving as a model for the humandisease [110, 111]. In view of the limited efficacy of DNAvaccines in clinical trials so far, these successful studies inmurine models reflect the difficulty of the transfer of resultsfrom mice to man.


However, there are important lessons to be learned fromanimal models such as the important role of adjuvanticityand prime/boost schedules. The majority of vaccinationstudies in mice were done with gp100 as melanoma targetantigen, and some results will be discussed here in moredetail. In early studies, vaccination of mice by s.c. injec-tion of plasmid encoding for hgp100 antigen alone or incombination with GM-CSF DNA was conducted in a B16model transfected with hgp100 DNA (B16/hgp100) [112,113]. Protection against melanoma challenge and reductionof established primary tumor growth was observed, andhgp100-specific CTLs and antibodies were found. However,immunity against mpg100 was poor. Changing the deliverysystem by incorporation of the plasmid DNA expressinghgp100 into liposomes was able to induce an mgp100mediated protective immunity [114]. Vaccination induced adelayed primary tumor growth, a phenomenon which wasalso seen in another study using an mgp100 plasmid [115].Vaccination of mice with attenuated Salmonella typhimurium(ST) transformed with mgp100 cDNA significantly reducedmelanoma growth, an effect which was increased by IL-2administration [116]. In the B16/hgp100 model, vaccinationwith hgp100 transformed ST was able to completely protectmore than 70% of the vaccinated mice mediated by astrong anti-hgp100 CTL response [117]. The superiorityof xenogeneic vaccination approaches became obvious in astudy were hgp100 plasmids were applied via helium-drivenDNA-gold complexes leading to a tumor protection in manyof the mice which was mediated by specific CD8+ T cellswith the typical side effect of autoimmune depigmentation[118]. A comparison of hgp100 and mgp100 full-lengthDNA and minigenes revealed that only the human constructswere able to induce a CD8+ dependent tumor protectiveimmunity further strengthening the xenogeneic vaccinationconcept [119]. Tumor formation was completely preventedin more than 30% of the treated C57BL/6 mice when ahgp100 plasmid was used together with synthetic peptidesof putative CTL epitopes, an effect which was not observedif plasmid or peptides were used alone [120]. The usedstudy design also demonstrated a therapeutic effect of thecombinatorial setting and a dependency on CD4+ andCD8+ T cells for melanoma protection. Codelivery of IL-12DNA by direct injection into tumors was able to furtherincrease the antitumorale effect induced by hgp100 DNAvaccination [121]. An autologous gp100 plasmid was able tobreak self-tolerance when a different mice model was used,showing that differences in genetic backgrounds are criticalparameters for success of DNA vaccination [122]. DBA/2mice were challenged with either mgp100 positive or negativesyngeneic M3 melanoma cells leading to an mgp100-specificT-cell-mediated immune response and a protection againstmelanoma growth only if the mgp100 expressing cells wereused.

Similar results were obtained using TRP-1/gp75 DNAvaccines in the B16 model in several syngeneic and xenoge-neic settings. Murine TRP-1 (mTRP-1) expressing plasmidswere not able to break self-tolerance against mTRP-1 incontrast to the human homologue which induced immu-nity against mTRP-1 and subsequent tumor protection

and eradication [123]. Rejection of a lethal challenge withB16 melanoma cells was achieved with mTRP-1 encodedby a recombinant vaccinia virus again emphasizing thatthe adjuvanticity of the delivery system is important tobreak self-tolerance [124]. Boosting of DNA vaccination byapplication of monoclonal antibodies is another strategyused with a hTRP-1 DNA vaccine [125]. Lung metastasesinduced by B16 were significantly decreased if hTRP-1DNA was used with TA99, an antibody targeting TRP-1,a synergistic effect which was not seen with vaccine orantibody alone. The TA99 antibody seemed to be responsiblefor the recruitment of TRP-1-specific CD8+ T cells to thetumor and subsequent tumor infiltration.

Also DNA vaccination using TRP-2 as a target revealedthat vaccination of mice with murine TRP-2 (mTRP-2)using vaccinia virus as a delivery system was more effectivethan naked DNA injection [126, 127]. Again the xenogeneicconcept was more effective since hTRP-2 DNA prevented thegrowth of B16 cells in the skin of treated mice by activationof CD4+ and CD8+ T cells [128]. Additional treatment withhTRP-2 after surgical resection of affected extremities alsoreduced the reoccurrence of local disease and the number oflung metastases. In summary, adjuvanticity, delivery systems,and prime/boost schedules are important factors to considerfrom preclinical models for the design of an effective DNAvaccination strategy in humans. The design of the antigenincluding signals for proteasomal degradation also plays animportant role in vaccine efficacy. In this regard, fusionof mTRP-2 to ubiquitin facilitated proteasome-dependentdegradation of antigen and subsequent presentation ofepitopes to MHC-class I leading to the generation of mTRP-2-specific CD8+ T cells. These T cells were not only capableto protect against melanoma but also had a therapeuticeffect on established melanomas [129]. Other strategiesmay enhance vaccine efficacy including lymphodepletionwith cyclophosphamide and antigen fusion with heat shockproteins. The role of lymphodepletion in adoptive T celltransfer strategies has been demonstrated [130, 131]. Similarconcepts may apply for DNA vaccines. Adenoviral delivery ofhTRP-2 in combination with high-dose cyclophosphamidehad a synergistic effect and improved the outcome of tumor-bearing mice [132]. Also fusion antigens are an interestingapproach to further enhance immunogenicity. Fusion of heatshock protein 70 (Hsp70) to tumor antigens led to efficientdelivery of antigen to APCs thereby breaking the immunetolerance against melanoma cells [133]. An Hsp70-mTRP-2 DNA vaccine was orally applied by transfected attenuatedS. typhimurium strain SL3261 protecting more than half ofthe treated mice from a lethal challenge with B16 melanomacells in a prophylactic setting and prevented or significantlyreduced tumor growth in a therapeutic setting.

There are promising results with a xenogeneic DNAvaccination strategy in dogs with melanoma, which raisehope that this strategy will succeed in humans as well.Therapy with xenogeneic tyrosinase DNA vaccines wasused in phase I trials of spontaneous advanced malig-nant melanoma in dogs, a disease very similar to humanmelanoma. Intramuscular injections with human tyrosinase(hTyr) plasmids significantly increased the expected survival


of the dogs compared to matched historical controls, and onedog with stage IV disease had a complete clinical response[134]. In three of the nine treated dogs, tyrosinase-specificantibodies were induced by vaccination with hTyr DNAwhich partially reacted with the syngeneic canine tyrosinase,a phenomenon which correlated with the observed clinicalresponse and possibly responsible for the tumor static effectsand long-term survival of the dogs [135].

In summary, the human clinical trials led to promisingresults like activation of melanoma antigen-specific CTLsbut have so far not resulted in significant improvements inoutcome. Studies in small and large animals were neverthe-less able to demonstrate efficacy of DNA vaccination againstmelanoma. Some critical parameters were identified includ-ing the delivery systems, adjuvants, and antigen design. Ingeneral, the use of xenogeneic antigens often showed betterresults in the treatment of melanoma by DNA vaccination,and the use of viral application methods and cytokinesfurther increased immunogenicity. The consideration of thelessons learned from animal models for the design of DNAvaccination strategies may lead to an effective approach inthe future.

3. Neuroblastoma

Neuroblastoma (NB) is the most common extracranialsolid tumor in childhood. The prognosis is still poor, andthe development of effective treatment strategies is oneof the main objectives in pediatric oncology. Despite thedevelopment of innovative treatment strategies like passiveimmunotherapy with the anti-GD2 antibody ch14.18, thelong-time survival rate especially of stage 4 tumor patientsremains poor, ranging between 35 and 40% [136, 137].The power of immunotherapy was demonstrated in a recentphase III trial combining ch14.18, GM-CSF, and IL-2 withstandard therapy raising the hope to significantly increasethe long time survival rate of high-risk patients in the nearfuture. Patients who received the immunotherapy showedan improved outcome compared to standard therapy witha two-year event-free survival (EFS) of 66% and 46%,respectively [138].

In animal models, there are several promising resultsshowing that DNA vaccination is able to protect mice froma lethal challenge with neuroblastoma tumor cells. Manyresults in this respect were generated using the syngeneicNXS2 neuroblastoma mouse model. This hybrid cell lineexpresses ganglioside GD2 which is highly expressed in NBand, as an established tumor marker, is used as a target inclinical trials of NB immunotherapy.

The NXS2 model mimics the human disease in severalaspects. It features spontaneous metastasis to bone marrowand liver after injection of the cells in syngeneic A/J mice,making this system an ideal model to study DNA vaccination[139]. One example is cyclic mimicking decapeptides ofGD2 which were successfully used for DNA vaccinationgenerating protection against tumor growth and a reductionof spontaneous liver metastases [140]. Codelivery of thecytokines IL-15 and IL-21 enhanced the induction of GD2

directed responses which increased the CD8+ T cell function.The effects were NK cell as well as CD4+ and CD8+ T cellmediated indicating the involvement of innate and adaptiveimmune responses [141]. A plasmid that encoded for thesecreted form of HuD was able to induce a strong andspecific anti-HuD response in a similar mouse model usingA/J mice with the Neuro2a NB cell line. Mice that werechallenged with constitutively HuD-expressing Neuro2a cellswere protected against tumor growth after immunizationwith HuD DNA vaccine but showed no signs of neurologicaldisease induction [142].

The neuroblastoma antigen tyrosine hydroxylase (TH)is another promising candidate for immunotherapy ofneuroblastoma. Tyrosine hydroxylase is involved in the firststep of catecholamine biosynthesis, a unique feature similarto melanin biosynthesis in melanoma, involving enzymesrestricted to the tumor tissue therefore providing tumor-associated antigens.

Prophylactic and therapeutic vaccination with murineTH cDNA or TH minigenes was able to protect againsttumor growth after delivering plasmid DNA by oral gavage ofattenuated S. thyphimurium to the mice. The used expressionvector contained a mutated ubiquitin leading to the expres-sion of ubiquitin-DNA fusion proteins that were efficientlydegraded in the proteasome. T cells recognized the TH self-antigen epitopes indicating that the self-tolerance against THcan be overcome with this approach [143].

In subsequent studies the same mTH-based minigenes,novel epitopes, and xenogeneic TH DNA vaccination wereeffective in therapeutic settings to suppress establishedneuroblastoma metastases. Modifications of mTH had addi-tional positive effects. The mutated ubiquitin of the usedplasmid was crucial for the strong antitumoral effect leadingto a CD8+ T cell immune response. Primary tumorswere infiltrated by CD8+ T cells, and TH-expressing cellswere specifically lysed in vitro. Depletion of CD8+ T cellscompletely abrogated the anti-NB immune response inducedby the hTH vaccine. Rechallenge of surviving mice resulted inreduced primary tumor growth, indicating the induction ofa memory immune response. An important observation wasthat immunization with the self-antigen TH did not lead toautoimmunity [18, 144, 145].

In a different study using the same mouse model, novelnatural MHC class I ligands from neuroblastoma werecharacterised and used in a DNA minigene approach. Immu-nization of mice induced protective immunity and thusunderlines the assumption that disruption of self-toleranceagainst neuroblastoma-associated epitopes is important foran effective neuroblastoma immunotherapy [146].

The inhibitor of apoptosis protein (IAP) survivin ishighly expressed in neuroblastoma and is associated with apoor prognosis. Therefore, survivin was chosen as a target forNB DNA vaccination. A survivin DNA minigene efficientlyinhibited the growth of primary tumor and metastases inthe NXS2 tumor model. The used DNA minigene was aseffective as a survivin full-length cDNA vaccine showing thepower of DNA minigene vaccination. Immunization withsurvivin minigene was associated with an increased presenceof CD8+ T cells in the primary tumor and production


of the proinflammatory cytokines INF-γ and TNF-α bysystemic CD8+ T cells. Depletion of CD8+ but not CD4+

cells led to a complete abrogation of the tumor immunity.Therapeutic vaccination with the minigene was able toeradicate neuroblastoma in more than half of the mice, andsurviving mice were protected from primary tumor growthafter rechallenge with tumor cells [147].

In summary, these preclinical studies suggest that DNAvaccination using S. thyphimurium as a delivery systemmay provide an important strategy to develop an activeimmunotherapy strategy for this challenging disease.

4. Prostate Cancer

Prostate cancer is the most common cancer and the fourthleading cause of cancer-related death in men in the developedcountries worldwide [1, 88, 148]. Standard prostate cancertherapy in early diagnosed patients involves prostatectomy,cryotherapy, radiotherapy, and antiandrogen therapy. Thesetreatments are effective but bear the risk of severe side effectslike incontinence and impotence [149–151]. Therefore, thereis an urgent need for novel approaches to treat this disease.

CD4+ and CD8+ T cells are detectable in prostate glandsof men with prostate cancer supporting the assumption thatprostate cancer might be a good candidate for immunother-apy. Prostate cancer cells are usually growing rather slow,permitting enough time to use vaccination as an approachto overcome immunosuppressive factors [43]. Currently,there are several clinical trials using immunotherapy againstprostate cancer targeting prostate cancer-associated anti-gens including Prostate-Specific Antigen (PSA) [152], Six-Transmembrane Epithelial Antigen of the Prostate (STEAP)[153], Prostate Stem Cell Antigen (PSCA) [154], Prostate-Specific Membrane Antigen (PSMA) [54, 155], and ProstaticAcid Phosphatase (PAP) [156].

PSMA was one of the first prostate cancer-associatedantigens used for DNA vaccination. Plasmid DNA andadenoviral vectors encoding for PSMA were used to immu-nize patients with prostate cancer in a phase I/II trial.Costimulation of plasmid DNA with the molecule CD86led to delayed-type hypersensibility to PSMA in half of thepatients, but additional boosting with the adenovirus wasnecessary to induce immunity in all of them. However,the success of this study is difficult to interpret due to theheterogeneity of the patients, and the concomitant hormonetherapy many of the patients received albeit local disease,distant metastases, and PSA levels changed positively [39]. Inanother study, patients were vaccinated against PSMA withplasmid DNA and adenovirus as well leading to the detectionof specific anti-PSMA antibodies in the sera of the patients[157]. A recent clinical phase I/II demonstrated the power ofelectroporation in induction of a humoral immune responseagainst prostate cancer. PSMA-specific DNA vaccines weredelivered by intramuscular injection or in combinationwith an additional delivery by electroporation (EP). Theboosting by EP significantly enhanced (24.5-fold increase)the immune response during an 18-month followup period[54].

In summary, the efforts made in clinical trials usingDNA vaccination against prostate cancer are promising,but the response rates have to be improved. Therefore,the evaluation and characterization of DNA vaccinationstrategies in preclinical models is an important venue. Thefollowing paragraphs summarize some of the research in thisrespect.

One approach is the use of xenogeneic vaccinationstrategies. The effectiveness of mouse PSMA and humanPSMA (hPSMA) DNA vaccines were tested in an animalmodel indicating that only xenogeneic hPSMA was able toinduce both antibody as well as T cell responses againstthe murine self-antigen. The antibodies induced were ableto recognize the human and the murine PSMA, and itwas concluded that xenogeneic DNA is a requirement toovercome the immunologic tolerance against the poorlyimmunogenic PSMA in contrast to other studies [158].These results were improved in a setting with immunizationof mice using xenogeneic hPSMA DNA followed by boostingwith hPSMA protein [17].

Further improvements can be achieved by modificationsof DNA vaccines leading to expression and proteasomaldegradation of hPSMA in combination with protein boost-ing. This resulted in antibody formation of the cytotoxic Th 1isotypes, and the best protection against tumor challenge wasobserved after immunization with the xenogeneic hPSMAconstruct following boosting with the syngeneic construct[159].

The use of cytokines in order to amplify subopti-mal immune responses following Prostate-Specific Antigen(PSA) DNA vaccination is another strategy to increase DNAvaccine efficacy. A DNA vaccine expressing PSA inducedPSA-specific CTLs when coinjected with the costimulatorycytokines IL-2 and GM-CSF and protected the majority ofimmunized mice against a lethal tumor challenge [160].

Also, the delivery system is an important factor to induceeffective immunity against prostate cancer. Intradermalimmunization of mice with PSA induced strong humoraland cellular immune responses of the Th 1 isotype indicatedby strong expression of INF-γ and IL-2 and protectedmice from challenge with PSA-expressing tumor cells [161].Intramuscular electroporation with human PSA (hPSA)DNA significantly reduced tumor growth and increased thesurvival of mice after a lethal challenge with hPSA-expressingTRAMP-C1 cells, a cell line developed from a prostatetumor of a TRAMP (transgenic adenocarcinoma mouseprostate) mouse. Production of hPSA-specific antibodies andexpression of IFN-γ was observed in the immunized animals[51]. Multiple CTL and T-helper cell epitopes of hPSMA,mPAP, and hPSA were combined to generate a DNA vaccinethat should have a stronger effect against prostate tumor cellsthan single antigen vaccines. The vaccine design was chosento overcome the problem that tumor cells often lose antigenicepitopes and escape immunologic detection. Vaccinationof mice by gene gun induced a strong immune responseagainst applied tumor cells and increased the survival timesignificantly [162]. A systematic comparison with otherdelivery systems including life-attenuated bacteria was notreported so far.


Other antigens investigated in preclinical models are theProstatic Acid Phosphatase (PAP) and Prostate Stem CellAntigen (PSCA). PAP is effective in inducing proliferatingPAP-specific CD4+ and CD8+ T cells of the Th 1 isotypeand expression of IFN-γ in rats immunized with a humanPAP DNA vaccine [163, 164]. Application of a PAP DNAvaccine to prostate cancer patients induced a PAP-specificT cell response and showed no side effects making the useof PAP in clinical stage II trials likely [165]. Vaccinationof mice with plasmid encoding for human PSCA inducedstrong PSCA-specific CD8+ T cell responses and inhibitedthe growth of PSCA-positive tumors [166]. There are nocomparative studies allowing a decision about which antigenmight be the best choice.

The models used to study DNA vaccination in prostatecancer are usually based on the TRAMP model. In TRAMPmice, the Simian virus 40 (SV40) tumor T-antigens areprostate-specifically expressed driving prostate neoplasia,which results in poorly differentiated adenocarcinomas ofthe prostate and metastasis in lung and pelvic lymph nodesresembling the human disease [167, 168]. TRAMP micewere strongly protected against tumor development whenvaccinated with PSCA-based DNA, an effect supposedlymediated by CD8+ T cells and expression of the cytokinesINF-γ, TNF-α, IL-2, IL-4, and IL-15 within prostate tumors.Importantly long-term protection was not accompanied byan induction of autoimmunity [169]. From this mousemodel, TRAMP-C1 tumor cells were isolated which growin syngeneic C57BL/6 mice. In this model, vaccination withPSCA DNA by intramuscular electroporation induced aneffective antitumor response, and the mice were either curedor showed a significant increase in survival, which wasmediated by an immunity of the Th 1 type [170].

An interesting discovery was made when two thera-peutic vaccination studies with the antigens PSCA andSTEAP in the TRAMP mouse model were compared. DNAvaccination at an early stage of disease resulted in animproved protection against tumor development and anincreased survival time when compared to vaccination afterthe development of invasive carcinoma. Regulatory T cellsas well as the expression of immunosuppressive factors likeTGF-β and indoleamine-2,3-dioxygenase were detected inmore advanced prostate cancer making the use of DNAvaccination at earlier stages of disease more promising [171].

In summary, DNA vaccination against prostate cancerhas demonstrated effectiveness in preclinical models, andpromising immune responses were observed in early clinicaltrials. Given the amount of preclinical information onselection and design of suitable and effective antigens,a system biology approach may provide an importantvenue to translate available information into an effectiveimmunotherapy.

5. Summary

DNA vaccination is a young field in immunotherapy ofcancer and has certainly not yet lived up to its expectations.However, considering the fact that the development of anti-bodies into effective cancer therapeutics followed a timeline

of over a century, DNA vaccination may be consideredto be on a fast track development. Preclinical data arevery promising and significant immune responses can bedemonstrated in several clinical trials especially in the field ofmelanoma DNA vaccination. The high versatility, the ease ofproduction, and the stability of DNA vaccines may provideimportant characteristics to further develop this approachinto effective cancer therapies of the 21st century.

Acknowledgments

This work was supported by NGFNplus (Bundesminis-terium fur Bildung und Forschung, ENGINE) and MedSys—Therapeutische Systemimmunologie (Bundesministeriumfur Bildung und Forschung).

References

[1] A. Jemal, R. Siegel, J. Xu, and E. Ward, “Cancer statistics,2010,” Cancer Journal for Clinicians, vol. 60, no. 5, pp. 277–300, 2010.

[2] M. Heron, D. L. Hoyert, S. L. Murphy, J. Xu, K. D. Kochanek,and B. Tejada-Vera, “Deaths: final data for 2006,” NationalVital Statistics Reports, vol. 57, no. 14, pp. 1–134, 2009.

[3] T. J. Stewart and S. I. Abrams, “How tumours escape massdestruction,” Oncogene, vol. 27, no. 45, pp. 5894–5903, 2008.

[4] M. Elrefaei, K. J. Blank, and D. M. Murasko, “Decreased IL-2,IFN-γ, and IL-10 production by aged mice during the acutephase of E55+ retrovirus infection,” Virology, vol. 299, no. 1,pp. 8–19, 2002.

[5] Z. F. Kapasi, K. Murali-Krishna, M. L. McRae, and R. Ahmed,“Defective generation but normal maintenance of memory Tcells in old mice,” European Journal of Immunology, vol. 32,no. 6, pp. 1567–1573, 2002.

[6] Y. Naito, K. Saito, K. Shiiba et al., “CD8+ T cells infiltratedwithin cancer cell nests as a prognostic factor in humancolorectal cancer,” Cancer Research, vol. 58, no. 16, pp. 3491–3494, 1998.

[7] L. Zhang, J. R. Conejo-Garcia, D. Katsaros et al., “Intratu-moral T cells, recurrence, and survival in epithelial ovariancancer,” New England Journal of Medicine, vol. 348, no. 3, pp.203–213, 2003.

[8] S. A. Birkeland, H. H. Storm, L. U. Lamm et al., “Cancerrisk after renal transplantation in the nordic countries, 1964–1986,” International Journal of Cancer, vol. 60, no. 2, pp. 183–189, 1995.

[9] I. Penn, “Malignant melanoma in organ allograft recipients,”Transplantation, vol. 61, no. 2, pp. 274–278, 1996.

[10] S. Ward, D. Casey, M.-C. Labarthe et al., “Immunothera-peutic potential of whole tumour cells,” Cancer Immunology,Immunotherapy, vol. 51, no. 7, pp. 351–357, 2002.

[11] A. N. Houghton, “Cancer antigens: immune recognition ofself and altered self,” Journal of Experimental Medicine, vol.180, no. 1, pp. 1–4, 1994.

[12] M. A. Liu and J. B. Ulmer, “Human clinical trials of plasmidDNA vaccines,” Advances in Genetics, vol. 55, pp. 25–40, 2005.

[13] W. W. Leitner, H. Ying, and N. P. Restifo, “DNA and RNA-based vaccines: principles, progress and prospects,” Vaccine,vol. 18, no. 9-10, pp. 765–777, 1999.

[14] W. W. Leitner, L. N. Hwang, M. J. De Veer et al., “Alphavirus-based DNA vaccine breaks immunological tolerance by


activating innate antiviral pathways,” Nature Medicine, vol. 9,no. 1, pp. 33–39, 2003.

[15] H. Ying, T. Z. Zaks, R.-F. Wang et al., “Cancer therapy using aself-replicating RNA vaccine,” Nature Medicine, vol. 5, no. 7,pp. 823–827, 1999.

[16] B. A. Ginsberg, H. F. Gallardo, T. S. Rasalan et al., “Immuno-logic response to xenogeneic gp100 DNA in melanomapatients: comparison of particle-mediated epidermal deliverywith intramuscular injection,” Clinical Cancer Research, vol.16, no. 15, pp. 4057–4065, 2010.

[17] P. D. Gregor, J. D. Wolchok, V. Turaga et al., “Inductionof autoantibodies to syngeneic prostate-specific membraneantigen by xenogeneic vaccination,” International Journal ofCancer, vol. 116, no. 3, pp. 415–421, 2005.

[18] N. Huebener, S. Fest, K. Hilt et al., “Xenogeneic immu-nization with human tyrosine hydroxylase DNA vaccinessuppresses growth of established neuroblastoma,” MolecularCancer Therapeutics, vol. 8, no. 8, pp. 2392–2401, 2009.

[19] J. D. Wolchok, J. Yuan, A. N. Houghton et al., “Safety andimmunogenicity of tyrosinase DNA vaccines in patients withmelanoma,” Molecular Therapy, vol. 15, no. 11, pp. 2044–2050, 2007.

[20] J. A. Wolff, R. W. Malone, P. Williams et al., “Direct genetransfer into mouse muscle in vivo,” Science, vol. 247,no. 4949, part I, pp. 1465–1468, 1990.

[21] T. Renno, S. Lebecque, N. Renard, S. Saeland, and A. Vicari,“What’s new in the field of cancer vaccines?” Cellular andMolecular Life Sciences, vol. 60, no. 7, pp. 1296–1310, 2003.

[22] I. F. Ciernik, J. A. Berzofsky, and D. P. Carbone, “Inductionof cytotoxic T lymphocytes and antitumor immunity withDNA vaccines expressing single T cell epitopes,” Journal ofImmunology, vol. 156, no. 7, pp. 2369–2375, 1996.

[23] A. Iwasaki, C. S. Dela Cruz, A. R. Young, and B. H.Barber, “Epitope-specific cytotoxic T lymphocyte inductionby minigene DNA immunization,” Vaccine, vol. 17, no. 15-16, pp. 2081–2088, 1999.

[24] F. Rodriguez, L. L. An, S. Harkins et al., “DNA immunizationwith minigenes: low frequency of memory cytotoxic T lym-phocytes and inefficient antiviral protection are rectified byubiquitination,” Journal of Virology, vol. 72, no. 6, pp. 5174–5181, 1998.

[25] Z. Yu, K. L. Karem, S. Kanangat, E. Manickan, and B. T.Rouse, “Protection by minigenes: a novel approach of DNAvaccines,” Vaccine, vol. 16, no. 17, pp. 1660–1667, 1998.

[26] J. P. Nkolola, E. G.-T. Wee, E.-J. Im et al., “EngineeringRENTA, a DNA prime-MVA boost HIV vaccine tailored forEastern and Central Africa,” Gene Therapy, vol. 11, no. 13,pp. 1068–1080, 2004.

[27] J. Zur Megede, G. R. Otten, B. Doe et al., “Expression andimmunogenicity of sequence-modified human immunodefi-ciency virus type 1 subtype B pol and gagpol DNA vaccines,”Journal of Virology, vol. 77, no. 11, pp. 6197–6207, 2003.

[28] J. B. Ulmer, J. J. Donnelly, S. E. Parker et al., “Heterologousprotection against influenza by injection of DNA encodinga viral protein,” Science, vol. 259, no. 5102, pp. 1745–1749,1993.

[29] D. M. Klinman, M. Takeno, M. Ichino et al., “DNAvaccines: safety and efficacy issues,” Springer Seminars inImmunopathology, vol. 19, no. 2, pp. 245–256, 1997.

[30] M. L. Disis, T. A. Gooley, K. Rinn et al., “Generationof T-cell immunity to the HER-2/neu protein after activeimmunization with HER-2/neu peptide-based vaccines,”Journal of Clinical Oncology, vol. 20, no. 11, pp. 2624–2632,2002.

[31] A. B. Heimberger, G. E. Archer, L. E. Crotty et al., “Dendriticcells pulsed with a tumor-specific peptide induce long-lastingimmunity and are effective against murine intracerebralmelanoma,” Neurosurgery, vol. 50, no. 1, pp. 158–164, 2002.

[32] B. J. Ledwith, S. Manam, P. J. Troilo et al., “Plasmid DNAvaccines: assay for integration into host genomic DNA,”Developments in Biologicals, vol. 104, pp. 33–43, 2000.

[33] S. Manam, B. J. Ledwith, A. B. Barnum et al., “Plasmid DNAvaccines: tissue distribution and effects of DNA sequence,adjuvants and delivery method on integration into hostDNA,” Intervirology, vol. 43, no. 4–6, pp. 273–281, 2000.

[34] S. M. Pupa, A. M. Invernizzi, S. Forti et al., “Prevention ofspontaneous neu-expressing mammary tumor developmentin mice transgenic for rat protoneu by DNA vaccination,”Gene Therapy, vol. 8, no. 1, pp. 75–79, 2001.

[35] M. J. Scanlan, A. O. Gure, A. A. Jungbluth, L. J. Old, andY.-T. Chen, “Cancer/testis antigens: an expanding family oftargets for cancer immunotherapy,” Immunological Reviews,vol. 188, pp. 22–32, 2002.

[36] J. A. C. Schalk, F. R. Mooi, G. A. M. Berbers, L. A. G. J. M.Van Aerts, H. Ovelgonne, and T. G. Kimman, “Preclinical andclinical safety studies on DNA vaccines,” Human Vaccines,vol. 2, no. 2, pp. 45–53, 2006.

[37] M. M. Soares, V. Mehta, and O. J. Finn, “Three differentvaccines based on the 140-amino acid MUC1 peptidewith seven tandemly repeated tumor-specific epitopes elicitdistinct immune effector mechanisms in wild-type versusMUC1-transgenic mice with different potential for tumorrejection,” Journal of Immunology, vol. 166, no. 11, pp. 6555–6563, 2001.

[38] P. Van der Bruggen, Y. Zhang, P. Chaux et al., “Tumor-specific shared antigenic peptides recognized by human Tcells,” Immunological Reviews, vol. 188, pp. 51–64, 2002.

[39] M. Mincheff, S. Tchakarov, S. Zoubak et al., “NakedDNA and adenoviral immunizations for immunotherapy ofprostate cancer: a phase I/II clinical trial,” European Urology,vol. 38, no. 2, pp. 208–217, 2000.

[40] S. A. Rosenberg, J. C. Yang, R. M. Sherry et al., “Inability toimmunize patients with metastatic melanoma using plasmidDNA encoding the gp100 melanoma-melanocyte antigen,”Human Gene Therapy, vol. 14, no. 8, pp. 709–714, 2003.

[41] S. T. Tagawa, P. Lee, J. Snively et al., “Phase I study ofintranodal delivery of a plasmid DNA vaccine for patientswith Stage IV melanoma,” Cancer, vol. 98, no. 1, pp. 144–154,2003.

[42] J. M. Timmerman, G. Singh, G. Hermanson et al., “Immuno-genicity of a plasmid DNA vaccine encoding chimericidiotype in patients with B-cell lymphoma,” Cancer Research,vol. 62, no. 20, pp. 5845–5852, 2002.

[43] C. G. Drake, E. Jaffee, and D. M. Pardoll, “Mechanismsof immune evasion by tumors,” Advances in Immunology,vol. 90, pp. 51–81, 2006.

[44] B. Seliger, “Strategies of tumor immune evasion,” Biodrugs,vol. 19, no. 6, pp. 347–354, 2005.

[45] G. Dranoff, “GM-CSF-based cancer vaccines,” ImmunologicalReviews, vol. 188, pp. 147–154, 2002.

[46] E. M. Jaffee, R. H. Hruban, B. Biedrzycki et al., “Novel allo-geneic granulocyte-macrophage colony-stimulating factor-secreting tumor vaccine for pancreatic cancer: a phase Itrial of safety and immune activation,” Journal of ClinicalOncology, vol. 19, no. 1, pp. 145–156, 2001.

[47] K. Aravindaram and N. S. Yang, “Gene gun delivery systemsfor cancer vaccine approaches,” Methods in Molecular Biology,vol. 542, pp. 167–178, 2009.


[48] R. A. Schwendener, B. Ludewig, A. Cerny, and O. Engler,“Liposome-based vaccines,” Methods in Molecular Biology,vol. 605, pp. 163–175, 2010.

[49] E. Manickan, S. Kanangat, R. J. D. Rouse, Z. Yu, and B. T.Rouse, “Enhancement of immune response to naked DNAvaccine by immunization with transfected dendritic cells,”Journal of Leukocyte Biology, vol. 61, no. 2, pp. 125–132, 1997.

[50] D. R. Sizemore, A. A. Branstrom, and J. C. Sadoff, “Attenu-ated bacteria as a DNA delivery vehicle for DNA-mediatedimmunization,” Vaccine, vol. 15, no. 8, pp. 804–807, 1997.

[51] S. Ahmad, G. Casey, P. Sweeney, M. Tangney, and G. C.O’Sullivan, “Optimised electroporation mediated DNA vac-cination for treatment of prostate cancer,” Genetic Vaccinesand Therapy, vol. 8, no. 1, article 1, 2010.

[52] P. Chiarella, V. M. Fazio, and E. Signori, “Application ofelectroporation in DNA vaccination protocols,” Current GeneTherapy, vol. 10, no. 4, pp. 281–286, 2010.

[53] L. C. Heller and R. Heller, “Electroporation gene therapypreclinical and clinical trials for melanoma,” Current GeneTherapy, vol. 10, no. 4, pp. 312–317, 2010.

[54] L. Low, A. Mander, K. McCann et al., “DNA vaccinationwith electroporation induces increased antibody responses inpatients with prostate cancer,” Human Gene Therapy, vol. 20,no. 11, pp. 1269–1278, 2009.

[55] A.-K. Yi, J.-G. Yoon, S.-C. Hong, T. W. Redford, and A.M. Krieg, “Lipopolysaccharide and CpG DNA synergize fortumor necrosis factor-α production through activation ofNF-κB,” International Immunology, vol. 13, no. 11, pp. 1391–1404, 2001.

[56] S. Akira, K. Takeda, and T. Kaisho, “Toll-like receptors:critical proteins linking innate and acquired immunity,”Nature Immunology, vol. 2, no. 8, pp. 675–680, 2001.

[57] R. Medzhitov, “CpG DNA: security code for host defense,”Nature Immunology, vol. 2, no. 1, pp. 15–16, 2001.

[58] K. Yasuda, H. Wagner, and Y. Takakura, “Role of immunos-timulatory DNA and TLR9 in gene therapy,” Critical Reviewsin Therapeutic Drug Carrier Systems, vol. 23, no. 2, pp. 89–109, 2006.

[59] T. Haas, J. Metzger, F. Schmitz et al., “The DNA sugarbackbone 2′ deoxyribose determines Toll-like receptor 9activation,” Immunity, vol. 28, no. 3, pp. 315–323, 2008.

[60] H. Wagner, “The sweetness of the DNA backbone drives Toll-like receptor 9,” Current Opinion in Immunology, vol. 20, no.4, pp. 396–400, 2008.

[61] K. J. Ishii, C. Coban, H. Kato et al., “A toll-like receptor-independent antiviral response induced by double-strandedB-form DNA,” Nature Immunology, vol. 7, no. 1, pp. 40–48,2006.

[62] A. Takaoka, Z. Wang, M. K. Choi et al., “DAI (DLM-1/ZBP1)is a cytosolic DNA sensor and an activator of innate immuneresponse,” Nature, vol. 448, no. 7152, pp. 501–505, 2007.

[63] Z. Wang, M. K. Choi, T. Ban et al., “Regulation of innateimmune responses by DAI (DLM-1/ZBP1) and other DNA-sensing molecules,” Proceedings of the National Academy ofSciences of the United States of America, vol. 105, no. 14,pp. 5477–5482, 2008.

[64] H. Ishikawa, Z. Ma, and G. N. Barber, “STING regulatesintracellular DNA-mediated, type i interferon-dependentinnate immunity,” Nature, vol. 461, no. 7265, pp. 788–792,2009.

[65] A. Ablasser, F. Bauernfeind, G. Hartmann, E. Latz, K. A.Fitzgerald, and V. Hornung, “RIG-I-dependent sensing ofpoly(dA:dT) through the induction of an RNA polymerase

III-transcribed RNA intermediate,” Nature Immunology,vol. 10, no. 10, pp. 1065–1072, 2009.

[66] Y.-H. Chiu, J. B. MacMillan, and Z. J. Chen, “RNApolymerase III detects cytosolic DNA and induces Type Iinterferons through the RIG-I pathway,” Cell, vol. 138, no. 3,pp. 576–591, 2009.

[67] S. Gurunathan, D. M. Klinman, and R. A. Seder, “DNA vac-cines: immunology, application, and optimization,” AnnualReview of Immunology, vol. 18, pp. 927–974, 2000.

[68] M. A. Liu, “DNA vaccines: a review,” Journal of InternalMedicine, vol. 253, no. 4, pp. 402–410, 2003.

[69] A. Darji, C. A. Guzman, B. Gerstel et al., “Oral somatictransgene vaccination using attenuated S. typhimurium,”Cell, vol. 91, no. 6, pp. 765–775, 1997.

[70] P. Paglia, E. Medina, I. Arioli, C. A. Guzman, and M.P. Colombo, “Gene transfer in dendritic cells, induced byoral DNA vaccination with Salmonella typhimurium, resultsin protective immunity against a murine fibrosarcoma,”Blood, vol. 92, no. 9, pp. 3172–3176, 1998.

[71] M. Corr, D. J. Lee, D. A. Carson, and H. Tighe, “Genevaccination with naked plasmid DNA: mechanism of CTLpriming,” Journal of Experimental Medicine, vol. 184, no. 4,pp. 1555–1560, 1996.

[72] J. B. Ulmer, R. R. Deck, C. M. Dewitt, J. J. Donnelly, andM. A. Liu, “Generation of MHC class I-restricted cytotoxicT lymphocytes by expression of a viral protein in musclecells: antigen presentation by non-muscle cells,” Immunology,vol. 89, no. 1, pp. 59–67, 1996.

[73] M. E. Dudley, J. R. Wunderlich, J. C. Yang et al., “Adop-tive cell transfer therapy following non-myeloablative butlymphodepleting chemotherapy for the treatment of patientswith refractory metastatic melanoma,” Journal of ClinicalOncology, vol. 23, no. 10, pp. 2346–2357, 2005.

[74] D. J. Powell Jr., M. E. Dudley, K. A. Hogan, J. R. Wun-derlich, and S. A. Rosenberg, “Adoptive transfer of vaccine-induced peripheral blood mononuclear cells to patients withmetastatic melanoma following lymphodepletion,” Journal ofImmunology, vol. 177, no. 9, pp. 6527–6539, 2006.

[75] C. Bourgeois, H. Veiga-Fernandes, A.-M. Joret, B. Rocha,and C. Tanchot, “CD8 lethargy in the absence of CD4 help,”European Journal of Immunology, vol. 32, no. 8, pp. 2199–2207, 2002.

[76] E. M. Janssen, E. E. Lemmens, T. Wolfe, U. Christen, M.G. Von Herrath, and S. P. Schoenberger, “CD4+ T cells arerequired for secondary expansion and memory in CD8+ Tlymphocytes,” Nature, vol. 421, no. 6925, pp. 852–856, 2003.

[77] D. J. Shedlock and H. Shen, “Requirement for CD4 T cell helpin generating functional CD8 T cell memory,” Science, vol.300, no. 5617, pp. 337–339, 2003.

[78] J. C. Sun and M. J. Bevan, “Defective CD8 T cell memoryfollowing acute infection without CD4 T cell help,” Science,vol. 300, no. 5617, pp. 339–342, 2003.

[79] K. Keusekotten, R. M. Leonhardt, S. Ehses, and M. R. Knittler,“Biogenesis of functional antigenic peptide transporter TAPrequires assembly of pre-existing TAP1 with newly synthe-sized TAP2,” Journal of Biological Chemistry, vol. 281, no. 26,pp. 17545–17551, 2006.

[80] P.-M. Kloetzel, “The proteasome and MHC class I antigenprocessing,” Biochimica et Biophysica Acta, vol. 1695, no. 1–3,pp. 225–233, 2004.

[81] M. Kovacsovics-Bankowski and K. L. Rock, “A phagosome-to-cytosol pathway for exogenous antigens presented onMHC class I molecules,” Science, vol. 267, no. 5195, pp. 243–246, 1995.


[82] A. N. Martinez, S. Tenzer, and H. Schild, “T-cell epitopeprocessing (the epitope flanking regions matter),” Methods inMolecular Biology, vol. 524, pp. 407–415, 2009.

[83] R. Kennedy and E. Celis, “Multiple roles for CD4+ T cellsin anti-tumor immune responses,” Immunological Reviews,vol. 222, no. 1, pp. 129–144, 2008.

[84] A. Yates, C. Bergmann, J. Leo Van Hemmen, J. Stark, andR. Callard, “Cytokine-modulated regulation of helper T cellpopulations,” Journal of Theoretical Biology, vol. 206, no. 4,pp. 539–560, 2000.

[85] M. G. Tovey and C. Lallemand, “Adjuvant activity ofcytokines,” Methods in Molecular Biology, vol. 626, pp. 287–309, 2010.

[86] G. Delogu, A. Howard, F. M. Collins, and S. L. Morris, “DNAvaccination against tuberculosis: expression of a ubiquitin-conjugated tuberculosis protein enhances antimycobacterialimmunity,” Infection and Immunity, vol. 68, no. 6, pp. 3097–3102, 2000.

[87] F. Rodriguez, J. Zhang, and J. L. Whitton, “DNA immuniza-tion: ubiquitination of a viral protein enhances cytotoxic T-lymphocyte induction and antiviral protection but abrogatesantibody induction,” Journal of Virology, vol. 71, no. 11,pp. 8497–8503, 1997.

[88] D. M. Parkin, F. Bray, J. Ferlay, and P. Pisani, “Global cancerstatistics, 2002,” Cancer Journal for Clinicians, vol. 55, no. 2,pp. 74–108, 2005.

[89] L. H. Brinckerhoff, L. W. Thompson, and C. L. Slingluff Jr.,“Melanoma vaccines,” Current Opinion in Oncology, vol. 12,no. 2, pp. 163–173, 2000.

[90] N. Halama, I. Zoernig, and D. Jaeger, “Advanced malig-nant melanoma: immunologic and multimodal therapeuticstrategies,” Journal of Oncology, vol. 2010, Article ID 689893,8 pages, 2010.

[91] Y.-T. Chen, E. Stockert, S. Tsang, K. A. Coplan, and L. J.Old, “Immunophenotyping of melanomas for tyrosinase:implications for vaccine development,” Proceedings of theNational Academy of Sciences of the United States of America,vol. 92, no. 18, pp. 8125–8129, 1995.

[92] G. J. Prud’homme, “DNA vaccination against tumors,”Journal of Gene Medicine, vol. 7, no. 1, pp. 3–17, 2005.

[93] G. Parmiani, C. Castelli, L. Rivoltini et al., “Immunotherapyof melanoma,” Seminars in Cancer Biology, vol. 13, no. 6,pp. 391–400, 2003.

[94] R. D. Cassaday, P. M. Sondel, D. M. King et al., “A phaseI study of immunization using particle-mediated epidermaldelivery of genes for gp100 and GM-CSF into uninvolvedskin of melanoma patients,” Clinical Cancer Research, vol. 13,no. 2, part I, pp. 540–549, 2007.

[95] M.-A. Perales, J. Yuan, S. Powel et al., “Phase I/II studyof GM-CSF DNA as an adjuvant for a multipeptide cancervaccine in patients with advanced melanoma,” MolecularTherapy, vol. 16, no. 12, pp. 2022–2029, 2008.

[96] J. Yuan, G. Y. Ku, H. F. Gallardo et al., “Safety and immuno-genicity of a human and mouse gp100 DNA vaccine in aphase I trial of patients with melanoma,” Cancer Immunity,vol. 9, article 5, 2009.

[97] P. L. Triozzi, W. Aldrich, K. O. Allen, R. R. Carlisle, A. F.LoBuglio, and R. M. Conry, “Phase I study of a plasmidDNA vaccine encoding MART-1 in patients with resectedmelanoma at risk for relapse,” Journal of Immunotherapy,vol. 28, no. 4, pp. 382–388, 2005.

[98] J. Weber, W. Boswell, J. Smith et al., “Phase 1 trial ofintranodal injection of a melan-A/MART-1 DNA plasmid

vaccine in patients with stage IV melanoma,” Journal ofImmunotherapy, vol. 31, no. 2, pp. 215–223, 2008.

[99] R. G. Meyer, C. M. Britten, U. Siepmann et al., “A phaseI vaccination study with tyrosinase in patients with stageII melanoma using recombinant modified vaccinia virusAnkara (MVA-hTyr),” Cancer Immunology, Immunotherapy,vol. 54, no. 5, pp. 453–467, 2005.

[100] K. R. Lindsey, L. Gritz, R. Sherry et al., “Evaluation of prime/boost regimens using recombinant poxvirus/tyrosinasevaccines for the treatment of patients with metastaticmelanoma,” Clinical Cancer Research, vol. 12, no. 8, pp. 2526–2527, 2006.

[101] J. N. Kochenderfer and R. E. Gress, “A comparison and crit-ical analysis of preclinical anticancer vaccination strategies,”Experimental Biology and Medicine, vol. 232, no. 9, pp. 1130–1141, 2007.

[102] J. Fang, Y. Lu, K. Ouyang et al., “Specific antibodies elicitedby a novel DNA vaccine targeting gastrin-releasing peptideinhibit murine melanoma growth in vivo,” Clinical andVaccine Immunology, vol. 16, no. 7, pp. 1033–1039, 2009.

[103] X. Sun, L. M. Hodge, H. P. Jones, L. Tabor, and J.W. Simecka, “Co-expression of granulocyte-macrophagecolony-stimulating factor with antigen enhances humoraland tumor immunity after DNA vaccination,” Vaccine,vol. 20, no. 9-10, pp. 1466–1474, 2002.

[104] M. C. Leslie, Y. J. Zhao, L. B. Lachman, P. Hwu, and M.Bar-Eli, “Immunization against MUC18/MCAM, a novelantigen that drives melanoma invasion and metastasis,” GeneTherapy, vol. 14, no. 4, pp. 316–323, 2007.

[105] J. C. Schmollinger and G. Dranoff, “Targeting melanomainhibitor of apoptosis protein with cancer immunotherapy,”Apoptosis, vol. 9, no. 3, pp. 309–313, 2004.

[106] A. Lladser, K. Ljungberg, H. Tufvesson et al., “IntradermalDNA electroporation induces survivin-specific CTLs, sup-presses angiogenesis and confers protection against mousemelanoma,” Cancer Immunology, Immunotherapy, vol. 59,no. 1, pp. 81–92, 2010.

[107] L. Rocha-Zavaleta, J. E. Alejandre, and A. Garcia-Carranca,“Parenteral and oral immunization with a plasmid DNAexpressing the human papillomavirus 16-L1 gene inducessystemic and mucosal antibodies and cytotoxic T lymphocyteresponses,” Journal of Medical Virology, vol. 66, no. 1, pp. 86–95, 2002.

[108] Y.-J. Wang, Y. Hou, H. Huang, G.-R. Liu, A. P. White, and S.-L. Liu, “Two oral HBx vaccines delivered by live attenuatedSalmonella: both eliciting effective anti-tumor immunity,”Cancer Letters, vol. 263, no. 1, pp. 67–76, 2008.

[109] M. Kamata, K. Denda-Nagai, N. Kubota, S. Aida, K. Takeda,and T. Irimura, “Vaccination of mice with MUC1 cDNAsuppresses the development of lung metastases,” Clinical andExperimental Metastasis, vol. 19, no. 8, pp. 689–696, 2002.

[110] B. Seliger, U. Wollscheid, F. Momburg, T. Blankenstein, andC. Huber, “Characterization of the major histocompatibilitycomplex class I deficiencies in B16 melanoma cells,” CancerResearch, vol. 61, no. 3, pp. 1095–1099, 2001.

[111] I. Peter, A. Mezzacasa, P. LeDonne, R. Dummer, andS. Hemmi, “Comparative analysis of immunocriticalmelanoma markers in the mouse melanoma cell lines B16,K1735 and S91-M3,” Melanoma Research, vol. 11, no. 1,pp. 21–30, 2001.

[112] A. L. Rakhmilevich, M. Imboden, Z. Hao et al., “Effec-tive particle-mediated vaccination against mouse melanomaby coadministration of plasmid DNA encoding gp100


and granulocyte-macrophage colony-stimulating factor,”Clinical Cancer Research, vol. 7, no. 4, pp. 952–961, 2001.

[113] M. W. J. Schreurs, A. J. De Boer, C. G. Figdor, and G. J.Adema, “Genetic vaccination against the melanocyte lineage-specific antigen gp100 induces cytotoxic T lymphocyte-mediated tumor protection,” Cancer Research, vol. 58, no. 12,pp. 2509–2514, 1998.

[114] W.-Z. Zhou, Y. Kaneda, S. K. S. Huang, R. Morishita, andD. S. B. Hoon, “Protective immunization against melanomaby gp100 DNA-HVJ-liposome vaccine,” Gene Therapy, vol. 6,no. 10, pp. 1768–1773, 1999.

[115] M. Nawrath, J. Pavlovic, R. Dummet et al., “Reducedmelanoma tumor formation in mice immunized with DNAexpressing the melanoma-specific antigen gp100/pmel17,”Leukemia, vol. 13, no. 1, pp. S48–S51, 1999.

[116] R. Weth, O. Christ, S. Stevanovic, and M. Zoller, “Genedelivery by attenuated Salmonella typhimurium: comparingthe efficacy of helper versus cytotoxic T cell priming in tumorvaccination,” Cancer Gene Therapy, vol. 8, no. 8, pp. 599–611,2001.

[117] B. Cochlovius, M. J. J. G. Stassar, M. W. Schreurs, A. Benner,and G. J. Adema, “Oral DNA vaccination: antigen uptake andpresentation by dendritic cells elicits protective immunity,”Immunology Letters, vol. 80, no. 2, pp. 89–96, 2002.

[118] W. G. Hawkins, J. S. Gold, R. Dyall et al., “Immunization withDNA coding for gp100 results in CD4+ T-cell independentantitumor immunity,” Surgery, vol. 128, no. 2, pp. 273–280,2000.

[119] J. S. Gold, C. R. Ferrone, J. A. Guevara-Patino et al., “A singleheteroclitic epitope determines cancer immunity after xeno-geneic DNA immunization against a tumor differentiationantigen,” Journal of Immunology, vol. 170, no. 10, pp. 5188–5194, 2003.

[120] M. Nawrath, J. Pavlovic, and K. Moelling, “Synergistic effectof a combined DNA and peptide vaccine against gp100 ina malignant melanoma mouse model,” Journal of MolecularMedicine, vol. 79, no. 2, pp. 133–142, 2001.

[121] L. Elzaouk, J. Pavlovic, and K. Moelling, “Analysis ofantitumor activity elicited by vaccination with combinationsof interleukin-12 DNA with gp100 DNA or the chemokineCCL21 in vivo,” Human Gene Therapy, vol. 17, no. 8, pp. 859–870, 2006.

[122] S. N. Wagner, C. Wagner, P. Luhrs et al., “Intracutaneousgenetic immunization with autologous melanoma-associatedantigen Pmel17/gp100 induces T cell-mediated tumor pro-tection in vivo,” Journal of Investigative Dermatology, vol. 115,no. 6, pp. 1082–1087, 2000.

[123] L. W. Weber, W. B. Bowne, J. D. Wolchok et al., “Tumorimmunity and autoimmunity induced by immunization withhomologous DNA,” Journal of Clinical Investigation, vol. 102,no. 6, pp. 1258–1264, 1998.

[124] W. W. Overwijk, D. S. Lee, D. R. Surman et al., “Vaccinationwith a recombinant vaccinia virus encoding a ”self” antigeninduces autoimmune vitiligo and tumor cell destruction inmice: requirement for CD4+ T lymphocytes,” Proceedingsof the National Academy of Sciences of the United States ofAmerica, vol. 96, no. 6, pp. 2982–2987, 1999.

[125] Y. M. Saenger, Y. Li, K. C. Chiou et al., “Improved tumorimmunity using anti-tyrosinase related protein-1 mono-clonal antibody combined with DNA vaccines in murinemelanoma,” Cancer Research, vol. 68, no. 23, pp. 9884–9891,2008.

[126] V. Bronte, E. Apolloni, R. Ronca et al., “Genetic vaccinationwith ’self ’ tyrosinase-related protein 2 causes melanomaeradication but not vitiligo,” Cancer Research, vol. 60, no. 2,pp. 253–258, 2000.

[127] M. Bellone, D. Cantarella, P. Castiglioni et al., “Relevanceof the tumor antigen in the validation of three vaccinationstrategies for melanoma,” Journal of Immunology, vol. 165,no. 5, pp. 2651–2656, 2000.

[128] W. G. Hawkins, J. S. Gold, N. E. Blachere et al., “XenogeneicDNA immunization in melanoma models for minimalresidual disease,” Journal of Surgical Research, vol. 102, no. 2,pp. 137–143, 2002.

[129] M. Zhang, C. Obata, H. Hisaeda et al., “A novel DNAvaccine based on ubiquitin-proteasome pathway targeting’self ’-antigens expressed in melanoma/melanocyte,” GeneTherapy, vol. 12, no. 13, pp. 1049–1057, 2005.

[130] M. E. Dudley, J. R. Wunderlich, P. F. Robbins et al.,“Cancer regression and autoimmunity in patients after clonalrepopulation with antitumor lymphocytes,” Science, vol. 298,no. 5594, pp. 850–854, 2002.


[132] N. Grinshtein, M. Ventresca, R. Margl et al., “High-dosechemotherapy augments the efficacy of recombinant ade-novirus vaccines and improves the therapeutic outcome,”Cancer Gene Therapy, vol. 16, no. 4, pp. 338–350, 2009.

[133] X. Zhu, J. Cai, J. Huang, X. Jiang, and D. Ren, “The treatmentand prevention of mouse melanoma with an oral dna vaccinecarried by attenuated salmonella typhimurium,” Journal ofImmunotherapy, vol. 33, no. 5, pp. 453–460, 2010.

[134] P. J. Bergman, J. McKnight, A. Novosad et al., “Long-termsurvival of dogs with advanced malignant melanoma afterDNA vaccination with xenogeneic human tyrosinase: a phaseI trial,” Clinical Cancer Research, vol. 9, no. 4, pp. 1284–1290,2003.

[135] J. C. Liao, P. Gregor, J. D. Wolchok et al., “Vaccination withhuman tyrosinase DNA induces antibody responses in dogswith advanced melanoma,” Cancer Immunity, vol. 6, p. 8,2006.

[136] J. Pritchard, S. J. Cotterill, S. M. Germond, J. Imeson,J. De Kraker, and D. R. Jones, “High dose melphalanin the treatment of advanced neuroblastoma: results of arandomised trial (ENSG-1) by the European NeuroblastomaStudy Group,” Pediatric Blood and Cancer, vol. 44, no. 4,pp. 348–357, 2005.

[137] S. L. Volchenboum and S. L. Cohn, “Progress in defining andtreating high-risk neuroblastoma: lessons from the benchand bedside,” Journal of Clinical Oncology, vol. 27, no. 7,pp. 1003–1004, 2009.

[138] A. L. Yu, A. L. Gilman et al., “Anti-GD2 antibody with GM-CSF, IL2 and isotretinoin for neuroblastoma: a children’soncology group (COG) phase III study,” New England Journalof Medicine, vol. 363, no. 14, 2010.

[139] H. N. Lode, R. Xiang, N. M. Varki, C. S. Dolman, S. D.Gillies, and R. A. Reisfeld, “Targeted interleukin-2 therapyfor spontaneous neuroblastoma metastases to bone marrow,”Journal of the National Cancer Institute, vol. 89, no. 21,pp. 1586–1594, 1997.

[140] S. Fest, N. Huebener, S. Weixler et al., “Characterizationof GD2 peptide mimotope DNA vaccines effective againstspontaneous neuroblastoma metastases,” Cancer Research,vol. 66, no. 21, pp. 10567–10575, 2006.


[141] A. Kowalczyk, A. Wierzbicki, M. Gil et al., “Induction ofprotective immune responses against NXS2 neuroblastomachallenge in mice by immunotherapy with GD2 mimotopevaccine and IL-15 and IL-21 gene delivery,” Cancer Immunol-ogy, Immunotherapy, vol. 56, no. 9, pp. 1443–1458, 2007.

[142] A. F. Carpentier, M. R. Rosenfeld, J.-Y. Delattre, R. G.Whalen, J. B. Posner, and J. Dalmau, “DNA vaccination withHuD inhibits growth of a neuroblastoma in mice,” ClinicalCancer Research, vol. 4, no. 11, pp. 2819–2824, 1998.

[143] H. N. Lode, U. Pertl, R. Xiang, G. Gaedicke, and R.A. Reisfeld, “Tyrosine hydroxylase-based DNA-vaccinationis effective against murine neuroblastoma,” Medical andPediatric Oncology, vol. 35, no. 6, pp. 641–646, 2000.

[144] N. Huebener, S. Fest, A. Strandsby et al., “A rationallydesigned tyrosine hydroxylase DNA vaccine induces specificantineuroblastoma immunity,” Molecular Cancer Therapeu-tics, vol. 7, no. 7, pp. 2241–2251, 2008.

[145] U. Pertl, H. Wodrich, J. M. Ruehlmann, S. D. Gillies,H. N. Lode, and R. A. Reisfeld, “Immunotherapy with aposttranscriptionally modified DNA vaccine induces com-plete protection against metastatic neuroblastoma,” Blood,vol. 101, no. 2, pp. 649–654, 2003.

[146] N. Huebener, B. Lange, C. Lemmel et al., “Vaccination withminigenes encoding for novel ’self ’ antigens are effectivein DNA-vaccination against neuroblastoma,” Cancer Letters,vol. 197, no. 1-2, pp. 211–217, 2003.

[147] S. Fest, N. Huebener, M. Bleeke et al., “Survivin minigeneDNA vaccination is effective against neuroblastoma,” Inter-national Journal of Cancer, vol. 125, no. 1, pp. 104–114, 2009.

[148] A. Jemal, A. Thomas, T. Murray, and M. Thun, “Cancerstatistics, 2002,” Cancer Journal for Clinicians, vol. 52, no. 1,pp. 23–47, 2002.

[149] E. D. Crawford, M. Rosenblum, A. M. Ziada, and P. H. Lange,“Overview: hormone refractory prostate cancer,” Urology,vol. 54, no. 6A, supplement, pp. 1–7, 1999.

[150] G. G. Hillman, J. A. Triest, M. L. Cher, S. V. Kocheril, andB. R. Talati, “Prospects of immunotherapy for the treatmentof prostate carcinoma—a review,” Cancer Detection andPrevention, vol. 23, no. 4, pp. 333–342, 1999.

[151] I. F. Tannock, D. Osoba, M. R. Stockler et al., “Chemotherapywith mitoxantrone plus prednisone or prednisone alone forsymptomatic hormone-resistant prostate cancer: a Canadianrandomized trial with palliative end points,” Journal ofClinical Oncology, vol. 14, no. 6, pp. 1756–1764, 1996.

[152] M. Noguchi, K. Kobayashi, N. Suetsugu et al., “Induction ofcellular and humoral immune responses to tumor cells andpeptides in HLA-A24 positive hormone-refractory prostatecancer patients by peptide vaccination,” Prostate, vol. 57,no. 1, pp. 80–92, 2003.

[153] M. D. L. L. Garcia-Hernandez, A. Gray, B. Hubby, and W. M.Kast, “In vivo effects of vaccination with six-transmembraneepithelial antigen of the prostate: a candidate antigen fortreating prostate cancer,” Cancer Research, vol. 67, no. 3,pp. 1344–1351, 2007.

[154] S. Ross, S. D. Spencer, I. Holcomb et al., “Prostate stemcell antigen as therapy target: tissue expression and in vivoefficacy of an immunoconjugate,” Cancer Research, vol. 62,no. 9, pp. 2546–2553, 2002.

[155] Z. Y. Zhu, C. P. Zhong, W. F. Xu et al., “PSMA mimotopeisolated from phage displayed peptide library can inducePSMA specific immune response,” Cell Research, vol. 9, no. 4,pp. 271–280, 1999.

[156] Y. Wang, M. Harada, H. Yano et al., “Prostatic acid phos-phatase as a target molecule in specific immunotherapyfor patients with nonprostate adenocarcinoma,” Journal ofImmunotherapy, vol. 28, no. 6, pp. 535–541, 2005.

[157] K. Todorova, I. Ignatova, S. Tchakarov et al., “Humoralimmune response in prostate cancer patients after immu-nization with gene-based vaccines that encode for a proteinthat is proteasomally degraded,” Cancer Immunity, vol. 5,pp. 1–8, 2005.

[158] J. D. Wolchok, P. D. Gregor, L. T. Nordquist, S. F. Slovin, andH. I. Scher, “DNA vaccines: an active immunization strategyfor prostate cancer,” Seminars in Oncology, vol. 30, no. 5,pp. 659–666, 2003.

[159] M. Mincheff, S. Zoubak, and Y. Makogonenko, “Immuneresponses against PSMA after gene-based vaccination forimmunotherapy—a: results from immunizations in ani-mals,” Cancer Gene Therapy, vol. 13, no. 4, pp. 436–444, 2006.

[160] A.-K. Roos, M. Pavlenko, J. Charo, L. Egevad, and P. Pisa,“Induction of PSA-specific CTLs and anti-tumor immunityby a genetic prostate cancer vaccine,” Prostate, vol. 62, no. 3,pp. 217–223, 2005.

[161] D. J. Marshall, L. R. San Mateo, K. A. Rudnick et al., “Induc-tion of Th1-type immunity and tumor protection with aprostate-specific antigen DNA vaccine,” Cancer Immunology,Immunotherapy, vol. 54, no. 11, pp. 1082–1094, 2005.

[162] H. Qin, C. Zhou, D. Wang et al., “Specific antitumor immuneresponse induced by a novel DNA vaccine composed ofmultiple CTL and T helper cell epitopes of prostate cancerassociated antigens,” Immunology Letters, vol. 99, no. 1,pp. 85–93, 2005.

[163] L. E. Johnson, T. P. Frye, N. Chinnasamy, D. Chinnasamy,and D. G. McNeel, “Plasmid DNA vaccine encoding prostaticacid phosphatase is effective in eliciting autologous antigen-specific CD8+ T cells,” Cancer Immunology, Immunotherapy,vol. 56, no. 6, pp. 885–895, 2007.

[164] L. E. Johnson, T. P. Frye, A. R. Arnot et al., “Safety andimmunological efficacy of a prostate cancer plasmid DNAvaccine encoding prostatic acid phosphatase (PAP),” Vaccine,vol. 24, no. 3, pp. 293–303, 2006.

[165] D. G. McNeel, E. J. Dunphy, J. G. Davies et al., “Safety andimmunological efficacy of a DNA vaccine encoding prostaticacid phosphatase in patients with stage D0 prostate cancer,”Journal of Clinical Oncology, vol. 27, no. 25, pp. 4047–4054,2009.

[166] X. Zhang, C. Yu, J. Zhao et al., “Vaccination with aDNA vaccine based on human PSCA and HSP70 adjuvantenhances the antigen-specific CD8+ T-cell response andinhibits the PSCA+ tumors growth in mice,” Journal of GeneMedicine, vol. 9, no. 8, pp. 715–726, 2007.

[167] J. R. Gingrich, R. J. Barrios, R. A. Morton et al., “Metastaticprostate cancer in a transgenic mouse,” Cancer Research,vol. 56, no. 18, pp. 4096–4102, 1996.

[168] N. M. Greenberg, F. DeMayo, M. J. Finegold et al., “Prostatecancer in a transgenic mouse,” Proceedings of the NationalAcademy of Sciences of the United States of America, vol. 92,no. 8, pp. 3439–3443, 1995.

[169] M. D. L. L. Garcia-Hernandez, A. Gray, B. Hubby, O.J. Klinger, and W. M. Kast, “Prostate stem cell antigenvaccination induces a long-term protective immune responseagainst prostate cancer in the absence of autoimmunity,”Cancer Research, vol. 68, no. 3, pp. 861–869, 2008.

[170] S. Ahmad, G. Casey, P. Sweeney, M. Tangney, and G. C.O’Sullivan, “Prostate stem cell antigen DNA vaccination


breaks tolerance to self-antigen and inhibits prostate cancergrowth,” Molecular Therapy, vol. 17, no. 6, pp. 1101–1108,2009.

[171] A. Gray, M. de la Luz Garcia-Hernandez, M. van West,S. Kanodia, B. Hubby, and W. M. Kast, “Prostate can-cer immunotherapy yields superior long-term survival inTRAMP mice when administered at an early stage of car-cinogenesis prior to the establishment of tumor-associatedimmunosuppression at later stages,” Vaccine, vol. 27, supple-ment 6, pp. G52–G59, 2009.


Review Article

Gene Carriers and Transfection Systems Used inthe Recombination of Dendritic Cells for EffectiveCancer Immunotherapy

Yu-Zhe Chen,1 Xing-Lei Yao,1 Yasuhiko Tabata,2 Shinsaku Nakagawa,3 and Jian-Qing Gao1

1 Institute of Pharmaceutics, College of Pharmaceutical Sciences, Zhejiang University, Hangzhou, Zhejiang 310058, China2 Department of Biomaterials, Field of Tissue Engineering, Institute for Frontier Medical Sciences, Kyoto University, 53 Kawanara-cho,Shogoin, Sakyo-ku, Kyoto 606-8507, Japan

3 Department of Biotechnology and Therapeutics, Graduate School of Pharmaceutical Sciences, Osaka University, 1-6 Yamadaoka,Suita, Osaka 565-0871, Japan

Correspondence should be addressed to Jian-Qing Gao, [email protected]

Received 1 July 2010; Accepted 28 October 2010

Academic Editor: Robert E. Cone

Copyright © 2010 Yu-Zhe Chen et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Dendritic cells (DCs) are the most potent antigen-presenting cells. They play a vital role in the initiation of immune responseby presenting antigens to T cells and followed by induction of T-cell response. Reported research in animal studies indicatedthat vaccine immunity could be a promising alternative therapy for cancer patients. However, broad clinical utility has not beenachieved yet, owing to the low transfection efficiency of DCs. Therefore, it is essential to improve the transfection efficiency ofDC-based vaccination in immunotherapy. In several studies, DCs were genetically engineered by tumor-associated antigens orby immune molecules such as costimulatory molecules, cytokines, and chemokines. Encouraging results have been achieved incancer treatment using various animal models. This paper describes the recent progress in gene delivery systems including viralvectors and nonviral carriers for DC-based genetically engineered vaccines. The reverse and three-dimensional transfection systemsdeveloped in DCs are also discussed.

1. Introduction

Cancer is a leading cause of death worldwide. Althoughprogress has been made in cancer therapy with conventionaltreatment modalities, such as surgery, chemotherapy, andradiotherapy over the last several decades [1], the totalnumber of cancer-related deaths is still increasing. Therefore,there is an urgent requirement to develop novel therapies forthe treatment of cancer. With the rapid developments in thefields of immunology and cancer biology, immunotherapyis expected to play a key role in next-generation cancertreatment. The goal of immunotherapy is to promote thepatient’s own immune system to kill cancer cells instead ofusing external helpers, that is, surgery or medicine. To inducea specific immune response against cancers, researchershave designed a variety of antitumor vaccines based on themolecular identities of tumor-associated antigens (TAAs).

Recent findings from this line of research suggest thatimmunotherapy strategies are feasible and promising [2–4].

DCs are professional and the most potent antigen-presenting cells (APCs) of T-cell special responses, whichplay an important role in initiating and regulating adaptiveimmune responses [5]. The major function of DCs inimmune system is capturing exogenous and endogenousantigens when infection or cancer occurs, and then pre-senting the antigens to T cells via major histocompatibilitycomplex (MHC) molecules [6]. Moreover, DCs are alsoinvolved in regulating immune tolerance and clonal selection[7, 8]. In 1990, it was firstly reported that injection ofDCs with protein antigens ex vivo could prime antigen-specific response in animal model [9]. After that, severalstudies demonstrated that DCs pulsed with TAAs couldproduce significant therapeutic immunity to tumors withlow toxicity. Because DCs could manipulate the immune


system by enhancing specific responses to infectious diseasesand cancer, DCs networking system became an attractiveapproach in cancer therapy [10, 11].

The results of early studies in animal models and somepreclinical trials indicated that TAA-presenting DC mightbe a promising treatment for cancer. However, it is difficultto induce long-term tumor-specific immune response inhumans. This may be due to the fact that most TAAs are self-antigens, which make cancer cells bypass normal immuneprotective mechanisms. Therefore, in order to overcometolerance against self-antigens, it is necessary for an efficientvaccine to induce autoimmune responses [12]. Additionally,the suppressive mechanisms in tumor microenvironmentcan also inhibit immune response to malignant cells [13].Hence, designing and developing an efficient and long-termDC vaccine, which could specifically target cancer cells,is urgently needed. Subsequent studies have shown thatvaccination using DCs in vitro transferred with transgeneencoding TAAs or immunomodulatory proteins are moreefficient than using cells directly pulsed with protein anti-gens, tumor peptides, lysates, or RNA [14].

This paper focuses on the recent findings in DC vaccina-tions genetically engineered by recombination biotechnologyvia different vectors and overviews the development of genedelivery systems for DCs.

2. Biological Characteristics ofDCs and the Process of DC-MediatedImmune Response

The DCs are generated from CD34+ bone marrow stemcells and from DC precursors in the peripheral blood.The concentration of DCs in normal tissue and blood isvery low, which makes it difficult to isolate DCs directlyfrom peripheral blood and bone marrow. Currently, theprevalent procedure is to differentiate the monocytes fromperipheral blood and bone marrow to DCs with the help ofleukapheresis technology and stimulation by cytokines [15].

According to biological properties of DCs, they couldbe divided into three major groups: plasmacytoid DCs(pDCs), inflammatory DCs (iDCs), and conventional DCs(cDCs) [16, 17]. cDCs are also named myeloid DC (mDCs)owing to their typical form and function [18]. They can befurther divided into lymphoid-tissue-resident DCs (splenic,thymic DCs, etc.) and migratory DCs (Langerhans cells,dermal DCs, etc.) [19]. Unlike migratory DCs, whichmigrate through the lymph, lymphoid-tissue-resident DCsare located mostly at lymphoid tissues to collect and presentantigens [20]. Both of the mDCs can be further classifiedbased on the levels of phenotype protein expression andfunction. For example, CD8+ and CD4+ mDCs, which werefound to preferentially express MHC I and II, respectively,induce different types of T-cell responses [21, 22].

On the other hand, on the basis of the differentphenotype and surface antigens, DCs could also be dividedinto immature and mature DCs. The term “immature” refersto DCs with the phenotypic features of low expression ofMHC II and molecules such as CD86. In contrast, mature

DCs are characterized by high expression of MHC II andT-cell costimulatory molecules such as CD40, CD80, CD83,and CD86 [23–25]. Under pathologic conditions, DCs arestimulated by microbes, products of damaged tissues, cellsof the innate or adaptive immune system, and inflamma-tory cytokines. These endogenous and exogenous antigensare taken up by DCs through the specialized endocyticsystem, which is mediated by a variety of receptors suchas Toll-like receptors (TLRs) [4, 26], nucleotide-binding,and oligomerization domain proteins (NODs) [27, 28].Then DCs undergo a complex process of activation makingimmature antigen-capturing DCs change into APCs. Theprocess is characterized by extended dendrites of DC’sexternal form. As a result, their cellular motility to migrateto the draining lymph node is increased [29]. Meanwhile,their surface costimulatory molecules such as CD40, CD80,and CD86 are upregulated [30–32], and MHC moleculesare expressed on the surface of cells [33]. One of themajor functions of DCs in immune system is capturingand presenting the antigen to T cells via MHC molecules.When the antigen is presented by APCs through MHC I,which interacts with CD8+ T cells, the activated T cells coulddifferentiate into cytotoxic T lymphocytes (CTLs). Activationof CD4+ T cells occurs with the help of MHC II. Afteractivation, the CD4+ T cells differentiate into T-helper 1(Th1) and T-helper 2 (Th2) cells, which are involved ininducing macrophages and B-cells responses [34, 35]. CTLsare the major killers of tumor cells; they accomplish thekilling with the help of CD4+ T cells, which can inducepotential long-term CD8+ T-cell responses by producingvarious cytokines [36, 37].

As it is well known that intracellular endogenous antigensand exogenous antigens are presented in MHC I andMHC II by DCs, respectively [38], the strategy for DC-based vaccines in cancer immunotherapy is to make DCscross-presentation. This can present MHC I to CTLs usinginternalized antigens generated from exogenous sources.Through this process, TAAs can be presented by the DC toboth CD4+ and CD8+ T cells (in MHC I), and a broad andstrong immune response against tumor could be induced[16, 39].

3. Methods Used in Genetically Engineered DCs

3.1. Modified/Pulsed Methods. It is generally believed that,the direct presentation of TAAs to CD8+ CTLs (e.g., directadministration of tumor peptides) is most tolerogenic [40,41]. Several evidences showed that the differentiation andmaturation of DCs were suppressed by cytokines existing inthe tumor microenvironment. Since DCs play a crucial rolein inducing antigen-specific T-cell responses, it is importantto deliver tumor antigens with CD8+ and CD4+ T-cellepitopes to DCs. There are several strategies to induce DCsto present exogenous antigens on MHC I molecules [42].

According to the mechanism of MHC-mediated antigenpresentation, early researchers tried to pulse DCs directlywith tumor-specific peptides. Synthetic MHC I-bindingpeptides have been used in DC-based vaccination. The


TAAs, which were derived from MHC I-binding peptides,including melanoma-related antigens [43, 44], carcinoem-bryonic antigen (CEA) [45, 46], folate binding protein (FBP)[47, 48], prostate-specific membrane antigen (PMSA) [49,50], and Mucin 1 (MUC-1) [27, 51, 52], were firstly usedto modify DCs. This strategy was easy to perform and wasshown to be successful in some animal studies and clinicaltrials. Further development in using Tat peptide, which isfrom transduction domain of Human ImmunodeficiencyVirus (HIV), makes this strategy more effective [46, 53].However, as this approach was mainly based on specificMHC I-restricted peptides, the importance of MHC II-restricted T-helper (Th) cells in mediating response (CD4+

T) and accelerating immune responses [54] was not fullyconsidered.

Because of the limitations of using one single TAApeptide to modify DCs, researchers began to design vac-cination utilizing DCs pulsed with whole tumor lysates.In this approach, as the tumor cell preparations containlots of relevant antigens, broad-spectrum TAAs includingunknown ones were presented to T cells and this inducedan immune response. The advantages of this method include(1) presenting multiple peptides and epitopes to T cells toinduce CTL response; (2) generating both CD8+ responseand CD4+ helper T-cell response, which help to inducemacrophages and B-cell response as well as prolonging theCTLs response; (3) reducing the workload of discovering andpreparing appropriate peptides and epitopes required to bepresented on DCs and then to be identified by T cells; (4)probably generating tumor lysate-specific memory T cells[55]. The results in animal models and in clinic trials usingtotal tumor lysate approaches have been demonstrated to behighly effective and have low toxicity in a variety of cancers[56, 57]. It was further observed that DCs pulsed withapoptotic tumor cell preparation showed a more pronouncedeffect in activating T cells [58]. The main limitation of thismethod is the limited amount of patient tumor cells wecould collect, making the preparation work difficult. Anotherproblem is that the presence of many irrelevant antigens intumor cell preparations could cause autoimmune responses.

Since the early reports on transfection with RNA showedstrong immune response against tumor [59, 60], mRNA-pulsed DCs have become a research hotspot these days.Because RNA-based vaccines have many advantages includ-ing easily preparation, low price, specificity, control, and norisk of incorporation into the host genome [61]. DCs weretransfected with tumor mRNA encoding TAA or epitopes byusing carriers [61, 62] or electroporation [63, 64]. This canalso induce tumor-specific immunity in vitro. By transfectionwith RNA, DCs can express specific or total antigen ontheir surface and finally present them to T cells [65]. Inaddition, using mRNA could be a promising therapy forthe patients who have only few available tumor cells formRNA preparation, because mRNA could be produced inlarge amounts through noninvasive biopsy procedures [66].Recent studies indicated that mRNA has been used notonly as a source of antigen, but also as a way to stimulateDC to produce immunostimulatory molecules [67]. Themain limitations of using mRNA, however are difficulty in

manipulation, having lower transfer efficiency and shorterlifespan (degradated by RNases rapidly) [68].

3.2. Modified/Pulsed Methods. To enhance DCs antitumorefficiency, the delivery of DNA encoding TAAs, immunos-timulatory molecules, cytokines, chemokines, and otherstimuli has been developed in the recent years. Comparedwith tumor antigen loading strategies described previously,genetical engineering of DCs has some special advantages,which include: (1) bypassing the work of understandingthe complex intracellular process of MHC-mediated pre-sentation; (2) achieving the purpose of cross presentationto induce a robust immune response; (3) showing a long-term antigen expression; (4) significantly reducing theautoimmune response; (5) easier preparation; (6) stabilityin transduction process. However, there are also somelimitations in DNA-based DC vaccination. Immunother-apy using DCs transferred by viral vectors may inducean autoimmune response and mutation. Nonviral transfershows very low transfer efficiency. DNA strategy also hasits intrinsic problems such as persisting expression andgenome incorporation risk. These problems are needed to beconsidered and solved before it becomes a better application.

To date, various vehicles and methods have been devel-oped for gene transfer of DCs. Vehicles can be dividedinto viral vectors and nonviral carriers. And there are alsosome other transfection methods such as ultrasound andelectroporation. By using different carries and transfectionmethods, the transduction efficiency and the preclinic trialresults are different [69, 70]. After long-term experimentsin vitro and in vivo, an increasing number of scholarsthink that the transduction of DCs using vehicles showsmore advantages than using naked DNA alone. The majoradvantage of using viral vectors is their high efficiency inthe transfection, which induced high protein expressionlevels. The limitations of this method are immune andmutation risk. For example, the host cells might expressthe viral proteins, and thereby might induce immunologicinterference. In contrast, the nonviral strategy, which isregarded as a safer alternative to virus-mediated transduc-tion, is considered to be promising treatment in clinic.Nonviral carriers could overcome the problems caused byviral vectors. In addition, nonviral carriers were more stableand controllable in preparation and application. The riskof virus-associated recombination mutation in host genomecould be avoided. It makes DNA vaccines feasible in clinicalapplication. But the major limitations of nonviral carriersare low efficiency in transfection and low levels of proteinexpression. It needs to be optimized by the carriers modifiedsystem. The mechanisms of transfection using viral vectorsand nonviral carriers are shown below (Figures 1 and 2).

3.2.1. DCs Transferred by Viral Vectors. In contrast withdirect viral vaccine, DCs transferred by viral vectors in vitrowould reduce the production of certain type of antibodies,which may cause side effects and would finally reduce theeffectiveness of cancer therapy in clinic [71]. The viruses


Endocytosis

Liposome

Purpose DNA

Polymer/DNAcomplex

Endocytosis

Endosome

DNA escape

Nuclear trafficking

Translating

Nucleus

Host genome

Figure 1: The mechanism of transfection using nonviral carriers.

employed as gene vectors include adenoviruses (Ad), adeno-associated viruses (AAV), retrovirus, lentivirus, and otherviruses. These viruses are all deleted critical genes neededfor their reproduction. And then, they are inserted withpurpose genes such as genes encoding TAAs. Recombinantviral vectors might be the most attractive gene transduc-tion vehicles because of their high transfection efficiency,although they are more immunogenic than nonviral carriersin clinical applications [72]. To minimize the risk of specificimmunity and to boost the clinical antitumor response,several improvements in viral strategies have been developed,such as replacing the genes required for viral replicationwith the helper plasmids and modifying the genome ofviral capsid [73]. However, to date, there is not a perfecttreatment for tumor by using viral vectors-transferred DCs,as the application of viruses in vivo could destroy the antigen-presenting function of DCs.

Adenovirus. Adenovirus (Ad), which is nonenveloped andmedium-sized (90–100 nm) viruses, has a double-strandedlinear DNA genome. The adenoviral genome comprises fourearly (E1, E2, E3, and E4), four intermediate, and one latetranscriptional units. To use Ad as a gene-transferring vector,the E1 gene, which contributes to reproduction, must bedeleted. Ad vectors (AdVs) are based on substitution ofthe E1 region by the therapeutic gene. AdVs are widelyused for basic and clinical research because of their hightransduction efficiency [74–76]. The primary and secondarybinding receptors of AdVs, Coxsackie adenovirus receptor(CAR), and V-integrin play important roles in mediatingthe uptake of immature DCs. AdVs can consistently inducepotent presentation of both MHC class I and class II-restricted epitopes. Several studies demonstrated that usingAdVs could be regarded as a valuable gene delivery systemeven in clinical application, for example, serotype 5 (rAd5)[77–80]. Interactions between AdVs and DCs were alsoinvestigated recently. These include virus-mediated DCmaturation, antigen processing machinery (APM) regulation

CD4 receptor

Lentivirus

Purpose DNA

Adenovirus

Endosome

Nucleus

Host genome

Purpose RNA

DNA

CAR

Escape

Nuclear trafficking

Integrating

TranslationReverse transcript

complex

usssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssg

Figure 2: The mechanism of transfection using viral vectors.

and T-cell activation. It was observed that the phenotypeand cytokine profile of DC transduced with Ads changed[81, 82], some selected modification of DCs by Ads arelisted in Table 1. These results provide the evidence for thedesigning human cancer vaccines.

Retrovirus and Lentivirus. Retrovirus is a single-stranded(ss) RNA virus, which is replicated from RNA to DNA bythe revertase. Then the produced DNA is integrated into thehost’s genome by an integrase enzyme. With the replicationof host genome, the viruses reproduce as part of the host’sDNA. Retroviruses also attract the DC researchers theseyears for their transduction capacity of bone marrow-derivedDCs (BMDCs) and cord blood-derived DCs to keep theirdifferentiation [93, 94]. The advantages of retroviral vectorsused as transduction vehicles include the stable expressionof full-length proteins and less immunologic responsesagainst viral antigens because their structural proteins arenot expressed [68]. However, most retroviruses are difficultto transfect nondividing cells such as mature DCs. Thisdisadvantage limits the application of retroviruses in clinic.

Lentiviruses derived from HIVs belong to retrovirusesfamily. It is easier for the lentivirus to infect nondividing cellscompared to other retroviruses because of its unique routeof viral transfection by expressing both integrase [95] andVpx proteins which interact with components of the nuclearpore complex [96, 97]. Another advantage of lentivirus isthe low prevalence of HIV infections, which lead relativelyrare pre-existing immune conditions [98]. Thus, recombinedLentivirus vectors (LVs) could be designed as efficient vectorsfor transduction of both mature and immature DCs. Recentstudies have demonstrated that, in comparison with lipofec-tion, electroporation and AAV, LV is the most effective vectorfor transduction of BMDCs [99]. Many studies reported thatthe highly efficient transduction of DCs in vivo [100–102]and ex vivo [103] is possible by using LVs, this can be alsoused for shRNA transduction [104, 105]. Moreover, LVs canalso be used to transfect monocytes before differentiating


Table 1: Overview of recent studies using Ad vectors for gene transfer in DCs.

Cancer Transfer molecule In vitro In vivo in animal model Reference

ProstatetPSMA and 4-1BBL High IFN-production Strong antitumor immunity [83]

STEAP High IFN-productionInhibition of tumor growth, vaccinationdelaying the growth of pre-establishedtumors

[84]

Hepatoma

mTERT High IFN-and IL-2production

Inhibition of the tumor growth [85]

hTERT Inducing strong CTLresponse

Inducing anti-tumour immunity [63]

HCC and CD40L Increasing DCs IL-12Inducing complete regression of establishedtumors and long-term immunity againsttumor recurrence

[86]

AFP and HBsAg Inducing CTLs killingHepG2.2.15 cell lines

Inhibition of tumor growth inimmunodeficiency mice

[87]

LeukemiaSurvivin and GM-CSF

Much higher activity ofCTL than DCs witheither

No data available [88]

IL-12 with tumor cell lysate No data available Prolonged survival time [89]

Metastatic lung cancer IL-12- and 4-1BBL High IFN-productionand CTLs response

Greater antitumor and antimetastatic effectsthan either treatment alone highermigratory abilities of DCs

[90]

Lung livin Inducing CTLs lysingLLC

Inducing a potent protective andtherapeutic antitumor immunity

[91]

Urologic cancer cells SurvivinInducing CTLs againstvarious bladder, kidney,and prostate cancer cells

No data available [92]

into DCs, which could bypass the preactivation agents suchas granulocyte-macrophage colony-stimulating actor (GM-CSF) and interleukin (IL)-4 [106]. Further modified workhas been performed these years and may be extended toclinical trials. However, gene therapy using LVs has a possiblerisk of insertional mutagenesis.

3.2.2. DCs Transferred by Nonviral Carriers. Although viralvectors have been demonstrated to be more efficient in genedelivery, the clinical application is limited due to their riskin safety and unexpected adverse effects [107]. In contrast,nonviral vectors, such as various liposomes and polyioncomplexes, have been increasingly developed these yearsbecause of their low immune response and ease of synthesisunder controllable conditions and ease to be modified. Themajor limitations are their inefficient transfer, low geneexpression and relatively high cytotoxicity by nature. Here isthe advanced development in several nonviral vectors.

Liposomes. Liposomes are artificial closed vesicles of lipidbilayer membranes. Liposomes, modified with specific tar-geting molecular structures on surface, can be used astransfection vectors for DCs. After the APCs interact withthe targeting liposomes which contain antigen peptidesor DNA, the APC-mediated CTL responses are effectivelyenhanced [108]. Different formulations of liposomes aredesigned to improve the uptake by DCs through dif-ferent receptor-mediated routes. These formulations of

liposomes include liposomes prepared with mannosylatedphosphatidylethanolamine (Man-PE), trimethyl ammoniumpropane [2], and phosphatidylserine [109] corresponding tomannose receptor (MR), negatively charged surface proteinsand PS receptor of DCs, respectively [110].

Over the last decade, several studies have demonstratedthat MR-mediated gene transfer into macrophages and DCsusing mannosylated cationic liposomes can elicit effectiveimmune responses [111–113]. MR is a typical receptor of C-type lectins, which are structurally related to surface-boundnonspecific pattern recognition receptors on the surface ofmonocytes, macrophages, and DCs. Using the affinity ofMR with mannose-containing ligands, researchers preparedseveral mannosylated cationic liposomes to encapsulateDNA or RNA for gene delivery purpose. It was also reportedthat the transfection efficiency of macrophages and DCs wasenhanced by a combination method using mannosylatedlipoplexes and bubble liposomes (BLs) with ultrasoundexposure. In the liver and spleen, the transfection efficiencyby using this combination method was higher than that ofnaked pDNA or combination of unmodified lipoplexes andBLs [114]. Moreover, besides DNA delivery, siRNA silencingof DCs with liposomes is also widely used [115, 116].

Fusogenic liposome (FL) encapsulating DNA is a novelbiological strategy to deliver antigen gene directly intothe cytoplasm of DCs via membrane fusion. It has beendemonstrated that FL-mediated OVA-gene delivery caninduce potent presentation of antigen via the MHC classI-dependent pathway in vitro and then can induce a series


of immune responses [117]. Complexes of lipoplexes withpH-sensitive fusogenic liposomes can not only transfectvarious malignant cells, but also can transfect a murineDC line (DC2.4) [118]. These complexes exhibited highertransfection efficiency to DC2.4 cells than some othercommercially available reagents. So these new complexesmay be valuable for the transfection of DCs.

Complex Particles. Recently, complex particles such ascationic polymers have been used as promising vectors forDNA delivery, because of their electrostatic interactions andease of modification in targeting ligands. The advantagesof cationic polymers used as gene carriers include (1)compression of the DNA into complex particles with smallsize and high density, which makes gene easier to transferinto cells; (2) electrostatic attraction with the cell membraneto facilitate endocytosis; (3) stability under the electrostaticrepulsion. Nowadays, chitosan and biodegradable micropar-ticles such as poly (ethylene-imine) (PEI), and so forth,attracted considerable attention in this field.

Chitosan has been used as gene delivery carrier becauseof good biocompatibility and high positive charge densityin recent years [119]. The transfection efficiency of chitosandepends on its molecular weight, DNA complexes chargeratio, pH, and particle sizes as well as the type of cells[120, 121]. To overcome the weakness of chitosan such aspoor solubility, low rate of DNA release and low efficiency oftransfection, hydrophilic, and hydrophobic structure modifi-cation have been carried out. Although unmodified chitosanmay not be a good gene delivery carrier for DCs becauseof its low transfection efficiency, some modified chitosanshowed better behavior in delivering genes into DCs. Forexample, to enhance the IL-12 gene delivery to DCs invivo, mannosylated chitosan (MC), which is used to inducemannose receptor-mediated endocytosis, was prepared toencapsulate IL-12 gene into DCs. MC not only has goodphysicochemical properties and low cytotoxicity, but alsoshows much more enhanced transfection efficiency to DCsrather than unmodified chitosan in vitro. And tumor growthin mouse model was suppressed by intratumoral injection ofMC/plasmid encoding murine IL-12 [122, 123]. Moreover,Zhou et al. [124] developed MC microspheres containingPEI/DNA complexes, and used this carrier to improve thedelivery of DNA into DCs. After in vivo immunization,the microspheres induced significantly enhanced serumantibody and cytotoxic T-lymphocyte (CTL) responses.Therefore, MC-mediated cytokine gene delivery system onDCs may be a potential approach for cancer immunotherapy.

Biodegradable microparticles which are easily clearedby physiological clearance systems can avoid the possi-ble cytotoxicity caused by accumulation in cells and tis-sues. Microparticles prepared from poly (lactide) [125],poly(lactide-coglycolide) (PLGA), poly (orthoesters) (POE),and other polymers microparticles have been well studiedin recent years. Their biodegradability, biocompatibility,and low toxicity properties make them suitable carriers forDNA vaccines. The virus-associated risk of adverse effectscan also be avoided because these microparticles do not

incorporate into the host cell’s nucleus. Several studieshave demonstrated that immune responses are inducedby particle-DNA vaccines. Recently, it was reported thatPLGA/PEI-DNA complex nanospheres have been developedas an efficient delivery system for the DCs. And the efficiencycan be significantly promoted by modifying with nuclearlocalization signal (NLS) [126]. Also, such material as POEwith lower cytoxicity was used to encapsulate plasmids and itinduced both cellular and humoral responses in vivo [127].The internalization of the particles into DCs is throughphagocytosis, and the microparticles are easily phagocytosedby DC in vitro or in vivo [128].

It is well known that PEI is the most effective nonviralcarrier for gene delivery. It has relative high transfectioncapacity due to its characteristics such as its ease in com-bining with DNA, binding with the cell and escaping fromthe endosome. Nowadays, the hotspots of research graduallyfocused on reducing its toxicity by various modifications.It is mainly because, cytotoxicity increases as the molecularweight increases, while the efficiency of gene loading andtransfection increases correspondingly. Recently, Ali andMooney [129] demonstrated that it showed sustained andlong-term presentation when DCs were transfected withthe PEI condensed with gene encoding GM-CSF. And theyalso use polymer PLG as scaffold fabrication continuouslystimulated DCs with both GM-CSF and PEI-DNA. Thisprocess led to a 20-fold increase in gene expression thanno scaffold groups, and 10 days expression in vitro. Theseresults largely encouraged the development of biomaterials,such as PEI, coordinated with other macroporous scaffoldsas a transfer system for DC-based vaccination. Besides, PEI-based nanoparticles could also be used to encapsulate siRNAto transfer DCs against tumor cells [130].

Transfection Systems for Nonviral Carriers Used in DCsTransfer. Although nonviral gene delivery system has manyadvantages as reported, it still cannot reach the hightransfection efficiency as viral vectors do, and the periodof transfection is also far from satisfactory. To improvethe efficiency, one of the methods is optimizing the wholetransfection system. For this purpose, the reverse and three-dimensional (3D) transfection systems have been proposed.

Some earlier studies demonstrated that reverse transfec-tion method was more effective in enhancing the level andduration of gene expression than that of the conventionalmethod on some cell lines [133, 134]. This may be dueto the fact that DNA complexes can more easily transfectcells if they are in the area nearer to cells. In addition, cellstend to adhere to the surface and bottom of culture dish.Thus, attaching the DNA-complexes to the bottom of culturedishes before adding cells could enhance and prolong geneexpression. Moreover, continuous interaction between DNA-complexes and cells would minimize the influence of serumin the transfection activity of DNA-complexes.

Several physical, chemical, and biochemical factors, caninfluence gene transfection efficiency. Increasing studieshave reported that, when cells are cultured in 3D sys-tems, the results of transfection in various cells such as


Table 2: Genes used for modification of DCs.

Groups Genes coding factors Effects

TAAsGp100, MART-1, PSA, CEA, MUC-1, p53, OVA,LAMP

Lastingly expressing tumor antigens to induce theadverse effects of T-cells special response

CytokinesIL-2, IL-7, IL-12, IL-15, IL-18 [131, 132], IL-21,IL-23, IFN, TNF-α

To enhance the activity of antigen-presenting functionof DCs,

Chemokines CCL21, CCL22, XCL1, CXCL9, CXCL19, CX3CL1To guide lymphocytes to the lymph nodes

To have angiostatic activity

Costimulatory andadhesion molecules

CD40L, CD70, 4-1BBL, OX40L RANKL, CD54,CD58, CD80

To enhance APC’s ability to generate antitumorimmune responses

To improve adhesion interaction between DCs and Tcells

MSCs [135, 136] and HEK293T [137] showed higherefficiency than results by using the conventional and reversemethods. And in 3D systems, cells also exhibited bettermorphology. The reasons may include (1) scaffolds providelarger surface area and space for the interaction betweenDNA-complexes and cells than that in two-dimensional(2D) systems. (2) DNA complexes can be fixed on thescaffolds and prevented from aggregation in 3D systems. (3)Signaling pathway can be influenced by 3D systems. Ali andMooney [129] reported that nonviral vector PEI/pDNA wasimmobilized on a nonwoven fabric with reverse transfectionmethod. The scaffold was treated with negative chargesto facilitate the adsorption of cationic DNA/PEI complex.DCs were effectively transfected in this 3D system, and thelevel of gene expression was significantly higher than thatof conventional transfection. It should be noted that, todate, the application of 3D or scaffold transfer system intransfection of DCs is still a new area. So the methods ofcell seeding and the properties of the scaffold are still neededfurther exploration.

4. Genes Employed in DCs Transfer

The DCs transferred with various genes can steadily andeffectively express the proteins when DCs are refused exvivo. After being transferred with TAA genes, DCs couldexpress multiple antigens and epitopes. These antigenscan be cross presented to MHC. DCs transferred withcytokines genes could produce large amounts of interleukin.Different genes encoding TAAs, cytokines, or chemokinesare utilized to engineer DCs to increase immunogenicity.Subsequently, DCs transferred with the genes encoding TAAscan present the encoded proteins to MHC molecules andthen to mediate T-cell responses. When DCs were transferredwith genes encoding cytokines such as IL-7 or IL-12, theefficacy of generating T cells and immune response canbe increased. When DCs were transferred with the genesencoding chemokines, the chemotaxis of DCs to T cells canbe enhanced. Although this approach has not been used inclinic, it would be a potential strategy for genetic engineeringtechnology on DCs. The genes used to transfer DCs are listedin Table 2.

5. Effective Cancer Immunotherapy Induced byGene-Transferred DCs

In recent years, DC vaccines, especially DNA-based DCvaccines, have been the focus of attention in cancerimmunotherapy. The main process is transferring DCsin vitro, and then implanting them ex vivo. Finally, thetumor-specific CTL response would be activated, and cancercells would be suppressed. These adoptive immunotherapyapproaches have been improved and have achieved partialsuccess in the treatment of malignant melanoma [34], renalcell carcinoma [138], malignant lymphoma [139], and othermalignant diseases [109, 140]. Most of them were phase I/IIclinical trials.

Although various tumor types were studied, melanoma,and prostate cancer are two predominant tumors treatedby gene-modified DC vaccine [141]. The clinical studiesfor DC-based genetically modified vaccination include bothviral and nonviral approaches. In viral vaccinations, recom-binant AdVs were mainly used, but their application in clinicis limited by the biosafety concerns. In contrast, nonviralcarriers are widely used in clinical trials because of their lowtoxicity. The major barrier of nonviral carriers is their lowtransfer efficiency compared with viral vectors. In additionto the gene delivery approach, the conformation of DCsand the route of administration are also considered to bethe important issues. The formulation of DCs includesnot only monocytederived DCs and BMDCs, but alsoimmature and mature ones. The administration methodsmainly include intradermal, intravenous, intranodal, andintratumoral delivery.

The above information suggests that DC-based vaccina-tion against cancer is a promising approach with low adverseeffects, but advanced efficacy studies need to be carried out.Despite this limitation, recombination DC vaccine would beconsidered as an encouraging tool to treat cancer.

6. Conclusions and Future Perspective

DC vaccines can kill the cancer cells with little damageto normal cells by inducing and enhancing patient’s owntumor-specific immune response. The function of DCs couldbe to optimize genetic modification by various TAAs or


immune-modulatory molecules. Therefore, the applicationof ex vivo DC-based vaccination for cancer immunother-apy has many advantages because of its tumor-specificstimulation. However, for clinical applications using DCvaccines, lots of problems need to be solved, such as the lowaffinity between tumor epitopes and MHC, the frequency ofvaccine delivery and immune procedures, and the difficultyto evaluate vaccines effect. Until now, many researchers usedgene carriers and transfection systems in the recombinationof DCs for effective cancer immunotherapy. With thedevelopment of materials science, targeted cell biology andmolecular cytology, and so forth, various strategies have beenintroduced to optimize both viral and nonviral vectors forgene delivery into DCs. In addition, there is a requirement forfurther investigation in the use of the reverse and 3D systemsto improve the nonviral transfection efficiency. These effortsare expected to facilitate future clinical applications of gene-modified DCs for cancer therapy.

Acknowledgments

This work was financially supported by National Natu-ral Science Foundation of China (no. 30973648), Zhe-jiang Provincial Natural Science Foundation of China (no.R2090176) and China-Japan Scientific Cooperation Program(no. 81011140077) supported by both NSFC, China andJSPS, Japan. The authors thank Dr. Yu-Lan Hu (Institute ofPharmaceutics, Zhejiang University) for her critical review ofthe paper.

References

[1] A. Jemal, R. Siegel, E. Ward, Y. Hao, J. Xu, and M. J. Thun,“Cancer statistics, 2009,” CA Cancer Journal for Clinicians,vol. 59, no. 4, pp. 225–249, 2009.

[2] T. Hinz, C. J. Buchholz, T. Van Der Stappen, K. Cichutek, andU. Kalinke, “Manufacturing and quality control of cell-basedtumor vaccines: a scientific and a regulatory perspective,”Journal of Immunotherapy, vol. 29, no. 5, pp. 472–476, 2006.

[3] C. Palena, S. I. Abrams, J. Schlom, and J. W. Hodge, “Cancervaccines: preclinical studies and novel strategies,” Advances inCancer Research, vol. 95, pp. 115–145, 2006.

[4] D. Dudziak, A. O. Kamphorst, G. F. Heidkamp et al.,“Differential antigen processing by dendritic cell subsets invivo,” Science, vol. 315, no. 5808, pp. 107–111, 2007.

[5] M. V. Dhodapkar, K. M. Dhodapkar, and A. K. Palucka,“Interactions of tumor cells with dendritic cells: balancingimmunity and tolerance,” Cell Death and Differentiation, vol.15, no. 1, pp. 39–50, 2008.

[6] C. A. Janeway and R. Medzhitov, “Innate immune recogni-tion,” Annual Review of Immunology, vol. 20, pp. 197–216,2002.

[7] A. A. Itano and M. K. Jenkins, “Antigen presentation to naiveCD4 T cells in the lymph node,” Nature Immunology, vol. 4,no. 8, pp. 733–739, 2003.

[8] G. J. Randolph, V. Angeli, and M. A. Swartz, “Dendritic-cell trafficking to lymph nodes through lymphatic vessels,”Nature Reviews Immunology, vol. 5, no. 8, pp. 617–628,2005.

[9] K. Inaba, J. P. Metlay, M. T. Crowley, and R. M. Steinman,“Dendritic cells pulsed with protein antigens in vitro canprime antigen-specific, MHC-restricted T cells in situ,”Journal of Experimental Medicine, vol. 172, no. 2, pp. 631–640, 1990.

[10] S. Paczesny, H. Ueno, J. Fay, J. Banchereau, and A. K. Palucka,“Dendritic cells as vectors for immunotherapy of cancer,”Seminars in Cancer Biology, vol. 13, no. 6, pp. 439–447, 2003.

[11] A. Rughetti, M. Biffoni, M. Sabbatucci et al., “Transfectedhuman dendritic cells to induce antitumor immunity,” GeneTherapy, vol. 7, no. 17, pp. 1458–1466, 2000.

[12] V. H. Engelhard, T. N. J. Bullock, T. A. Colella, S. L. Sheasley,and D. W. Mullins, “Antigens derived from melanocytedifferentiation proteins: self-tolerance, autoimmunity, anduse for cancer immunotherapy,” Immunological Reviews, vol.188, pp. 136–146, 2002.

[13] E. M. Shevach, “Fatal attraction: tumors beckon regulatory Tcells,” Nature Medicine, vol. 10, no. 9, pp. 900–901, 2004.

[14] X. Zhang, J. R. Gordon, and J. Xiang, “Advances in dendriticcell-based vaccine of cancer,” Cancer Biotherapy and Radio-pharmaceuticals, vol. 17, no. 6, pp. 601–619, 2002.

[15] F. Sallusto and A. Lanzavecchia, “Efficient presentation of sol-uble antigen by cultured human dendritic cells is maintainedby granulocyte/macrophage colony-stimulating factor plusinterleukin 4 and downregulated by tumor necrosis factor α,”Journal of Experimental Medicine, vol. 179, no. 4, pp. 1109–1118, 1994.

[16] J. A. Villadangos, W. R. Heath, and F. R. Carbone, “Outsidelooking in: the inner workings of the crosspresentationpathway within dendritic cells,” Trends in Immunology, vol.28, no. 2, pp. 45–47, 2007.

[17] K. Shortman and S. H. Naik, “Steady-state and inflammatorydendritic-cell development,” Nature Reviews Immunology,vol. 7, no. 1, pp. 19–30, 2007.

[18] B. Dubois, J. M. Bridon, J. Fayette et al., “Dendritic cellsdirectly modulate B cell growth and differentiation,” Journalof Leukocyte Biology, vol. 66, no. 2, pp. 224–230, 1999.

[19] K. Shortman and Y. J. Liu, “Mouse and human dendritic cellsubtypes,” Nature Reviews Immunology, vol. 2, no. 3, pp. 151–161, 2002.

[20] M. Merad and M. G. Manz, “Dendritic cell homeostasis,”Blood, vol. 113, no. 15, pp. 3418–3427, 2009.

[21] R. Maldonado-Lopez, T. De Smedt, B. Pajak et al., “Role ofCD8α and CD8α dendritic cells in the induction of primaryimmune responses in vivo,” Journal of Leukocyte Biology, vol.66, no. 2, pp. 242–246, 1999.

[22] B. Pulendran, J. L. Smith, G. Caspary et al., “Distinct den-dritic cell subsets differentially regulate the class of immuneresponse in vivo,” Proceedings of the National Academy ofSciences of the United States of America, vol. 96, no. 3, pp.1036–1041, 1999.

[23] N. S. Wilson and J. A. Villadangos, “Regulation of antigenpresentation and cross-presentation in the dendritic cell net-work: facts, hypothesis, and immunological implications,”Advances in Immunology, vol. 86, pp. 241–305, 2005.

[24] C. Reis e Sousa, “Dendritic cells in a mature age,” NatureReviews Immunology, vol. 6, no. 6, pp. 476–483, 2006.

[25] J. A. Villadangos and W. R. Heath, “Life cycle, migrationand antigen presenting functions of spleen and lymph nodedendritic cells: limitations of the Langerhans cells paradigm,”Seminars in Immunology, vol. 17, no. 4, pp. 262–272, 2005.


[26] S. Akira and K. Takeda, “Toll-like receptor signalling,” NatureReviews Immunology, vol. 4, no. 7, pp. 499–511, 2004.

[27] H. Kondo, S. Hazama, T. Kawaoka et al., “Adoptiveimmunotherapy for pancreatic cancer using MUC1 peptide-pulsed dendritic cells and activated T lymphocytes,” Anti-cancer Research, vol. 28, no. 1, pp. 379–387, 2008.

[28] E. Meylan, J. Tschopp, and M. Karin, “Intracellular patternrecognition receptors in the host response,” Nature, vol. 442,no. 7098, pp. 39–44, 2006.

[29] P. Salgame, J. S. Abrams, C. Clayberger et al., “Differinglymphokine profiles of functional subsets of human CD4 andCD8 T cell clones,” Science, vol. 254, no. 5029, pp. 279–282,1991.

[30] C. Caux, C. Massacrier, B. Vanbervliet et al., “Activation ofhuman dendritic cells through CD40 cross-linking,” Journalof Experimental Medicine, vol. 180, no. 4, pp. 1263–1272,1994.

[31] A. Macagno, G. Napolitani, A. Lanzavecchia, and F. Sallusto,“Duration, combination and timing: the signal integrationmodel of dendritic cell activation,” Trends in Immunology,vol. 28, no. 5, pp. 227–233, 2007.

[32] S. P. Schoenberger, R. E. M. Toes, E. I. H. Van Dervoort,R. Offringa, and C. J. M. Melief, “T-cell help for cytotoxicT lymphocytes is mediated by CD40-CD4OL interactions,”Nature, vol. 393, no. 6684, pp. 480–483, 1998.

[33] P. Pierre, S. J. Turley, E. Gatti et al., “Developmentalregulation of MHC class II transport in mouse dendriticcells,” Nature, vol. 388, no. 6644, pp. 787–792, 1997.

[34] L. Engell-Noerregaard, T. H. Hansen, M. H. Andersen, P.Thor Straten, and I. M. Svane, “Review of clinical studieson dendritic cell-based vaccination of patients with malig-nant melanoma: assessment of correlation between clini-cal response and vaccine parameters,” Cancer Immunology,Immunotherapy, vol. 58, no. 1, pp. 1–14, 2009.

[35] E. Gilboa, “DC-based cancer vaccines,” Journal of ClinicalInvestigation, vol. 117, no. 5, pp. 1195–1203, 2007.

[36] S. M. Kaech and R. Ahmed, “Immunology: CD8 T cellsremember with a little help,” Science, vol. 300, no. 5617, pp.263–265, 2003.

[37] K. Kawamura, N. Kadowaki, R. Suzuki et al., “Dendritic cellsthat endocytosed antigen-containing IgG-liposomes eliciteffective antitumor immunity,” Journal of Immunotherapy,vol. 29, no. 2, pp. 165–174, 2006.

[38] E. S. Trombetta and I. Mellman, “Cell biology of antigen pro-cessing in vitro and in vivo,” Annual Review of Immunology,vol. 23, pp. 975–1028, 2005.

[39] W. R. Heath, G. T. Belz, G. M. N. Behrens et al., “Cross-presentation, dendritic cell subsets, and the generation ofimmunity to cellular antigens,” Immunological Reviews, vol.199, pp. 9–26, 2004.

[40] A. F. Ochsenbein, S. Sierro, B. Odermatt et al., “Roles oftumour localization, second signals and cross priming incytotoxic T-cell induction,” Nature, vol. 411, no. 6841, pp.1058–1064, 2001.

[41] G. Willimsky and T. Blankenstein, “Sporadic immunogenictumours avoid destruction by inducing T-cell tolerance,”Nature, vol. 437, no. 7055, pp. 141–146, 2005.

[42] S. Raychaudhuri and K. L. Rock, “Fully mobilizing hostdefense: building better vaccines,” Nature Biotechnology, vol.16, no. 11, pp. 1025–1033, 1998.

[43] D. S. Bangari and S. K. Mittal, “Current strategies and futuredirections for eluding adenoviral vector immunity,” CurrentGene Therapy, vol. 6, no. 2, pp. 215–226, 2006.

[44] A. Tuettenberg, C. Becker, E. Huter, J. Knop, A. H.Enk, and H. Jonuleit, “Induction of strong and persistentMelanA/MART-1-specific immune responses by adjuvantdendritic cell-based vaccination of stage II melanomapatients,” International Journal of Cancer, vol. 118, no. 10, pp.2617–2627, 2006.

[45] J. Babatz, C. Rollig, B. Lobel et al., “Induction of cellularimmune responses against carcinoembryonic antigen inpatients with metastatic tumors after vaccination with alteredpeptide ligand-loaded dendritic cells,” Cancer Immunology,Immunotherapy, vol. 55, no. 3, pp. 268–276, 2006.

[46] M. Y. Bae, N. H. Cho, and S. Y. Seong, “Protective anti-tumour immune responses by murine dendritic cells pulsedwith recombinant Tat-carcinoembryonic antigen derivedfrom Escherichia coli,” Clinical and Experimental Immunol-ogy, vol. 157, no. 1, pp. 128–138, 2009.

[47] D. K. Kim, T. V. Lee, A. Castilleja et al., “Folate bindingprotein peptide 191-199 presented on dendritic cells canstimulate CTL from ovarian and breast cancer patients,”Anticancer Research, vol. 19, no. 4, pp. 2907–2916, 1999.

[48] D. K. Kim, J. H. Kim, Y. T. Kim, J. W. Kim, and C. G.Ioannides, “The comparoison of cytotoxic T-lymphocyteeffects of dendritic cells stimulated by the folate bindingprotein peptide cultured with IL-15 and IL-2 in solid tumor,”Yonsei Medical Journal, vol. 43, no. 6, pp. 691–700, 2002.

[49] S. Garetto, F. Sizzano, D. Brusa, A. Tizzani, F. Malavasi, andL. Matera, “Binding of prostate-specific membrane antigento dendritic cells: a critical step in vaccine preparation,”Cytotherapy, vol. 11, no. 8, pp. 1090–1100, 2009.

[50] Y. Waeckerle-Men, E. Uetz-Von Allmen, M. Fopp etal., “Dendritic cell-based multi-epitope immunotherapy ofhormone-refractory prostate carcinoma,” Cancer Immunol-ogy, Immunotherapy, vol. 55, no. 12, pp. 1524–1533, 2006.

[51] H. Yang, N. H. Cho, and S. Y. Seong, “The Tat-conjugatedN-terminal region of mucin antigen 1 (MUC1) inducesprotective immunity against MUC1-expressing tumours,”Clinical and Experimental Immunology, vol. 158, no. 2, pp.174–185, 2009.

[52] J. Wierecky, M. Mueller, and P. Brossart, “Dendritic cell-based cancer immunotherapy targeting MUC-1,” CancerImmunology, Immunotherapy, vol. 55, no. 1, pp. 63–67, 2006.

[53] S. J. Woo, C. H. Kim, M. Y. Park et al., “Co-administrationof carcinoembryonic antigen and HIV TAT fusion proteinwith CpG-oligodeoxynucleotide induces potent antitumorimmunity,” Cancer Science, vol. 99, no. 5, pp. 1034–1039,2008.

[54] L. Fong and E. G. Engleman, “Dendritic cells in cancerimmunotherapy,” Annual Review of Immunology, vol. 18, pp.245–273, 2000.

[55] M. Schott, J. Feldkamp, D. Schattenberg et al., “Inductionof cellular immunity in a parathyroid carcinoma treatedwith tumor lysate-pulsed dendritic cells,” European Journalof Endocrinology, vol. 142, no. 3, pp. 300–306, 2000.

[56] M. Schnurr, P. Galambos, C. Scholz et al., “Tumor celllysate-pulsed human dendritic cells induce a T-cell responseagainst pancreatic carcinoma cells: an in vitro model for theassessment of tumor vaccines,” Cancer Research, vol. 61, no.17, pp. 6445–6450, 2001.

[57] X. Zhao, Y. Q. Wei, and Z. L. Peng, “Induction of T cellresponses against autologous ovarian tumors with wholetumor cell lysate-pulsed dendritic cells,” ImmunologicalInvestigations, vol. 30, no. 1, pp. 33–45, 2001.


[58] N. Inzkirweli, B. Guckel, C. Sohn, D. Wallwiener, G. Bastert,and M. Lindner, “Antigen loading of dendritic cells withapoptotic tumor cell-preparations is superior to that usingnecrotic cells or tumor lysates,” Anticancer Research, vol. 27,no. 4, pp. 2121–2129, 2007.

[59] D. Boczkowski, S. K. Nair, D. Snyder, and E. Gilboa, “Den-dritic cells pulsed with RNA are potent antigen-presentingcells in vitro and in vivo,” Journal of Experimental Medicine,vol. 184, no. 2, pp. 465–472, 1996.

[60] D. M. Ashley, B. Faiola, S. Nair, L. P. Hale, D. D. Bigner,and E. Gilboa, “Bone marrow-generated dendritic cellspulsed with tumor extracts or tumor RNA induce antitumorimmunity against central nervous system tumors,” Journal ofExperimental Medicine, vol. 186, no. 7, pp. 1177–1182, 1997.

[61] R. K. Tyagi, S. Mangal, N. Garg, and P. K. Sharma,“RNA-based immunotherapy of cancer: role and therapeuticimplications of dendritic cells,” Expert Review of AnticancerTherapy, vol. 9, no. 1, pp. 97–114, 2009.

[62] J. J. Hernando, T. W. Park, H. P. Fischer et al., “Vaccinationwith dendritic cells transfected with mRNA-encoded folate-receptor-α for relapsed metastatic ovarian cancer,” LancetOncology, vol. 8, no. 5, pp. 451–454, 2007.

[63] L. Chen, X. D. Tang, S. T. Yu et al., “Induction of anti-tumour immunity by dendritic cells transduced with hTERTrecombinant adenovirus in mice,” Journal of Pathology, vol.217, no. 5, pp. 685–692, 2009.

[64] C. H. Kim, J. S. Yoon, H. J. Sohn et al., “Direct vac-cination with pseudotype baculovirus expressing murinetelomerase induces anti-tumor immunity comparable withRNA-electroporated dendritic cells in a murine gliomamodel,” Cancer Letters, vol. 250, no. 2, pp. 276–283, 2007.

[65] D. Weissman, H. Ni, D. Scales et al., “HIV gag mRNAtransfection of dendritic cells (DC) delivers encoded antigento MHC class I and II molecules, causes DC maturation, andinduces a potent human in vitro primary immune response,”Journal of Immunology, vol. 165, no. 8, pp. 4710–4717, 2000.

[66] C. A. Nicolette, D. Healey, I. Tcherepanova et al., “Dendriticcells for active immunotherapy: optimizing design andmanufacture in order to develop commercially and clinicallyviable products,” Vaccine, vol. 25, no. 2, pp. B47–B60, 2007.

[67] D. Boczkowski and S. Nair, “RNA as performance-enhancersfor dendritic cells,” Expert Opinion on Biological Therapy, vol.10, no. 4, pp. 563–574, 2010.

[68] M. A. Morse and H. K. Lyerly, “DNA and RNA modifieddendritic cell vaccines,” World Journal of Surgery, vol. 26, no.7, pp. 819–825, 2002.

[69] V. F. I. Van Tendeloo, P. Ponsaerts, F. Lardon et al., “Highlyefficient gene delivery by mRNA electroporation in humanhematopoietic cells: superiority to lipofection and passivepulsing of mRNA and to electroporation of plasmid cDNAfor tumor antigen loading of dendritic cells,” Blood, vol. 98,no. 1, pp. 49–56, 2001.

[70] A. Ribas, “Genetically modified dendritic cells for cancerimmunotherapy,” Current Gene Therapy, vol. 5, no. 6, pp.619–628, 2005.

[71] C. A. Mack, W. R. Song, H. Carpenter et al., “Circumventionof anti-adenovirus neutralizing immunity by administrationof an adenoviral vector of an alternate serotype,” HumanGene Therapy, vol. 8, no. 1, pp. 99–109, 1997.

[72] T. L. Wu and H. C. J. Ertl, “Immune barriers to successfulgene therapy,” Trends in Molecular Medicine, vol. 15, no. 1,pp. 32–39, 2009.

[73] N. Okada, T. Saito, Y. Masunaga et al., “Efficient antigengene transduction using Arg-Gly-Asp fiber-mutant aden-ovirus vectors can potentiate antitumor vaccine efficacy andmaturation of murine dendritic cells,” Cancer Research, vol.61, no. 21, pp. 7913–7919, 2001.

[74] H. Mizuguchi and T. Hayakawa, “Targeted adenovirus vec-tors,” Human Gene Therapy, vol. 15, no. 11, pp. 1034–1044,2004.

[75] Y. Eto, Y. Yoshioka, Y. Mukai, N. Okada, and S. Nakagawa,“Development of PEGylated adenovirus vector with targetingligand,” International Journal of Pharmaceutics, vol. 354, no.1-2, pp. 3–8, 2008.

[76] X. Yao, Y. Yoshioka, T. Morishige et al., “Systemic administra-tion of a PEGylated adenovirus vector with a cancer-specificpromoter is effective in a mouse model of metastasis,” GeneTherapy, vol. 16, no. 12, pp. 1395–1404, 2009.

[77] M. L. Edelstein, M. R. Abedi, J. Wixon, and R. M. Edelstein,“Gene therapy clinical trials worldwide 1989–2004—anoverview,” Journal of Gene Medicine, vol. 6, no. 6, pp. 597–602, 2004.

[78] W. C. Adams, E. Bond, M. J. E. Havenga et al., “Adenovirusserotype 5 infects human dendritic cells via a coxsackievirus-adenovirus receptor-independent receptor pathway medi-ated by lactoferrin and DC-SIGN,” Journal of GeneralVirology, vol. 90, no. 7, pp. 1600–1610, 2009.

[79] N. Okada, S. Iiyama, Y. Okada et al., “Immunologicalproperties and vaccine efficacy of murine dendritic cellssimultaneously expressing melanoma-associated antigen andinterleukin-12,” Cancer Gene Therapy, vol. 12, no. 1, pp. 72–83, 2005.

[80] N. Okada, N. Mori, R. Koretomo et al., “Augmentation of themigratory ability of DC-based vaccine into regional lymphnodes by efficient CCR7 gene transduction,” Gene Therapy,vol. 12, no. 2, pp. 129–139, 2005.

[81] A. Lundqvist, A. Choudhury, T. Nagata et al., “Recombinantadenovirus vector activates and protects human monocyte-derived dendritic cells from apoptosis,” Human Gene Ther-apy, vol. 13, no. 13, pp. 1541–1549, 2002.

[82] L. Schumacher, A. Ribas, V. B. Dissette et al., “Humandendritic cell maturation by adenovirus transductionenhances tumor antigen-specific T-cell responses,” Journal ofImmunotherapy, vol. 27, no. 3, pp. 191–200, 2004.

[83] K. Youlin, W. Xiaodong, L. Xiuheng et al., “Anti-tumorimmune response induced by dendritic cells transduced withtruncated PSMA IRES 4-1BBL recombinant adenoviruses,”Cancer Letters, vol. 293, no. 2, pp. 254–262, 2010.

[84] S. Kim, J. B. Lee, K. L. Geon, and J. Chang, “Vaccinationwith recombinant adenoviruses and dendritic cells express-ing prostate-specific antigens is effective in eliciting CTLand suppresses tumor growth in the experimental prostatecancer,” Prostate, vol. 69, no. 9, pp. 938–948, 2009.

[85] N. Jiang, G. S. Wang, H. Li et al., “Immunization withdendritic cells infected with mTERT adenovirus vectoreffectively elicits immunity against mouse H22 hepatoma invivo,” Zhonghua Zhong Liu Za Zhi, vol. 31, no. 6, pp. 405–409,2009.

[86] M. A. Gonzalez-Carmona, V. Lukacs-Kornek, A. Timmer-man et al., “CD40Ligand-expressing dendritic cells induceregression of hepatocellular carcinoma by activating innateand acquired immunity in vivo,” Hepatology, vol. 48, no. 1,pp. 157–168, 2008.


[87] J. Y. Yang, D. Y. Cao, Y. Xue, Z. C. Yu, and W. C. Liu,“Improvement of dendritic-based vaccine efficacy againsthepatitis B virus-related hepatocellular carcinoma by twotumor-associated antigen gene-infected dendritic cells,”Human Immunology, vol. 71, no. 3, pp. 255–262, 2010.

[88] X. P. Zhu, Z. Z. Chen, C. T. Li et al., “In vitro inducingeffect of dendritic cells cotransfected with survivin andgranulocyte-macrophage colony-stimulating factor on cyto-toxic T cell to kill leukemic cells,” Chinese Medical Journal,vol. 121, no. 21, pp. 2180–2184, 2008.

[89] B. Y. Tsai, Y. L. Lin, and B. L. Chiang, “Application ofinterleukin-12 expressing dendritic cells for the treatmentof animal model of leukemia,” Experimental Biology andMedicine, vol. 234, no. 8, pp. 952–960, 2009.

[90] J. H. Huang, S. N. Zhang, K. J. Choi et al., “Therapeuticand tumor-specific immunity induced by combination ofdendritic cells and oncolytic adenovirus expressing IL-12 and4-1BBL,” Molecular Therapy, vol. 18, no. 2, pp. 264–274,2010.

[91] J. Xie, L. Xiong, X. Tao et al., “Antitumor effects ofmurine bone marrow-derived dendritic cells infected withxenogeneic livin α recombinant adenoviral vectors againstLewis lung carcinoma,” Lung Cancer, vol. 68, no. 3, pp. 338–345, 2010.

[92] K. Kikkawa, R. Fujii, T. Kuramoto et al., “Dendritic cellswith transduced survivin gene induce specific cytotoxic Tlymphocytes in human urologic cancer cell lines,” Urology,vol. 74, no. 1, pp. 222–228, 2009.

[93] C. Bello-Fernandez, M. Matyash, H. Strobl et al., “Efficientretrovirus-mediated gene transfer of dendritic cells generatedfrom CD34+ cord blood cells under serum-free conditions,”Human Gene Therapy, vol. 8, no. 14, pp. 1651–1658, 1997.

[94] P. Szabolcs, H. F. Gallardo, D. H. Ciocon, M. Sadelain, andJ. W. Young, “Retrovirally transduced human dendritic cellsexpress a normal phenotype and potent T-cell stimulatorycapacity,” Blood, vol. 90, no. 6, pp. 2160–2167, 1997.

[95] F. D. Bushman, “Integration site selection by lentiviruses:biology and possible control,” Current Topics in Microbiologyand Immunology, vol. 261, pp. 165–177, 2002.

[96] C. Goujon, L. Jarrosson-Wuilleme, J. Bernaud, D. Rigal, J. L.Darlix, and A. Cimarelli, “With a little help from a friend:increasing HIV transduction of monocyte-derived dendriticcells with virion-like particles of SIV,” Gene Therapy, vol. 13,no. 12, pp. 991–994, 2006.

[97] C. Goujon, L. Riviere, L. Jarrosson-Wuilleme et al.,“SIV/HIV-2 Vpx proteins promote retroviral escape froma proteasome-dependent restriction pathway present inhuman dendritic cells,” Retrovirology, vol. 4, article 2, 2007.

[98] C. Esslinger, P. Romero, and H. R. Macdonald, “Efficienttransduction of dendritic cells and induction of a T-cellresponse by third-generation lentivectors,” Human GeneTherapy, vol. 13, no. 9, pp. 1091–1100, 2002.

[99] G. B. Li and G. X. Lu, “Gene delivery efficiency in bonemarrow-derived dendritic cells: comparison of four methodsand optimization for lentivirus transduction,” MolecularBiotechnology, vol. 43, no. 3, pp. 250–256, 2009.

[100] S. Han and L. J. Chang, “Immunity of lentiviral vector-modified dendritic cells,” Methods in Molecular Biology, vol.542, pp. 245–259, 2009.

[101] L. Yang, H. Yang, K. Rideout et al., “Engineered lentivectortargeting of dendritic cells for in vivo immunization,” NatureBiotechnology, vol. 26, no. 3, pp. 326–334, 2008.

[102] K. Breckpot, P. Emeagi, M. Dullaers, A. Michiels, C.Heirman, and K. Thielemans, “Activation of immaturemonocyte-derived dendritic cells after transduction withhigh doses of lentiviral vectors,” Human Gene Therapy, vol.18, no. 6, pp. 536–546, 2007.

[103] R. Stripecke, “Lentiviral vector-mediated genetic program-ming of mouse and human dendritic cells,” Methods inMolecular Biology, vol. 506, pp. 139–158, 2009.

[104] L. Zhang, M. Procuik, T. Fang, and S. K. P. Kung, “Functionalanalysis of the quantitative expression of a costimulatorymolecule on dendritic cells using lentiviral vector-mediatedRNA interference,” Journal of Immunological Methods, vol.344, no. 2, pp. 87–97, 2009.

[105] Y. Zhang, H. Yang, B. Xiao et al., “Dendritic cells trans-duced with lentiviral-mediated RelB-specific ShRNAs inhibitthe development of experimental autoimmune myastheniagravis,” Molecular Immunology, vol. 46, no. 4, pp. 657–667,2009.

[106] P. Veron, S. Boutin, J. Bernard, O. Danos, J. Davoust, and C.Masurier, “Efficient transduction of monocyte- and CD34+-derived Langerhans cells with lentiviral vectors in the absenceof phenotypic and functional maturation,” Journal of GeneMedicine, vol. 8, no. 8, pp. 951–961, 2006.

[107] S. Hacein-Bey-Abina, C. Von Kalle, M. Schmidt et al.,“A serious adverse event after successful gene therapy forX-linked severe combined immunodeficiency,” The NewEngland Journal of Medicine, vol. 348, no. 3, pp. 255–256,2003.

[108] J. Ren, J. Jia, H. Zhang et al., “Dendritic cells pulsed withα-fetoprotein and mutant P53 fused gene induce bi-targetedcytotoxic T lymphocyte response against hepatic carcinoma,”Cancer Science, vol. 99, no. 7, pp. 1420–1426, 2008.

[109] S. J. Antonia, N. Mirza, I. Fricke et al., “Combination of p53cancer vaccine with chemotherapy in patients with extensivestage small cell lung cancer,” Clinical Cancer Research, vol. 12,no. 3, pp. 878–887, 2006.

[110] C. Foged, C. Arigita, A. Sundblad, W. Jiskoot, G. Storm,and S. Frokjaer, “Interaction of dendritic cells with antigen-containing liposomes: effect of bilayer composition,” Vaccine,vol. 22, no. 15-16, pp. 1903–1913, 2004.

[111] M. Yamada, M. Nishikawa, S. Kawakami et al., “Tissue andintrahepatic distribution and subcellular localization of amannosylated lipoplex after intravenous administration inmice,” Journal of Controlled Release, vol. 98, no. 1, pp. 157–167, 2004.

[112] Y. Lu, S. Kawakami, F. Yamashita, and M. Hashida, “Devel-opment of an antigen-presenting cell-targeted DNA vaccineagainst melanoma by mannosylated liposomes,” Biomateri-als, vol. 28, no. 21, pp. 3255–3262, 2007.

[113] Y. Hattori, S. Kawakami, K. Nakamura, F. Yamashita, andM. Hashida, “Efficient gene transfer into macrophages anddendritic cells by in vivo gene delivery with mannosylatedlipoplex via the intraperitoneal route,” Journal of Pharmacol-ogy and Experimental Therapeutics, vol. 318, no. 2, pp. 828–834, 2006.

[114] R. Suzuki, Y. Oda, N. Utoguchi et al., “A novel strategyutilizing ultrasound for antigen delivery in dendritic cell-based cancer immunotherapy,” Journal of Controlled Release,vol. 133, no. 3, pp. 198–205, 2009.

[115] X. Zheng, C. Vladau, X. Zhang et al., “A novel in vivosiRNA delivery system specifically targeting dendritic cellsand silencing CD40 genes for immunomodulation,” Blood,vol. 113, no. 12, pp. 2646–2654, 2009.


[116] A. Orlacchio, S. Martino, I. Di Girolamo, R. Tiribuzi,F. D’Angelo, and A. Datti, “Efficient siRNA delivery bythe cationic liposome DOTAP in human hematopoieticstem cells differentiating into dendritic cells,” Journal ofBiomedicine and Biotechnology, vol. 2009, Article ID 410260,2009.

[117] T. Yoshikawa, S. Imazu, J. Q. Gao et al., “Augmentationof antigen-specific immune responses using DNA-fusogenicliposome vaccine,” Biochemical and Biophysical ResearchCommunications, vol. 325, no. 2, pp. 500–505, 2004.

[118] E. Yuba, C. Kojima, N. Sakaguchi, A. Harada, K. Koiwai,and K. Kono, “Gene delivery to dendritic cells mediated bycomplexes of lipoplexes and pH-sensitive fusogenic polymer-modified liposomes,” Journal of Controlled Release, vol. 130,no. 1, pp. 77–83, 2008.

[119] W. Weecharangsan, P. Opanasopit, T. Ngawhirunpat et al.,“Evaluation of chitosan salts as non-viral gene vectors inCHO-K1 cells,” International Journal of Pharmaceutics, vol.348, no. 1-2, pp. 161–168, 2008.

[120] M. Lavertu, S. Methot, N. Tran-Khanh, and M. D.Buschmann, “High efficiency gene transfer using chi-tosan/DNA nanoparticles with specific combinations ofmolecular weight and degree of deacetylation,” Biomaterials,vol. 27, no. 27, pp. 4815–4824, 2006.

[121] T. Dastan and K. Turan, “In vitro characterization anddelivery of chitosan-DNA microparticles into mammaliancells,” Journal of Pharmacy and Pharmaceutical Sciences, vol.7, no. 2, pp. 205–214, 2004.

[122] T. H. Kim, J. W. Nah, M. H. Cho, T. G. Park, and N. S.Cho, “Receptor-mediated gene delivery into antigen present-ing cells using mannosylated chitosan/DNA nanoparticles,”Journal of Nanoscience and Nanotechnology, vol. 6, no. 9-10,pp. 2796–2803, 2006.

[123] T. H. Kim, H. Jin, H. W. Kim, M. H. Cho, and C. S. Cho,“Mannosylated chitosan nanoparticle-based cytokine genetherapy suppressed cancer growth in BALB/c mice bearingCT-26 carcinoma cells,” Molecular Cancer Therapeutics, vol.5, no. 7, pp. 1723–1732, 2006.

[124] X. Zhou, B. Liu, X. Yu et al., “Controlled release of PEI/DNAcomplexes from mannose-bearing chitosan microspheres as apotent delivery system to enhance immune response to HBVDNA vaccine,” Journal of Controlled Release, vol. 121, no. 3,pp. 200–207, 2007.

[125] S. C. Ng, S. Plamondon, M. A. Kamm et al., “Immuno-suppressive effects via human intestinal dendritic cells ofprobiotic bacteria and steroids in the treatment of acuteulcerative colitis,” Inflammatory Bowel Diseases, vol. 16, no.8, pp. 1286–1298, 2010.

[126] T. Kanazawa, Y. Takashima, M. Murakoshi, Y. Nakai,and H. Okada, “Enhancement of gene transfection intohuman dendritic cells using cationic PLGA nanosphereswith a synthesized nuclear localization signal,” InternationalJournal of Pharmaceutics, vol. 379, no. 1-2, pp. 187–195,2009.

[127] C. Wang, Q. Ge, D. Ting et al., “Molecularly engineeredpoly(ortho ester) microspheres for enhanced delivery ofDNA vaccines,” Nature Materials, vol. 3, no. 3, pp. 190–196,2004.

[128] L. Thiele, H. P. Merkle, and E. Walter, “Phagocytosisand phagosomal fate of surface-modified microparticles indendritic cells and macrophages,” Pharmaceutical Research,vol. 20, no. 2, pp. 221–228, 2003.

[129] O. A. Ali and D. J. Mooney, “Sustained GM-CSF andPEI condensed pDNA presentation increases the level andduration of gene expression in dendritic cells,” Journal ofControlled Release, vol. 132, no. 3, pp. 273–278, 2008.

[130] J. R. Cubillos-Ruiz, X. Engle, U. K. Scarlett et al.,“Polyethylenimine-based siRNA nanocomplexes reprogramtumor-associated dendritic cells via TLR5 to elicit therapeu-tic antitumor immunity,” Journal of Clinical Investigation, vol.119, no. 8, pp. 2231–2244, 2009.

[131] L. Vujanovic, E. Ranieri, A. Gambotto, W. C. Olson, J. M.Kirkwood, and W. J. Storkus, “IL-12p70 and IL-18 gene-modified dendritic cells loaded with tumor antigen-derivedpeptides or recombinant protein effectively stimulate specificType-1 CD4+ T-cell responses from normal donors andmelanoma patients in vitro,” Cancer Gene Therapy, vol. 13,no. 8, pp. 798–805, 2006.

[132] X. M. Tong, S. E. Zheng, A. Bader et al., “Construction ofexpression vector of HTERT- HIL18 fusion gene and induc-tion of cytotoxic T lymphocyte response against hTERT,”European Journal of Medical Research, vol. 13, no. 1, pp. 7–14, 2008.

[133] O. Mannherz, D. Mertens, M. Hahn, and P. Lichter,“Functional screening for proapoptotic genes by reversetransfection cell array technology,” Genomics, vol. 87, no. 5,pp. 665–672, 2006.

[134] A. Okazaki, J. I. Jo, and Y. Tabata, “A reverse transfectiontechnology to genetically engineer adult stem cells,” TissueEngineering, vol. 13, no. 2, pp. 245–251, 2007.

[135] C. Holladay, M. Keeney, U. Greiser, M. Murphy, T. O’Brien,and A. Pandit, “A matrix reservoir for improved control ofnon-viral gene delivery,” Journal of Controlled Release, vol.136, no. 3, pp. 220–225, 2009.

[136] Y. Lei and T. Segura, “DNA delivery from matrix metallopro-teinase degradable poly(ethylene glycol) hydrogels to mousecloned mesenchymal stem cells,” Biomaterials, vol. 30, no. 2,pp. 254–265, 2009.

[137] J. H. Jang, Z. Bengali, T. L. Houchin, and L. D. Shea,“Surface adsorption of DNA to tissue engineering scaffoldsfor efficient gene delivery,” Journal of Biomedical MaterialsResearch A, vol. 77, no. 1, pp. 50–58, 2006.

[138] B. Frankenberger, S. Regn, C. Geiger et al., “Cell-basedvaccines for renal cell carcinoma: genetically-engineeredtumor cells and monocyte-derived dendritic cells,” WorldJournal of Urology, vol. 23, no. 3, pp. 166–174, 2005.

[139] X. P. Zhu, Z. Z. Chen, J. D. Hu, C. T. Li, T. Yang, and Z.S. Xu, “Induction of anti-lymphoma cytotoxic T cell effectby dendritic cells transfected with recombinant adenovirusvectors carrying survivin gene,” Zhongguo Shi Yan Xue Ye XueZa Zhi, vol. 15, no. 3, pp. 591–593, 2007.

[140] W. Xiong, M. Candolfi, C. Liu et al., “Human Flt3L generatesdendritic cells from canine peripheral blood precursors:implications for a dog glioma clinical trial,” PLoS One, vol.5, no. 6, Article ID e11074, 2010.

[141] E. L. Smits, S. Anguille, N. Cools, Z. N. Berneman, and V.F. Van Tendeloo, “Dendritic cell-based cancer gene therapy,”Human Gene Therapy, vol. 20, no. 10, pp. 1106–1118, 2009.


Review Article

Theoretical Modeling Techniques and Their Impact onTumor Immunology

Anna Lena Woelke, Manuela S. Murgueitio, and Robert Preissner

Structural Bioinformatics Group, Institute for Physiology, Charite - Universitatsmedizin Berlin, Lindenberger Weg 80,13125 Berlin, Germany

Correspondence should be addressed toAnna Lena Woelke, [email protected] and Manuela S. Murgueitio, [email protected]

Received 30 June 2010; Revised 10 October 2010; Accepted 11 October 2010

Academic Editor: Robert E. Cone

Copyright © 2010 Anna Lena Woelke et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Currently, cancer is one of the leading causes of death in industrial nations. While conventional cancer treatment usually resultsin the patient suffering from severe side effects, immunotherapy is a promising alternative. Nevertheless, some questions remainunanswered with regard to using immunotherapy to treat cancer hindering it from being widely established. To help rectify thisdeficit in knowledge, experimental data, accumulated from a huge number of different studies, can be integrated into theoreticalmodels of the tumor-immune system interaction. Many complex mechanisms in immunology and oncology cannot be measuredin experiments, but can be analyzed by mathematical simulations. Using theoretical modeling techniques, general principles oftumor-immune system interactions can be explored and clinical treatment schedules optimized to lower both tumor burdenand side effects. In this paper, we aim to explain the main mathematical and computational modeling techniques used in tumorimmunology to experimental researchers and clinicians. In addition, we review relevant published work and provide an overviewof its impact to the field.

1. Introduction

Biological systems possess a high degree of complexity. Therole a component plays in the organism is not only defined byits function but also by its interaction network [1]. The fieldof systems biology is concerned with that topic and aims atunderstanding the interactions between various componentsof the living cell, such as genes, proteins, and metabolites[2]. A huge mass of biological facts has been uncovered bymolecular biology, but understanding biological complexityon a systems level can only be achieved by a combination ofexperimental and computational approaches [3].

The immune response to tumor formation representsa complex system that can solely be understood by usingseveral different research strategies. In recent time, it hasbeen shown that the immune system plays a pivotal role inthe regulation of cancer, enhancing its growth by certainmechanisms [4, 5] and being capable of recognizing and

eradicating tumors as well [5, 6]. As conventional cancertherapy usually involves severe side effects, increased researchefforts have been made in order to stimulate the immuneresponse against the tumor [7, 8]. Various approachesincluding vaccination [9, 10] or direct injection of antibodies[11], lymphocytes [12–14], or cytokines [15, 16] have beendeveloped.

Complex systems can be analyzed through several math-ematical and computational approaches. An overview of thedistinct steps in model generation is given in Figure 1. Theavailable knowledge about a given biological phenomenonis used to build up a model. Once validated through thecomparison with experimental results or the literature, themodel can be used to perform in silico experiments, whichallow the generation of hypotheses on the behavior of thebiological system. These findings can in turn be verified invivo or in vitro and the data acquired this way can be usedto refine the model, hence enhancing its significance. This


Biological phenomenon

New data In vivo/in vitroexperiment

In silico experiment(simulation) Hypothesis

Experimental results,data analysis

Mathematical orcomputational model

New data

Acquisition of new knowledgeand understanding

Figure 1: Steps in knowledge generation in system biology. Knowledge on a given biological phenomenon is generated in a stepwise mannerby combining modeling and experimental techniques. A model is built up enabling the researcher to perform in silico experiments in orderto predict the behavior of the biological system under given conditions. These predictions have, in turn, to be validated via in vivo or in vitrosystems, leading to further refinement of the model and the underlying hypothesis.

Table 1: Examples of possible data sources for model building and refinement.

Experimental results from the literature or own laboratory Databases

Experimental results: General databases:

(i) Cell division rates (i) KEGG [17] (http://www.genome.jp/kegg/)

(ii) Expression data (ii) Reactome [18] (http://www.reactome.org/)

(iii) Growth curves Immunological databases:

(iv) Binding affinities (i) AntiJen [21] (http://www.darrenflower.info/antijen/)

(v) Diffusion coefficients (ii) IEDB [19] (http://www.immuneepitope.org/)

(iii) InnateDB [22] (http://www.innatedb.org/)

iterative course of action leads to a better understandingof biological systems. In this work, we introduce the basicsteps in model generation and present the two most commonmethods in simulating the interaction between a growingtumor and the immune system: differential equations andrule-based models. We compare them and point out theirimpact on the field of tumor immunology in diverseapplications as for example the identification of cancermechanisms or tumor therapy.

2. Construction of a Mathematical Model

Constructing a model always starts with the collection ofrelevant data in order to define the problem. Afterwards, themodel is built in an iterative process including parameterfitting and validation steps. In the next paragraph, weprovide an overview of what kind of information is collectedand how it is built into the network. We will then go on toexplain how the model is refined and optimized.

2.1. Data Retrieval. The first step in model construction isthe retrieval of information about the given parameters, forexample, cell division rates or enzyme kinetics, which can beextracted from the literature or from public databases, forexample KEGG [17] or Reactome [18] (see Table 1). Focusedimmunological information can be found in specializeddatabases (e.g. IEDB [19]). More detailed information aboutspecific processes is extracted from experimental results.These can be gained by carrying out own experimentsor cooperation with wet-lab researchers, or by carefullymining the published literature. In some cases, parametersare not experimentally accessible; here, data fitting must beperformed to estimate them. The values must be adjusted sothat the biological feasibility of the model is preserved (for adetailed review, see [20]).

2.2. Building and Refining the Model. The model is builtand refined in an iterative process. The basic assumptionsof the behavior of the system are formulated as differential


equations or as computational rules and the parameters aredefined as described above. The resulting model is usedto simulate the system under different conditions and theresults are compared to wet-lab experiments or clinical data.If the simulation differs from the experimental results, themodel parameters get adjusted, or the model is refined toaccount for a more divergent systems behavior. In turn, themodel can explain the system more in detail and gives morerealistic predictions.

A model relating to the nuclear factor “kappa-light-chain-enhancer” of activated B-cells (NF-κB) signaling path-way illustrates clearly how a mathematical model is refinedin a stepwise manner by the addition of new experimen-tal knowledge which in turn leads to an ever growingunderstanding of the system. Lipniacki et al. developed amathematical model of the NF-κB regulatory module [23].The model exhibits two-compartment kinetics built as asystem of differential equations and it can reproduce the timebehavior of the involved protein and mRNA levels and of thecatalytic activity of IκB kinase. After further research, thismodel was refined to include two classes of switches thatare invoked stochastically, namely, the cell-surface receptoractivation by the tumor necrosis factor-α (TNF-α) ligandand the activation of genes by NF-κB [24]. These stochasticswitches allow single cells to respond differently to theirneighbors, with each individual response being unequivocal.In cooperation with a wet-lab research group, it was shownthat the activation by TNF-α is indeed heterogeneous andsingle cells respond in a digital process with fewer cellsresponding at lower doses [25]. In addition, they found thatsome parameters change analogously, for example, NF-κBpeak intensity, response time, and number of oscillations.Consequently, the mathematical model was refined againwith a cellular variation in the amount of TNFα-receptorand a nonlinear activation profile of IκB kinase. This finalmodel is able to reproduce both the digital and analoguedynamics as well as most gene expression profiles at allmeasured conditions. It can also predict the fraction ofcells responding to consecutive short pulses of low-doseTNF-α with high accuracy. Our understanding of theTNF-α-induced NF-κB-signaling has improved significantlythrough the close cooperation and reciprocal influence ofexperimental and theoretical research.

3. Differential Equations Systems

Differential equations are used to describe several principlesin physics or chemistry or to simulate complex systems inbiology and economics. A system of differential equationsallows modeling of time-dependent cellular phenomena suchas individual biochemical reactions, signal transduction cas-cades, or even the interaction between whole cell populations[26]. Each entity that is considered to be important for thequestion of interest is modeled by one differential equationdescribing its production and its decay or its influx andefflux from a compartment. In Figure 2, a simplified modelis shown consisting of two equations describing the amountof tumor and CD8 T-cells over time. By changing only

one parameter, the result can switch from an exponentiallygrowing tumor to a tumor being recognized and destroyedby the immune system. An equivalent system of differentialequations can be analyzed for several criteria. In sensitivityanalysis, the impact on the amount of entities is determinedfor any change of the parameter’s magnitude. In equilibriumanalysis, parameter values are identified for which theentities meet a steady state, meaning their amount does notchange over time; for example, in a tumor-free equilibrium,the tumor is kept under control by the immune system.Threshold criteria are defined at which the behavior of thesystem changes from one state to the other. In bifurcationanalysis, the solution space is scanned for discontinuousparts. Using this tool, the point can be found where oneparameter changes the behavior of the system all of a sudden,jumping from one state to another; for example, below thebifurcation point, the patient remains tumor-free and aboveit, he develops a growing tumor.

In tumor immunology, several different problems havebeen addressed using differential equations systems. Thesimplest form consists of ordinary differential equations(ODEs) that can be solved analytically to find maxima andminima, for example, the maximal survival probability of thepatient.

3.1. Ordinary Differential Equations to Find Generic Princi-ples. Some of the mathematical models describe the tumor-immune interaction generally to find common mechanisms.A generic model of the influence of cytotoxic T-cells ispresented by Kuznetsov et al. [29] which helps to explainthe phenomena of tumor dormancy and sneaking throughin a mathematical way. Leon et al. [30] explore the impactof regulatory CD25 CD4 T-cells on cancer. They proposetwo alternative modes of unbounded tumor growth. Eitherthe tumor induces the production of effector T-cells thatoutcompete regulatory T-cells but are not able to eradicatethe tumor, or a balanced expansion of both effector andregulatory T-cells is induced by the tumor, which preventsit from being destroyed by the immune cells.

The different roles of NK cells and CD8 T-cells in tumorsuppression were investigated in another study [27]. Theauthors highlight the importance of CD8 T-cells in tumoreradication and suggest that immunotherapy should focuson the increase of their activity.

3.2. Ordinary Differential Equations in Specialized Therapy.The effect of innovative new cancer therapies can beestimated using differential equations systems. The influenceof the newly characterized IL-21 in cancer immunotherapywas explored by Cappuccio et al. [31]. This interleukin hasa role in the transition from innate immunity to adaptiveimmunity, and thus the authors suggest that lower dosesof IL-21 should be used for low immunogenic tumors andhigher doses for highly immunogenic ones. In addition, theyfind that cytokine gene therapy is more promising thanhydrodynamics-based gene delivery.

Bunimovich-Mendrazitsky et al. [32, 33] focus on a morespecific type of cancer and explore the effect of pulsed


hSteepness coefficient of the CD8

T-cell competition term 2.02∗105 cell2

gProportionality factor for

immunogenicity of the tumor

0.3 (non-im.)1.4 (im.)

Proportionality factor forCD8 T-cell death rate

2∗10−4/(cell∗day)f

eSteepness coefficient of thetumor competition term

2.73

c Proportionality factor for tumorlysis rate by T-cells

1.4/day

b 5.14∗10−10/dayTumor cell necrosis rate

a Tumor growth rate 1.1/day

Parameters Description Value

Celldeath

Clonal division

dK

dt= g

T2

h + T2 K − f K2

dT

dt= aT − bT − c

(K/T

e + K/T

)T

Cellbirth

Celldeath

Lysisby T-cells

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

Nu

mbe

rof

cells

1 6 11 16 21 26 31 36

Time steps

Immunogenic tumor

Tumor cellsCD8 T-cells

Non-immunogenic tumor

0

200

400

600

800

1000

1200

1400

1600

1800

2000

Nu

mbe

rof

cells

1 6 11 16 21 26 31 36

Time steps

Figure 2: Differential equations system. A simplified differential equations system describes the interaction between tumor cells (T) andCD8 T-cells (K). The only difference between the simulation of a nonimmunogenic tumor (upper graph) and an immunogenic tumor(lower graph) is that parameter g is increased leading to an increase in tumor immunogenicity. Lysis of tumor cells by the CD8 T-cells andthe activation of CD8 T-cells by tumor cells follow a Michaelis-Menten kinetic to account for saturation at high cell numbers. The deathrate of CD8 T-cells is proportional to the square of the cell number to assure for fast declining waves of T-cell expansion. Equations andparameters modified from [27], parameters originally obtained from a published mouse study [28].

and continuous immunotherapy with Bacillus Calmette-Guerin—an attenuated strain of Mycobacterium bovis—totreat superficial bladder cancer. They calculate the amount ofbacterial solution by which the tumor is eradicated but onlylittle side effects are induced.

Kronik et al. [34] investigate the dynamics of Glioblas-toma, a highly aggressive primary brain tumor, treated withex vivo activated cytotoxic T-cells. Model analysis suggeststhat tumor eradication requires a 20-fold higher dose thanhad been administered in clinical studies.

Another role of the immune system in tumor formationis concerned in the cancer treatment using an oncolyticvirus. In this case, the immune response against the hostis considered to be negligible, while the immune cells candestroy the virus, before it is able to enter the tumor cell.To find the optimal virus administration circumventing astrong immune response against the virus, Wein et al. [35]considered different sites of virus application. They concludethat injections should be distributed equally within a solidtumor; core or rim injections alone will eventually result in


tumor escape. Tao and Guo [36] extend this model, focusingon the diffusion of the virus and the immune cells andnarrow down possible forms of optimal treatment.

3.3. Ordinary Differential Equations in Molecular Detail. Adifferential equations system can also focus on a specificmolecular interaction. Accordingly, the importance of theFas/FasL system was emphasized by Webb et al. [37]. TheFas ligand (FasL) can induce apoptosis in the target cellexpressing the Fas receptor, while both ligand and receptorare expressed on tumor cells and T-cells at different levelsdepending on developmental state. The model shows thattumor regression could be enhanced by upregulated Fasreceptor expression in tumor cells, but an even greatersuccess would be gained by constitutive FasL expression inactivated T-cells. This has an implication for the clinicaluse of broad spectrum matrix metalloproteinase (MMP)inhibitors as antiangiogenic agents. In the model, MMP inac-tivation results in increased transmembrane FasL and leadsto a higher rate of Fas-mediated apoptosis in lymphocytesthan in tumor cells; therefore, MMP treatment might becounterproductive.

3.4. Principle of Optimal Control. In the principle of optimalcontrol, a control function is established that quantifies thewanted and unwanted impact of the parameters on theoutcome of the model. The control function is minimized tofind the desired solution. This method is based on the workof Kacser and Burns [38] and Heinrich and Rapoport [39]. Atypical application nowadays is to find the best vaccinationstrategy with the lowest vaccine burden that is still ableto eliminate the tumor. Exploring this question in humanpatients would be very labor intensive or almost impossible,but can easily be done using differential equations.

De Pilis et al. [40, 41] examine the effect of immunother-apy and chemotherapy on cancer growth and find that eachtherapy alone is not able to control the tumor; therefore, theyrecommend combination therapy. Also, IL-2 and adoptivecellular immunotherapy are compared in another study(Kirschner 1998), and the combination of both is ableto reduce the tumor and has the least risk of inducingautoimmunity.

Castiglione and Piccoli [42, 43] apply the principle ofoptimal control to determine an exact vaccination schedulein immunotherapy with autologous dendritic cell transfec-tion. For that purpose, they build a cost function summingup the burden of the tumor and of all side effects of theimmune therapy depending on the vaccination scheduleand then minimizing that function to receive the optimalschedule. They recommend a vaccination schedule with onehigh-dose injection at the beginning of the treatment, andthe other injections being smaller dosages distributed almostequally over the rest of the six months treatment period [43].

3.5. Delay Differential Equations. Differential equations canbe enhanced to become delay differential equations (DDEs)that can account for time consumption in processes like celldivision or other specific behavior. Delayed feedback and

oscillatory behavior can be efficiently described using thismethod. For instance, Kim et al. [44] study the dynamicsof chronic myelogenous leukemia (CML) under imatinibtreatment including the influence of immune cells. They usedelay differential equations, whereby the delay term is usedto incorporate the time for cell division. The model suggestsa combination of immunotherapy and imatinib treatment tooptimally sustain the antileukemia T-cell response. Anotherstudy elucidates the effect of immunotherapy in leukemiapatients after bone marrow transplantation to study specif-ically the graft-versus-leukemia effect [45]. A delay term inthe differential equations system is used to account for theprogression of cells through different modes of behavior.The authors conclude that high concentrations of donorT-cells slightly favor tumor elimination, but also increasethe risk of graft-versus-host disease. Interestingly, higherinitial concentrations of general host blood cells enhance thesuccess rate more significantly whilst also avoiding the risk ofgraft-versus-host disease. This result can be applied directlyto clinical treatment.

3.6. Partial Differential Equations. Partial differential equa-tions (PDEs) are the most advanced form of differentialequations; hence, they are also most demanding mathemat-ically. PDEs are commonly used either to account for theprogression of cells through a developmental process (age-structured model) or to model spatiality (spatiotemporalmodel). Matzavinos et al. [46] make use of PDEs to studythe geometry of a tumor interacting with tumor-infiltratingcytotoxic lymphocytes (TICLs). This approach focuses onthe motility of TICLs that can move at random or towardsincreasing chemokine concentration inside the tumor. Themechanism elucidated in this work may help to explainthe phenomenon of tumor dormancy. A similar approach[13] tries to illuminate the growth pattern of a solid tumordepending on the attack of tumor-associated macrophagesand their movement.

PDEs can also be used to model the movement oftumor cells, as in the study of Eikenberry et al. [47], wheremelanoma invasion into healthy tissue was simulated. Theauthors observe that immune cells can have opposed effectsas they can both destroy tumors or can induce tumorigenicexpansion through the production of angiogenic factors.

4. Rule-Based Modeling

In immunology, the two main simulation approaches inrule-based modeling are Agent-Based Models (ABMs) andCellular Automata (CA), which are closely related. In ABMs,discrete autonomous units or agents interact with eachother at discrete time steps following a set of logical rules,depending on the state of their environment. The agents areidentifiable, and their environment is represented by a grid.They are able to adapt to changes in their surroundings and,therefore, need some sort of memory. A simplified ABM isshown in Figure 3, illustrating how observed phenomena aretranslated into behavioral rules for the entities in a grid.


Tumor cell

CD8 T-cell

CD4 T-cell

Dendritic cell

B cell

Antibody

Legend:

(a) (b) (c)

Tumor cell:

If (close to active CD8 T-cell)−→ die

else−→ divide

CD8 T-cellIf (close to tumor cell)−→ get active, lyse tumor cell

else−→move or die unstimulated

Figure 3: Agent-based modeling. (a) Lymphocytes infiltrating an urothelial carcinoma of the bladder. (b) Translation to an ABM. Cellsmove and interact in a grid. Each cell can occupy one grid space, while antibodies or cytokines have a continuous concentration at each gridspace. (c) A simplified part of the underlying rules for the agents is shown.

CA are closely related to ABMs even though there aresome important differences. In ABMs, agents are mobilewhereas in CA they have fixed positions. Updating of theagents’ state is usually performed in a synchronous way in CAwhile this is not always the case in ABMs. The understandingof the environment in ABMs also differs from the one in CA.In ABMs, it is discretized into micro-compartments whichcan hold a variety of information whereas the environmentof agents in CA is described by the von Neumann or Mooreneighborhood which consider four or eight neighbors foreach agent, respectively [48]. Another aspect is that CArule sets mostly comprise strictly deterministic rules, andABMs often include a mixture of stochastic and deterministicelements. An example of CA is shown in Figure 4, where aprostate tumor is reconstructed using a CA called CancerSim[49]. The three-dimensional visualization of CancerSim canbe compared to the observed tumor and the model cansimulate the progression of cancer.

It has to be kept in mind that the modeling approaches,ABMs and CA, are very similar to each other as they areboth rule-based, and it is not unusual to find hybrid formsin which elements from both methods are used.

In immunology, rule-based models are particularly use-ful because cells and molecules are modeled as individualagents that may have specific ligands or receptors ontheir surface. An additional advantage is the possibilityof including the three-dimensional space explicitly in thesimulation, where each cell can be exactly located and changeits activation state depending on its direct environment. Thisway, complex patterns evolve from a set of simple behavioralrules.

4.1. Tumor-Immune System Interaction in Rule-Based Mod-eling. When simulating the immune response to tumor

formation, the use of discrete modeling techniques as CA orABMs is advantageous due to the possibility of consideringthe activity of individual cells and their interactions. The ideaof simulating the immune system using discrete automatawas first introduced by Kaufman et al. [50]. In theirmodel, Boolean values are used to represent different cellularpopulations whereas the interactions between each other aredefined by simple rules.

In early approaches to simulating tumor-immune inter-actions, each automaton describes the concentration of onecell type [51]. Discrete two-state variables are used to specifythe concentration of particular cell types (high or low) andthe functionality of their epitopes (is recognized, is notrecognized). In this model, the killer role of macrophages aswell as the difference between antigen recognition by T- or B-cells is ignored; still, the model captures many crucial aspectsof the immune response to tumor formation.

Agent-based modeling proves to be particularly usefulfor modeling the early stages of tumor growth before solidtumor or metastasis formation, because in this phase oftumor development, it is necessary to consider the activityof each single cell. A model dealing with the interaction ofthe immune system with a tumor at this stage was presentedby Mallet and de Pillis [52]. In this model, a cluster oftumor cells is studied which are supplied with nutrientsthrough a blood vessel. The dependence of different tumormorphologies such as spherical and papillary or lymphocyte-infiltrated growth on several key model parameters relatedto the interplay between the immune system and the tumoris shown. For this purpose, a hybrid modeling approach isused. Cells’ behavior is described by a set of probabilisticrules whereas chemical diffusion is simulated via determinis-tic PDEs. The simulation comprises NK cells and cytotoxic T-lymphocytes and considers their spatiotemporal interactionwith normal and tumor cells.


(a) (b)

Figure 4: Cellular Automata. (a) Human prostate gland containing carcinoma at the lower right part, seen as a yellowish mass [23]. (b)Cellular automata named CancerSim [49] simulating tumor growth. Tumor cells are shown in blue, healthy tissue in gray, and blood vesselsin red, dual view with and without healthy tissue. The three-dimensional shape of the prostate gland and the simulated structure can becompared to improve the model and to explain the observed phenomena.

4.2. Rule-Based Modeling in Immune Therapy. Recently,immune therapy of tumors is becoming increasingly wellestablished [55, 56]. Nevertheless, several questions con-cerning the dosage, vaccination schedules, usage of carriers,and so forth are still open. Answering these questionsexperimentally remains difficult due to the high number ofexperiments required, which result in high costs and arevery time consuming. Simulations can be extremely helpfulin order to predict optimal therapy strategies and reducethe amount of experiments necessary. Several attempts toanswer these questions have been performed by differentgroups based on a former ABM of the immune system calledIMMSIM [57].

Among them, Castiglione et al. developed an elaboratedmodel of the immune response to tumor antigens [58]. Themodel includes several behavioral patterns of the immunesystem: hematopoesis, antigen digestion and presentationby B-lymphocytes, macrophages and dendritic cells, thehypermutation of antibodies, and last but not least, cyto-toxicity by CD8 T-cells. In order to simulate the immunerecognition, epitopes, and peptides are represented by binarystrings, and their immunogenicity is defined by the hammingdistance, which is the number of complementary bits in bit-wise comparison. The aim of this work is to evaluate theeffects of repeated injections of tumor-associated antigen(TAA) together with carriers on the humoral as well ason the cellular immune response. The authors find thatthe administration of TAA with multiple carriers causesthe strongest immune response against the tumor. Theadministration of TAA with one single carrier fails becausea stronger immune response against the carrier and not theTAA takes place.

ABMs are also used to describe tumor vaccination inmice with the Triplex vaccine. The model SimTriplex is anABM that describes the relevant processes of the competitionbetween mammary carcinoma and the immune system [59,60]. Its results show a strong correlation with experimentalresults. This modeling framework is also based on IMMSIM[57]. In a further approach, the model is used to optimizevaccination schedules by the use of a genetic algorithm to

drive the simulator [61] and later by a simulated annealingapproach [62]. This makes it possible to use the model as avirtual mouse with which extensive in silico experiments canbe performed.

4.3. Virotherapy Simulated by Agent-Based Modeling.Another approach in tumor therapy is to attack the cancerwith oncolytic viruses, which are capable of killing cancercells or inducing an immune response against them.Agent-based modeling is especially suitable to simulateoncolytic virotherapy due to the possibility of describingthe interplay between cancer cells, viruses, and the immunesystem individually and in a multiscale way. The authorsof this work [63], construct a hybrid ABM-PDE modelwhich illustrates virotherapy in a stage of avascular tumorgrowth. The multiscale dynamics of tumor growth aredefined by probabilistic CA rules whereas the dynamics ofnutrients and viruses are described by reaction-diffusionequations. The modeling result suggests that for a successfulsingle-agent virotherapy, the host immune system must bestrongly inhibited, and a potent virus with high intratumoralmobility is to be used.

5. Application of Modeling Results

The most elaborate modeling approach is futile if it cannotbe applied to reality. Therefore, it is essential to comparesimulation results of a model to experimental data from thelaboratory or the clinic. As an example, a comparison of themodel of Kim et al. [44] to clinical data is shown in Figure 5.The parameters of three different patients (Figures 5(a), 5(b),and 5(c)) are incorporated into this model simulating theimmune response to Chronic Myelogenous Leukemia underimatinib treatment. Clinically observed T-cell numbers [53]are compared to the simulated curve that represents thedata points sufficiently. Based on several of these tests, themodel is used to simulate the course of disease shown inthe “Leukemia” curve and is compared to a similar modelprediction that does not account for immune reaction (“Noimmune response”).


0

0.01

0.02

0.03

0.04

0.05

0.06

Cel

lcon

cen

trat

ion

(k/μ

L)

0 10 20 30 40 50

Time (months)

Leukemia

T-cells

No immuneresponse

(a)

0

0.01

0.02

0.03

0.04

0.05

0.06

Cel

lcon

cen

trat

ion

(k/μ

L)

0 10 20 30 40 50

Time (months)

Leukemia

T-cells

No immuneresponse

(b)

0

0.01

0.02

0.03

0.04

0.05

0.06

Cel

lcon

cen

trat

ion

(k/μ

L)

0 10 20 30 40 50

Time (months)

LeukemiaT-cells

No immuneresponse

(c)

Figure 5: Differential equations model validation with experimental data of chronic myelogenous leukemia treatment with imatinib [44].(a), (b), and (c) show data from different patients along with simulated results using individual parameters for each of the patients.Simulated cell concentrations are compared to measurements of T-cell numbers from [53] (cell numbers are scaled down by 2500 to showrelative magnitudes). Simulated curves are shown as lines and experimental data as black squares; the horizontal dashed line indicates theapproximate level of complete cytogenetic remission. “No immune response” corresponds to the predicted tumor cell number using a modelthat does not include the immune response [54], “Leukemia” corresponds to the results of another model that takes the immune responseinto account [44]. The “T cells” curve is obtained with the model by Kim et al. and is compared to experimentally observed T-cell numbersfrom different patients [53].

After the validity of a model is proven using experimentaldata, it can be used to predict a clinical outcome, to improvea certain treatment, or to elucidate unknown mechanisms.Through collaboration with wet-lab researchers, it is possibleto achieve an optimal interplay in which both sides profitfrom the knowledge gained from the analyzed system. Chenget al. [64] describe the fruitful cooperation between hisgroup and a collaborating wet-lab group, which gainedimportant insights into the dynamics of the memory T-cell responses under sequences of heterologous viral infec-tions. The mathematical model (IMMSIM) not only couldsimulate the biological findings but also could predict theexperimental outcome correctly. From this collaboration, it

could be shown that long-term memory loss is accounted forby active attrition by virus-induced type 1 interferon and notby the competition between memory cells.

6. Advantages and Disadvantages ofthe Modeling Types

The advantages of theoretical models over experimentalwork and clinical studies are obvious, mathematical andcomputational techniques are by far less expensive, lesstime consuming, and it is possible to change environmentalinfluences and parameter scales easily.


As presented, both continuous differential equations sys-tems and discrete ABMs or CA have been applied successfullyto tumor immunology, each having its own advantages anddisadvantages.

Differential equations are easier to analyze, parametersensitivity is measurable, and the solution space can bedetermined. It is also more straightforward to adjust globalparameters and fit the model to experimental data usingdifferential equations. On the other hand, differential equa-tions systems are mainly limited to a specific observablephenomenon, and it is nearly impossible to capture thewhole complexity of a biological system. By contrast, rule-based models can deal with a lot of different entities andcan be easily extended with new insights from experimentalresearch. In addition, rule-based models reproduce complexpatterns from simple behavioral rules. However, rule-basedmodels are difficult to analyze in terms of parameter sensitiv-ity and solution space. Furthermore, most rule-based modelsare not completely deterministic, but they include stochasticelements which complicate the analysis additionally. Thesame holds true for the computational efficiency; differentialequations systems are not too computationally demanding,while rule-based models might be limited by computationalcapacity.

The great advantage of rule-based models is their capa-bility of distinguishing every single cell or molecule in itslocation, developmental state, and specificity. Using differ-ential equations, one is limited to homogeneous populationsthat might not correctly represent immune cells with theirspecific receptors.

After all, the choice of the modeling technique alwaysdepends on the question of interest. If the advantages of bothmodeling approaches are desired, the newly emerging hybridmodels might be favored. In hybrid models, the underly-ing architecture of an ABM is extended with differentialequations to simulate continuous parts of the system, asfor example the interaction strength between two cells withmatching receptors.

7. Outlook

Theoretical models are established tools in medical scienceand support experimental work in various ways. Their majordrawback is that models can only be as good as the dataor the theory they are based on, and every result has tobe verified experimentally. Several parameters have to beestimated as they are not known or are not even accessiblefrom experiments. Theoretical models can focus on mainmechanisms leaving out perturbing environmental effects.Thus, they can be used to support or to contradict a theoryor to search for optimal conditions for a desired outcome.One application is to improve individualized medicine asparameters of a theoretical model can be easily adjusted tothe requirements of a specific patient, and the individualoptimal treatment schedule can be gained.

Certainly, theoretical models and applied immunologywill grow hand in hand, as experimental data is needed toestablish theoretical models, and the results from simulationcan help in a more efficient design of experiments.

Acknowledgments

This study was supported by the Deutsche Forschungsge-meinschaft (SFB 449) and the BMBF project MedSys. Theopen access charge was funded by the project DFG SFB 449.We would like to thank Catherine Worth for correcting themanuscript and for her helpful suggestions.

References

[1] H. Kitano, “Systems biology: a brief overview,” Science, vol.295, no. 5560, pp. 1662–1664, 2002.

[2] J. Synnergren, B. Olsson, and J. Gamalielsson, “Classificationof information fusion methods in systems biology,” In SilicoBiology, vol. 9, no. 3, pp. 65–76, 2009.

[3] H. Kitano, “Computational systems biology,” Nature, vol. 420,no. 6912, pp. 206–210, 2002.

[4] A. Mantovani, P. Allavena, A. Sica, and F. Balkwill, “Cancer-related inflammation,” Nature, vol. 454, no. 7203, pp. 436–444,2008.

[5] K. E. de Visser, A. Eichten, and L. M. Coussens, “Paradoxicalroles of the immune system during cancer development,”Nature Reviews Cancer, vol. 6, no. 1, pp. 24–37, 2006.

[6] S. A. Rosenberg, “Progress in human tumour immunologyand immunotherapy,” Nature, vol. 411, no. 6835, pp. 380–384,2001.

[7] S. Kim-Schulze, B. Taback, and H. L. Kaufman, “Cytokinetherapy for cancer,” Surgical Oncology Clinics of North America,vol. 16, no. 4, pp. 793–818, 2007.

[8] C. R. Parish, “Cancer immunotherapy: the past, the presentand the future,” Immunology and Cell Biology, vol. 81, no. 2,pp. 106–113, 2003.

[9] S. Feyerabend, S. Stevanovic, C. Gouttefangeas et al., “Novelmulti-peptide vaccination in Hla-A2+ hormone sensitivepatients with biochemical relapse of prostate cancer,” Prostate,vol. 69, no. 9, pp. 917–927, 2009.

[10] H. Van Poppel, S. Joniau, and S. W. Van Gool, “Vaccinetherapy in patients with renal cell carcinoma,” EuropeanUrology, vol. 55, no. 6, pp. 1333–1344, 2009.

[11] L. M. Weiner, R. Surana, and S. Wang, “Monoclonal anti-bodies: versatile platforms for cancer immunotherapy,” NatureReviews Immunology, vol. 10, no. 5, pp. 317–327, 2010.

[12] S. Koido, E. Hara, S. Homma et al., “Cancer vaccine by fusionsof dendritic and cancer cells,” Clinical and DevelopmentalImmunology, vol. 2009, Article ID 657369, 2009.

[13] C. A. Kruse, L. Cepeda, B. Owens, S. D. Johnson, J. Stears, andK. O. Lillehei, “Treatment of recurrent glioma with intracav-itary alloreactive cytotoxic T lymphocytes and interleukin-2,”Cancer Immunology Immunotherapy, vol. 45, no. 2, pp. 77–87,1997.

[14] A. Lin, A. Schildknecht, L. T. Nguyen, and P. S. Ohashi, “Den-dritic cells integrate signals from the tumor microenviron-ment to modulate immunity and tumor growth,” ImmunologyLetters, vol. 127, no. 2, pp. 77–84, 2010.

[15] Z. Kirkali and E. Tuzel, “Systemic therapy of kidney cancer:tyrosine kinase inhibitors antiagiogenesis or IL-2?” FutureOncology, vol. 5, no. 6, pp. 871–888, 2009.

[16] D. F. McDermott, “Immunotherapy of metastatic renal cellcarcinoma,” Cancer, vol. 115, no. 10, pp. 2298–2305, 2009.

[17] M. Kanehisa, S. Goto, M. Furumichi, M. Tanabe, and M.Hirakawa, “KEGG for representation and analysis of molec-ular networks involving diseases and drugs,” Nucleic AcidsResearch, vol. 38, supplement 1, pp. D355–D360, 2009.


[18] L. Matthews, G. Gopinath, M. Gillespie et al., “Reactomeknowledgebase of human biological pathways and processes,”Nucleic Acids Research, vol. 37, no. 1, pp. D619–D622, 2009.

[19] R. Vita, L. Zarebski, J. A. Greenbaum et al., “The immune epi-tope database 2.0,” Nucleic Acids Research, vol. 38, supplement1, pp. D854–D862, 2009.

[20] M. Ashyraliyev, Y. Fomekong-Nanfack, J. A. Kaandorp, andJ. G. Blom, “Systems biology: parameter estimation forbiochemical models,” The FEBS Journal, vol. 276, no. 4, pp.886–902, 2009.

[21] C. P. Toseland, D. J. Clayton, H. McSparron et al., “AntiJen:a quantitative immunology database integrating functional,thermodynamic, kinetic, biophysical, and cellular data,”Immunome Research, vol. 1, no. 1, p. 4, 2005.

[22] D. J. Lynn, G. L. Winsor, C. Chan et al., “InnateDB: facilitatingsystems-level analyses of the mammalian innate immuneresponse,” Molecular Systems Biology, vol. 4, article 218, 2008.

[23] T. Lipniacki, P. Paszek, A. R. Brasier, B. Luxon, and M. Kimmel,“Mathematical model of NF-κB regulatory module,” Journal ofTheoretical Biology, vol. 228, no. 2, pp. 195–215, 2004.

[24] T. Lipniacki, K. Puszynski, P. Paszek, A. R. Brasier, and M.Kimmel, “Single TNFα trimers mediating NF-κB activation:stochastic robustness of NF-κB signaling,” BMC Bioinformat-ics, vol. 8, article 376, 2007.

[25] S. Tay, J. J. Hughey, T. K. Lee, T. Lipniacki, S. R. Quake, and M.W. Covert, “Single-cell NF-B dynamics reveal digital activationand analogue information processing,” Nature, vol. 466, no.7303, pp. 267–271, 2010.

[26] V. Helms, Principles of Computational Cell Biology, WILEY-VCH, Weinheim, Germany, 2008.

[27] L. G. de Pillis, A. E. Radunskaya, and C. L. Wiseman,“A validated mathematical model of cell-mediated immuneresponse to tumor growth,” Cancer Research, vol. 65, no. 17,pp. 7950–7958, 2005.

[28] A. Diefenbach, E. R. Jensen, A. M. Jamieson, and D. H. Raulet,“Rae1 and H60 ligands of the NKG2D receptor stimulatetumour immunity,” Nature, vol. 413, no. 6852, pp. 165–171,2001.

[29] V. A. Kuznetsov, I. A. Makalkin, M. A. Taylor, and A.S. Perelson, “Nonlinear dynamics of immunogenic tumors:parameter estimation and global bifurcation analysis,” Bulletinof Mathematical Biology, vol. 56, no. 2, pp. 295–321, 1994.

[30] K. Leon, K. Garcia, J. Carneiro, and A. Lage, “How regulatoryCD25+CD4+T cells impinge on tumor immunobiology? Onthe existence of two alternative dynamical classes of tumors,”Journal of Theoretical Biology, vol. 247, no. 1, pp. 122–137,2007.

[31] A. Cappuccio, M. Elishmereni, and Z. Agur, “Cancerimmunotherapy by interleukin-21: potential treatment strate-gies evaluated in a mathematical model,” Cancer Research, vol.66, no. 14, pp. 7293–7300, 2006.

[32] S. Bunimovich-Mendrazitsky, H. Byrne, and L. Stone, “Math-ematical model of pulsed immunotherapy for superficialbladder cancer,” Bulletin of Mathematical Biology, vol. 70, no.7, pp. 2055–2076, 2008.

[33] S. Bunimovich-Mendrazitsky, E. Shochat, and L. Stone,“Mathematical model of BCG immunotherapy in superficialbladder cancer,” Bulletin of Mathematical Biology, vol. 69, no.6, pp. 1847–1870, 2007.

[34] N. Kronik, Y. Kogan, V. Vainstein, and Z. Agur, “Improvingalloreactive CTL immunotherapy for malignant gliomas usinga simulation model of their interactive dynamics,” CancerImmunology, Immunotherapy, vol. 57, no. 3, pp. 425–439,2008.

[35] L. M. Wein, J. T. Wu, and D. H. Kirn, “Validation and analysisof a mathematical model of a replication-competent oncolyticvirus for cancer treatment: implications for virus design anddelivery,” Cancer Research, vol. 63, no. 6, pp. 1317–1324, 2003.

[36] Y. Tao and Q. Guo, “The competitive dynamics between tumorcells, a replication-competent virus and an immune response,”Journal of Mathematical Biology, vol. 51, no. 1, pp. 37–74, 2005.

[37] S. D. Webb, J. A. Sherratt, and R. G. Fish, “Cells behavingbadly: a theoretical model for the Fas/FasL system in tumourimmunology,” Mathematical Biosciences, vol. 179, no. 2, pp.113–129, 2002.

[38] H. Kacser and J. A. Burns, “The control of flux,” Symposia ofthe Society for Experimental Biology, vol. 27, pp. 65–104, 1973.

[39] R. Heinrich and T. A. Rapoport, “A linear steady statetreatment of enzymatic chains: general properties, control andeffector strength,” European Journal of Biochemistry, vol. 42,no. 1, pp. 89–95, 1974.

[40] L. G. de Pillis et al., “Chemotherapy for tumors: an analysisof the dynamics and a study of quadratic and linear optimalcontrols,” Mathematical Biosciences, vol. 209, pp. 292–315,2007.

[41] L. G. de Pillis, W. Gu, and A. E. Radunskaya, “Mixedimmunotherapy and chemotherapy of tumors: modeling,applications and biological interpretations,” Journal of Theo-retical Biology, vol. 238, pp. 841–862, 2006.

[42] F. Castiglione and B. Piccoli, “Optimal control in a model ofdendritic cell transfection cancer immunotherapy,” Bulletin ofMathematical Biology, vol. 68, no. 2, pp. 255–274, 2006.

[43] F. Castiglione and B. Piccoli, “Cancer immunotherapy, math-ematical modeling and optimal control,” Journal of TheoreticalBiology, vol. 247, no. 4, pp. 723–732, 2007.

[44] P. S. Kim, P. P. Lee, D. Levy et al., “Dynamics and potentialimpact of the immune response to chronic myelogenousleukemia,” PLoS Computational Biology, vol. 4, no. 6, ArticleID e1000095, 2008.

[45] R. DeConde, P. S. Kim, D. Levy, and P. P. Lee, “Post-transplantation dynamics of the immune response to chronicmyelogenous leukemia,” Journal of Theoretical Biology, vol.236, no. 1, pp. 39–59, 2005.

[46] A. Matzavinos, M. A. J. Chaplain, and V. A. Kuznetsov,“Mathematical modelling of the spatio-temporal response ofcytotoxic T-lymphocytes to a solid tumour,” MathematicalMedicine and Biology, vol. 21, no. 1, pp. 1–34, 2004.

[47] S. Eikenberry, C. Thalhauser, and Y. Kuang, “Tumor-immuneinteraction, surgical treatment, and cancer recurrence ina mathematical model of melanoma,” PLoS ComputationalBiology, vol. 5, no. 4, Article ID e1000362, 2009.

[48] L. Gray, “A mathematician looks at Wolfram’s new kind ofscience,” Notices of the American Mathematical Society, vol. 50,no. 2, pp. 200–211, 2003.

[49] http://www.cs.unm.edu/∼forrest/software/cancersim/.

[50] M. Kaufman, J. Urbain, and R. Thomas, “Towards a logicalanalysis of the immune response,” Journal of TheoreticalBiology, vol. 114, no. 4, pp. 527–561, 1985.

[51] D. Chowdhury, M. Sahimi, and D. Stauffer, “A discretemodel for immune surveillance, tumor immunity and cancer,”Journal of Theoretical Biology, vol. 152, no. 2, pp. 263–270,1991.

[52] D. G. Mallet and L. G. De Pillis, “A cellular automata modelof tumor-immune system interactions,” Journal of TheoreticalBiology, vol. 239, no. 3, pp. 334–350, 2006.


[53] E. M. Higham, C.-H. Shen, K. D. Wittrup, and J. Chen,“Cutting edge: delay and reversal of T cell tolerance byintratumoral injection of antigen-loaded dendritic cells in anautochthonous tumor model,” Journal of Immunology, vol.184, no. 11, pp. 5954–5958, 2010.

[54] F. Michor, T. P. Hughes, Y. Iwasa et al., “Dynamics of chronicmyeloid leukaemia,” Nature, vol. 435, no. 7046, pp. 1267–1270, 2005.

[55] M. Dougan and G. Dranoff, “Immune therapy for cancer,”Annual Review of Immunology, vol. 27, pp. 83–117, 2009.

[56] R. R. Jenq and M. R. M. van den Brink, “Allogeneichaematopoietic stem cell transplantation: individualized stemcell and immune therapy of cancer,” Nature Reviews Cancer,vol. 10, no. 3, pp. 213–221, 2010.

[57] F. Celada and P. E. Seiden, “A computer model of cellularinteractions in the immune system,” Immunology Today, vol.13, no. 2, pp. 56–62, 1992.

[58] F. Castiglione, F. Toschi, M. Bernaschi et al., “Computationalmodeling of the immune response to tumor antigens,” Journalof Theoretical Biology, vol. 237, no. 4, pp. 390–400, 2005.

[59] S. Motta, F. Castiglione, P. Lollini, and F. Pappalardo, “Mod-elling vaccination schedules for a cancer immunopreventionvaccine,” Immunome Research, vol. 1, no. 1, p. 5, 2005.

[60] F. Pappalardo, P.-L. Lollini, F. Castiglione, and S. Motta,“Modeling and simulation of cancer immunopreventionvaccine,” Bioinformatics, vol. 21, no. 12, pp. 2891–2897, 2005.

[61] P.-L. Lollini, S. Motta, and F. Pappalardo, “Discovery of cancervaccination protocols with a genetic algorithm driving anagent based simulator,” BMC Bioinformatics, vol. 7, article 352,2006.

[62] F. Pappalardo, M. Pennisi, F. Castiglione, and S. Motta, “Vac-cine protocols optimization: in silico experiences,” Biotechnol-ogy Advances, vol. 28, no. 1, pp. 82–93, 2010.

[63] L. R. Paiva, C. Binny, S. C. Ferreira Jr., and M. L. Martins,“A multiscale mathematical model for oncolytic virotherapy,”Cancer Research, vol. 69, no. 3, pp. 1205–1211, 2009.

[64] Y. Cheng, D. Ghersi, C. Calcagno, L. K. Selin, R. Puzone,and F. Celada, “A discrete computer model of the immunesystem reveals competitive interactions between the humoraland cellular branch and between cross-reacting memory andnaıve responses,” Vaccine, vol. 27, no. 6, pp. 833–845, 2009.


Review Article

Prognostic and Diagnostic Value of SpontaneousTumor-Related Antibodies

Sebastian Kobold, Tim Luetkens, Yanran Cao, Carsten Bokemeyer,and Djordje Atanackovic

Department of Oncology, Hematology, Stem Cell Transplantation with the section Pneumology, University Cancer Center Hamburg(Hubertus Wald Tumorzentrum), University Medical Center Hamburg-Eppendorf, 20246 Hamburg, Germany

Correspondence should be addressed to Sebastian Kobold, [email protected]

Received 30 June 2010; Revised 11 October 2010; Accepted 29 November 2010

Academic Editor: Richard L. Gallo

Copyright © 2010 Sebastian Kobold et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

There is an urgent need for earlier diagnosis of malignancies and more stringent monitoring of relapses after antitumortherapy. In addition, new prognostic markers are needed for risk stratification and design of individualized cancer therapies.New diagnostic and prognostic parameters should overcome the impairments of current standards in a cost-effective manner.Serological approaches measuring spontaneous antibody responses against tumor-associated antigens could be of use as diagnosticand prognostic markers and could also be employed to evaluate response to therapy in cancer patients. Autoantibodies havebeen suggested to be of frequent and specific occurrence in patients with malignancies and to correlate with clinical parameters.Screening the relevant literature on this topic, we suggest that the analysis of single antibody specificities is unlikely to providesufficient diagnostic and prognostic accuracy. The combined analysis of autoantibodies targeting different antigens, however, mayreach high sensitivity and specificity. In addition, screening cancer patients for autoantibodies might identify subgroups with highrelapse risk and a worse prognosis. Larger prospective trials should be initiated to identify sets of tumor-associated autoantibodiessuited for the use in diagnostic algorithms for cancer detection and followup.

1. Introduction

For close to 150 years, human malignancies and the immunesystem have been suspected to be interaction partners [1].While data supporting this relationship has accumulatedin recent years, the exact biological role of spontaneouslyoccurring anti-tumor immune responses is still a matter ofcontroversy [2, 3]. In any case, the characterization of thecrosstalk between tumors and their immune environmenthas led to a systematic analysis of the antibody repertoire ofcancer patients [4]. The relatively high frequency of spon-taneous antibody responses against cancer-related antigensled to the assumption that these antibodies could be ofuse in the clinical setting [5]. Accordingly, a lot of effortwas invested in correlating the presence of such antibodieswith clinical parameters to assess their use as prognosticparameters. Furthermore, the highly cancer-specific natureof some of these antibodies resulted in the evaluation of

their diagnostic utility [6]. Both approaches seemed verypromising as a serological detection of cancer, and a serologicrisk stratification would be easy to handle, of low cost, andmuch more likely to be accepted by a wide majority ofpatients hesitant to undergo invasive procedures [7].

Nevertheless, the initial euphoria was dampened bycontroversial results regarding the prognostic reliability oftumor-associated autoantibodies throughout different can-cers [8]. Autoantibodies were either reported to improve theprognosis of cancer patients, to worsen the clinical outcome,or even to be irrelevant for the course of the disease [9]. Froma diagnostic point of view, the results did not meet the highexpectations perhaps as the analysis of single autoantibodiesproved to be of insufficient sensitivity for clinical routine [8].

Very recently, the idea that tumor-associated autoanti-bodies could be developed into meaningful diagnostic andprognostic tools has been revived [10, 11] as researchersaimed at increasing the sensitivity of serological assays


by combining several autoantibodies [12]. In the presentpaper, we will try to answer the question whether and howautoantibodies could be used to enhance early diagnosisof malignant conditions and how they might contribute toperform appropriate risk stratifications in these patients.

2. Serological Analyses in Cancer Patients

2.1. Tumor-Associated Autoantibodies against Single AntigensLack Sensitivity to Reach Diagnostic Relevance. Since tumor-associated autologous antibodies have first been observed, ithas been investigated whether they could be used as an earlydisease marker in a minimally invasive diagnostic approach[6]. In order to be applicable as diagnostic markers, tumor-associated autoantibodies should only be present in cancerpatients, they should be detectable in as many patients aspossible, and they should ideally appear early in the courseof the disease.

Choosing an appropriate antigen is a difficult task inlight of the overwhelming amount of antigens elicitingautoantibodies in cancer patients. The Cancer ImmunomeDatabase [13] currently lists 2,743 sequences for 2,316clones, and this number is constantly growing. However,most antigens are unsuitable for diagnostic purposes becausethey are too low-titered, occur only in a subgroup of cancerpatients, and/or are also found in healthy subjects or patientswith benign diseases [14].

We screened all available studies evaluating autoantibod-ies as possible diagnostic parameters in cancer patients inthe pubmed database. Autoantibodies had to be investigatedin at least five studies in order to be included into the finalanalysis. We proposed three quality criteria for the 9 antibodyspecificities which were analyzed for their diagnostic value.At least two of these criteria had to be fulfilled by a certainantibody in order to qualify as a promising candidate fordiagnostic purposes. Serological responses (1) had to behigh-titered (defined as a titer above 1:1,000), (2) had tobe detectable in at least 40% of patients with a respectivemalignancy, and (3) had to be absent from the peripheralblood of healthy subjects. Only three types of antibodiesfulfilled our quality criteria: polymorphic epithelial mucin(MUC-1), p53, and Sry-like high mobility group superfamily (SOX).

MUC-1 is overexpressed by breast, colorectal, and lungcancer and is deficiently glycosylated on the cell surface[15]. As a surface antigen, the immunogenicity of MUC-1was of particular interest and has been extensively studied[16, 17]. Despite the fact that high-titered MUC-1-specificautoantibodies are frequently found in cancer patients, theiruse as single diagnostic parameters is hampered by thecircumstance that they can be detected at an almost similarfrequency in healthy donors [18].

In contrast, in a study examining patients with lungcancer, p53-specific autoantibodies could be observed yearsprior to detection of malignant lesions [8]. All patientsdeveloped cancer after first detection of anti-p53 autoan-tibodies and the serological response remained detectablethroughout the remaining course of the disease [8] In lightof such data, p53-specific autoantibodies would seem to

be very well suited for diagnostic purposes [19, 20] and,accordingly, anti-p53 autoantibodies have been extensivelystudied regarding their diagnostic and prognostic applicabil-ity for a wide variety of malignancies [21–23]. Unfortunately,one conclusion from these investigations is that screeningfor anti-p53 antibodies—at least when used as a singleantigen—is of insufficient sensitivity, which is indicated bya maximal cancer detection rate of 46% [24]. In addition,p53 antibodies are not restricted to a single cancer type [23]and the observation that anti-p53 antibodies are detectablein subgroups of patients with inflammatory diseases whonever develop a malignancy represents another importantdownside of the use of p53 serology for diagnostic purposes.

Antibodies against SOX families B1 and B2 are found[25] in up to 43% of patients with small cell lung cancer(SCLC) [10]. Unfortunately, the presence of SOX-specificantibodies in patients with benign diseases [10, 26] resultsin an insufficient specificity of SOX when used as a singleantibody. Anti-SOX antibodies have recently been suggestedto help differentiating between patients with paraneoplastic(cancer associated) and sporadic Lambert-Eaton syndrome,respectively [10]. On the other hand, many tumor patientswill evidence SOX antibodies without any hint for neuro-logical symptoms [25]. Additional well-designed studies areneeded before the question whether anti-SOX antibodies areusable for diagnostic purposes can be resolved.

A very recent meta-analysis [27] has concluded that EBV-specific antibodies are of diagnostic value for the diagnosis ofnasopharyngeal carcinoma. Specificity and sensitivity weresuggested to be extremely high for a single antigen (over90%). However, these retrospective analyses still need to bevalidated in a prospective manner. At this time, one hasto conclude that antibodies with a single specificity cannotreliably be used for cancer screening and early detection.

2.2. The Combination of Tumor Antigen-Specific Autoantibod-ies May Reach Diagnostic Relevance. As a single cancer typepotentially elicits antibody responses against a large numberof antigens, another approach is to combine several types ofantibody responses in a diagnostic algorithm. Such a strategyshould allow for increased sensitivity without significantlycomplicating the diagnostic procedure. However, as of yetonly a small number of studies has addressed this issue.We will herein only discuss combinations of antibodyspecificities (1) showing a specificity of at least 75% beinginvestigated in studies (2) including at least 100 patients.Most published studies did not fulfill these criteria (Table 1).Most of these studies used samples of patients in whom thediagnosis of cancer had already been established.

It has recently been shown that up to 75% of allbreast cancer patients bear antibodies against at least oneof the following antigens: p53, c-myc, NY-ESO-1, BRCA1,BRCA2, HER2, or MUC1. With a combined analysis ofthese antibodies, specificity for breast cancer increased to85% while sensitivity remained relatively low with 64% [17].On the other hand, the combined analysis of antibodiesto p53, HER-2 (Human Epidermal growth factor Receptor-2), IGFBP-2 (Insulin like Growth Factor Binding Protein2), and TOPO2α (Topoisomerase-2-alpha) increased both


Table 1: Combinations of autoantibodies evaluated as diagnostic measures.

Antibody specificity Number of patients Tumor type Specificity/Sensitivitiy Ref.

PIM1, MAPKAPK3, ACVR2B 114 Colon cancer 74/83 [28]

ASB-9, SERAC1, RELT 87 Breast cancer 100/80 [29]

PPIA, PRDX2, FKBP52, MUC-1, HSP60 142 Breast cancer 73/85 [30]

Calnuc, p53, Cyclin D1, Cyclin B1, myc 447 Mixed 59/62 [31]

myc, p53, Cyclin B1, p62, Koc, IMP1, Survivin 976 Mixed 50/91 [22]

CCCAP, HDAC5, p53, NMDAR, NY-CO-16 94 Colon cancer 59/78 [32]

SEREX clones (N = 6) 48 Colon cancer 96/92 [33]

SEREX clones (N = 10) 77 HCC 87/66 [34]

Imp1, p62, Koc, p53, myc, Cyclin B1, Survivin, p16 142 HCC 50/60 [35]

SEREX clones (N = 80) 39 Head and neck cancer 90/80 [36]

O-glycopeptides 56 Mixed 85/25 [37]

Nucleophosmin, Cathepsin D, p53, SSX 125 Ovarian cancer NN [38]

Autoantibodies to MIAPACA cell line 238 Pancreatic cancer 80/93 [39]

Phage display clones (N = 22) 119 Prostate cancer 88/82 [40]

p53, c-myc, HER2, NY-ESO-1, BRCA1, BRCA2, MUC1 97 Breast cancer 85/64 [17]

p53, NY-ESO-1, CAGE; GBU4-5, Annexin 1 626 Lung cancer 90/40 [41]

diagnostic specificity and sensitivity to up to 75% for breastcancer patients [12].

Combined serological approaches have also been appliedto other malignancies. Using lysates of a human pancreaticcarcinoma cell line as a mixture of unknown target antigens,antibody responses were detected in over 90% of patientswith pancreatic cancer with a specificity of 80% [39, 42]. Inanother study 22 phage-displayed tumor-associated antigenswere combined to diagnose prostate cancer in a serologicalanalysis. Sensitivity of this approach was remarkably highwith 88% while still detecting 82% of true-positive patients[40]. Very recently, a large study including over 600 patientswith lung cancer and utilizing a combination of five antigensreached a remarkable 90% specificity, while sensitivityremained relatively low with 40% [41].

In conclusion, the combined analysis of different tumor-related autoantibodies might represent a promising approachfor the diagnosis of cancer leading to an up to threefoldincrease in specificity and sensitivity (Table 1). However,despite the general scalability of this approach, increasingcosts and the methodological complexicity of large antibodyscreens become important considerations. To be applicablein the clinical setting, standardized and reliable methodsfor the performance of serological concepts are needed.Unfortunately, a large variety of different methods forantibody detection has been used worldwide and results fromthese analyses are often difficult to compare [43]. Recently,a semiautomatic ELISA assay has been proposed to possesshigh specificity, reproducibility, and reliability [41]. Thebroad introduction of comparable methods probably repre-sents a prerequisite for the future use of serological analysesof antibodies in clinical routine. Comprehensive prospectivestudies are needed to provide a definitive rationale for theirapplication in the clinical routine.

2.3. Tumor-Related Autoantibodies Are Predictors of ClinicalOutcome in Cancer Patients. In spite of significant progressin the diagnosis and followup of cancer patients, it isstill difficult to stratify patients according to their risk ofrelapse [11]. Such an assessment would be of great use toboth patient and clinician, as it would help to differentiatebetween those who might benefit from additional therapyand those who may only suffer from therapy-related sideeffects without achieving an improved outcome. An idealstratification technique would also be of low cost and causeminimal discomfort for the patient. Obviously, a serologicalapproach can fulfill both of these criteria and add furtherdiagnostic aspects not covered by current stratifications.Many studies have examined the correlation between theoccurrence of tumor-related autoantibodies and the clinicaloutcome of cancer patients (Table 2). When we screened theliterature, we focused on three types of target antigens: (1)shared cancer-specific antigens, (2) antigens overexpressedin tumors compared to normal tissues, and (3) mutatedantigens. Overall, among the studies dealing with theprognostic impact of autoantibodies in cancer patients, 50%were unable to demonstrate an impact on the course of thedisease, 33.33% did show a worse outcome, and 16.67%involved a better prognosis.

A large body of studies particularly involving SEREX(Serological Identification of antigens by recombinantexpression cloning) identified antibodies that did not revealany significant impact on clinical outcome. Even antibodiesagainst antigens that are involved in the pathogenesis of thedisease (Cyclin B1 or p53 in oral cancer) or associated witha more aggressive course did not seem to affect the outcome(Table 2).

Antigen p53 probably represents the tumor-associatedprotein that has most extensively been studied with regard


Table 2: Influence of autoantibodies against tumor-associated antigens on the prognosis of cancer patients.

Autoantibody specificity Number of patients Tumor type Impact Ref.

Laminin 71 Breast cancer Decreased OS [44]

HSP90 327 Breast cancer Decreased OS [45]

p53 9489 Breast, gastric, colon cancer, NSCLC and oral cancer Decreased OS [46]

NY-ESO-1 207 Prostate cancer Decreased OS [47]

Nucleophosmin 100 Breast cancer Decreased RFS [48]

Panel of 29 antigens 60/59 Ovarian cancer/pancreatic cancer Increased OS [49]

CTSP-1 147 Prostate cancer Increased RFS [50]

p53 130 HCC Increased RFS [51]

Laminin-Receptor 67 CLL Increased RFS [52]

CML66 15 CML Increased RFS [53]

GLEA3 and PHF3 62 Glioblastoma Increased OS [54]

MUC1 30 NSCLC Increased RFS [18]

MUC1 100 Ovarian cancer Increased OS [55]

MUC5AC 30 Colon cancer Increased OS [56]

SOX1 90 SCLC Increased OS [57]

CEA 52 Breast cancer Increased RFS [58]

SEREX clones 12 Meningioma None [59]

SCP1 100 Pancreatic cancer None [60]

Cyclin B1 42 AML None [61]

p53 120 Oral cancer None [6]

Hsp90 116 Ovarian cancer None [62]

ALK 21 ALL None [63]

SEREX clones 25 SCC None [64]

MUC-1 125 Breast cancer None [65]

Survivin 76 NSCLC None [66]

NY-ESO-1 12 Different cancers None [67]

NY-ESO-1 69 Esophageal cancer None [68]

Phage Display clones 176 Breast cancer None [69]

Braf 372 Melanoma None [70]

OS: overall survival, RFS: recurrence-free survival.

to its prognostic value. Several small studies yielded variableresults, ranging between a missing effect and a negativeinfluence on the patients’ outcome [71]. Interestingly, onelarger study performed in hepatocellular carcinoma patientssuggested that the presence of p53-specific antibodies mightbe associated with an increased overall survival [72]. How-ever, a number of other large studies in breast, lung,colon and oral cancer patients as well as a meta-analysisclearly highlighted the correlation between the presenceof p53-specific autoantibodies and decreased overall andprogression-free survival [19, 73].

Another well-studied example of antibody responsesassociated with a poor prognosis is found among CancerTestis Antigens (CTA). CTA, also called shared cancer-specific antigens, represent a unique class of tumor proteinswith an expression restricted to normal testis and cancertissue. Interestingly, antibodies against some CTA, such asNY-ESO-1, were primarily found in patients with advanceddisease [74]. Levels of NY-ESO-1-specific antibody responsesseemed to correlate with the volume of tumor tissue present

in the patient’s body [67]. Accordingly, spontaneous occur-rence of NY-ESO-1 antibodies has particularly been foundin multiple myeloma patients and melanoma patients whorelapsed and/or suffered from progressive disease [67, 75].Moreover, in prostate cancer, the presence of anti-NY-ESO-1 antibodies was associated with a decreased overall survivalwhile no influence of NY-ESO-1 seropositivity was observedin patients with esophageal cancer [47, 68]. These findingsmight, on the one hand, suggest that NY-ESO-1-specific anti-bodies merely represent markers of an increased tumor loadand do not indicate a biologically relevant immune response.Conversely, NY-ESO-1-specific antibodies might in fact exerta relevant anti-tumor effect by directly and/or indirectlysuppressing tumor growth in earlier stages of the disease.As the disease progresses, the net effects of these cancer-specific immune responses might decrease and, finally, theywill be overwhelmed by cancer-related immunosuppressionand the kinetics of tumor growth [67]. This hypothesiswould, to some degree, be supported by observations inpatients vaccinated with CTA where response to therapy and


enhanced overall survival have been associated with CTA-specific T cell and antibody responses [76].

On the other hand, a number of studies have suggesteda positive influence of the occurrence of tumor-associatedautoantibodies on the outcome of the patient. In onestudy with NSCLC patients, the presence of anti-MUC-1antibodies was clearly linked to earlier stages of the diseaseas well as a lower recurrence rate [15]. Based on thesefindings and on in vitro data, it was hypothesized thatspontaneously occurring anti-MUC-1 antibodies might bindto the surface of tumor cells uncovering adhesion moleculesmasked by the deficiently glycosylated MUC-1. Accordingto this hypothesis, this would then render MUC-1 specificantibodies a part of an effective immune response againstcancer cells.

Compared to the overall number of studies analyzingautoantibody profiles in cancer patients, relatively fewstudies evaluate the impact of these immune responseson the patients’ prognosis. This observation may be anexpression of publication bias, as negative results are lesslikely to be reported than positive ones [77]. Accordingly,controversial results are found even for a given antigen [19].In contrast, antigens such as p53 or MUC-1 are well studiedin large cohorts and results may be considered reliable.Accordingly, even if most studies suggest for autoantibodiesto have no impact on the course of human malignancies,confirmation of these results is strongly needed. Importantly,no study has yet revealed how such antibodies impact on thepatient’s prognosis, and it is unclear whether autoantibodiesthemselves actively contribute to the patient’s outcome orwhether they are only an indicator of the clinical condition.

As in the case of diagnostic approaches, one couldtheoretically analyze antibody responses against differentantigens in a combined manner in order to enhance theprognostic impact of such assays. A very recent workby Gnjatic et al. has addressed this issue and found thesimultaneous presence of four different autoantibodies todefine a subgroup of patients who showed an improvedprognosis [49]. This pioneering work might lead the waytowards a more efficient prognostic classification of cancerpatients based on serological analyses.

Unfortunately, to date little is known regarding thebehavior of humoral responses against tumor antigens overthe course of the disease. In the vaccination setting antibodyresponses have been routinely monitored over time, butlongitudinal studies on spontaneously occurring antibodyresponses are scarce [78]. The few data available suggestthat, for example, the appearance and persistence of NY-ESO-1-specific antibodies are tightly linked to the amountof antigen present in the patient’s body. Accordingly, diseaseprogression appears to promote the development of NY-ESO-1-specific antibody responses and to lead to increasedantibody titers. Future studies will have to determine in astringent manner the clinical conditions contributing to thedevelopment of autoantibodies in cancer patients.

In any case, the biological basis of either of these obser-vations remains elusive. Therefore, although it is impossibleto design randomized studies in this particular setting,prospective analyses evaluating such antibody responses over

the course of the patient’s disease may help to addresstwo related questions: (1) under which clinical situationsdoes the patient develop antibodies against tumor-associatedantigens and (2) what is the temporal association betweenclinical incidents (i.e., response to therapy or relapse) and thedevelopment of humoral immune responses.

3. Conclusions

The large diversity of the antibody repertoire directed againstautologous tumor-related antigens illustrates the complexand heterogeneous nature of the human seromic repertoire.As summarized in this analysis, the combination of a fewfrequently occurring types of antibodies may be of diagnosticvalue for certain malignancies. The combination approachappears particularly attractive because of the lack of sec-ondary prevention tools for many tumor types. However,prospective studies evaluating the true value of this methodare required prior to clinical introduction.

As primary cancer prevention remains a frustrating issuedue to the multifactorial origins of cancer, early detection ofcancer lesions is an essential goal to enhance the patient’sprognosis and to reduce costs. Current strategies focus onapparative diagnostics and only to a much lesser extendon serological testing [79, 80]. As of yet, only mammog-raphy, colonoscopy, cervical smear, and skin examinationare accepted early cancer detection tools for the respectivecancer types [79, 81–83]. While the reduction in overallmortality provided by these screening methods alone is notwell defined, specificity and the sensitivity reach almost 90%in given studies [79]. In contrast, it is still not completelyclear to what extent approaches, such as the measurement ofprostate-specific antigen in the serum, improve the patient’soutcome [80].

Currently, most prognostic markers used in a routineclinical setting are based on clinical parameters such asperformance status and stage and vary between cancer types[84]. In addition, molecular markers such as the expressionof HER2/NEU have been established to further stratifypatients according to their relapse risk and their need foradditional therapies [85]. It is very likely that, in the nearfuture more sophisticated tools such as gene expressionprofiles will become available [86].

Two major hurdles still hamper the wide spread use ofthe above-mentioned early detection tools in cancer care: (1)the majority of patients fear invasive and time-consumingprocedures and (2) costs of modern diagnostics. Accordingly,there is a need for minimal invasive tools which are cost-effective, and the analysis of cancer-specific antibodies mightprovide an acceptable solution to this clinical dilemma.Such a role would certainly be complementary as serologicalapproaches are unlikely to replace “classical” pathologicalmarkers or staging procedures.

In our opinion, the available data justifies cautiousoptimism regarding the development of serological analysesinto diagnostic and/or prognostic tools. Detailed analysesof the human antibody repertoire will hopefully contributeto clarifying the clinical value of such humoral responses.


Combined data from these studies could vastly improvethe identification and selection of appropriate diagnosticand prognostic markers as well as of future targets forimmunotherapies [87, 88].

References

[1] A. M. Silberstein, History of Immunology, Academic Press, SanDiego, Calif, USA, 1989.

[2] J. D. Bui and R. D. Schreiber, “Cancer immunosurveillance,immunoediting and inflammation: independent or interde-pendent processes?” Current Opinion in Immunology, vol. 19,no. 2, pp. 203–208, 2007.

[3] G. Willimsky and T. Blankenstein, “Sporadic immunogenictumours avoid destruction by inducing T-cell tolerance,”Nature, vol. 437, no. 7055, pp. 141–146, 2005.

[4] R. Bei, L. Masuelli, C. Palumbo, M. Modesti, and A. Modesti,“A common repertoire of autoantibodies is shared by cancerand autoimmune disease patients: inflammation in theirinduction and impact on tumor growth,” Cancer Letters, vol.281, no. 1, pp. 8–23, 2009.

[5] M. M. Black, S. Kerpe, and F. D. Speer, “Lymph node structurein patients with cancer of the breast,” American Journal ofPathology, vol. 29, pp. 505–521, 1953.

[6] R. Ralhan, N. Nath, S. Agarwal, M. Mathur, B. Wasylyk, andN. K. Shukla, “Circulating p53 antibodies as early markers oforal cancer: correlation with p53 alterations,” Clinical CancerResearch, vol. 4, no. 9, pp. 2147–2152, 1998.

[7] H. T. Tan, J. Low, S. G. Lim, and M. C. M. Chung, “Serumautoantibodies as biomarkers for early cancer detection,” FEBSJournal, vol. 276, no. 23, pp. 6880–6904, 2009.

[8] R. Lubin, G. Zalcman, L. Bouchet et al., “Serum p53 antibodiesas early markers of lung cancer,” Nature Medicine, vol. 1, no. 7,pp. 701–702, 1995.

[9] D. G. DeNardo and L. M. Coussens, “Inflammation andbreast cancer. Balancing immune response: crosstalk betweenadaptive and innate immune cells during breast cancerprogression,” Breast Cancer Research, vol. 9, no. 4, p. 212, 2007.

[10] M. J. Titulaer, R. Klooster, M. Potman et al., “SOX antibodiesin small-cell lung cancer and Lambert-Eaton myasthenicsyndrome: frequency and relation with survival,” Journal ofClinical Oncology, vol. 27, no. 26, pp. 4260–4267, 2009.

[11] K. Ait-Tahar, C. Damm-Welk, B. Burkhardt et al., “Correlationof the autoantibody response to the ALK oncoantigen in pedi-atric anaplastic lymphoma kinase-positive anaplastic large celllymphoma with tumor dissemination and relapse risk,” Blood,vol. 115, no. 16, pp. 3314–3319, 2010.

[12] H. Lu, V. Goodell, and M. L. Disis, “Humoral immu-nity directed against tumor-associated antigens as potentialbiomarkers for the early diagnosis of cancer,” Journal ofProteome Research, vol. 7, no. 4, pp. 1388–1394, 2008.

[13] Research LIfC, “Cancer Immunome Database,” 2010.[14] K. D. Preuss, C. Zwick, C. Bormann, F. Neumann, and

M. Pfreundschuh, “Analysis of the B-cell repertoire againstantigens expressed by human neoplasms,” ImmunologicalReviews, vol. 188, pp. 43–50, 2002.

[15] S. Von Mensdorff-Pouilly, M. M. Gourevitch, P. Kenemanset al., “Humoral immune response to polymorphic epithelialmucin (MUC-1) in patients with benign and malignant breasttumours,” European Journal of Cancer A, vol. 32, no. 8, pp.1325–1331, 1996.

[16] H. Nakamura, Y. Hinoda, N. Nakagawa et al., “Detectionof circulating anti-MUC1 mucin core protein antibodies in

patients with colorectal cancer,” Journal of Gastroenterology,vol. 33, no. 3, pp. 354–361, 1998.

[17] C. Chapman, A. Murray, J. Chakrabarti et al., “Autoantibodiesin breast cancer: their use as an aid to early diagnosis,” Annalsof Oncology, vol. 18, no. 5, pp. 868–873, 2007.

[18] Y. Hirasawa, N. Kohno, A. Yokoyama, K. Kondo, K. Hiwada,and M. Miyake, “Natural autoantibody to MUC1 is a prognos-tic indicator for non-small cell lung cancer,” American Journalof Respiratory and Critical Care Medicine, vol. 161, no. 2 I, pp.589–594, 2000.

[19] R. Lubin, B. Schlichtholz, J. L. Teillaud et al., “p53 antibodiesin patients with various types of cancer: assay, identification,and characterization,” Clinical Cancer Research, vol. 1, no. 12,pp. 1463–1469, 1995.

[20] R. Lubin, B. Schlichtholz, D. Bengoufa et al., “Analysis ofp53 antibodies in patients with various cancers define B-cellepitopes of human p53: distribution on primary structure andexposure on protein surface,” Cancer Research, vol. 53, no. 24,pp. 5872–5876, 1993.

[21] R. Ralhan, S. Arora, T. K. Chattopadhyay, N. K. Shukla, andM. Mathur, “Circulating p53 antibodies, p53 gene mutationalprofile and product accumulation in esophageal squamous-cell carcinoma in India,” International Journal of Cancer, vol.85, no. 6, pp. 791–795, 2000.

[22] J. Y. Zhang, C. A. Casiano, X. X. Peng, J. A. Koziol, E. K. L.Chan, and E. M. Tan, “Enhancement of antibody detectionin cancer using panel of recombinant tumor-associated anti-gens,” Cancer Epidemiology Biomarkers and Prevention, vol. 12,no. 2, pp. 136–143, 2003.

[23] B. Abendstein, C. Marth, E. Muller-Holzner, M. Wid-schwendter, G. Daxenbichler, and A. G. Zeimet, “Clinicalsignificance of serum and ascitic p53 autoantibodies inepithelial ovarian carcinoma,” Cancer, vol. 88, no. 6, pp. 1432–1437, 2000.

[24] F. D. Vogl, E. Stickeler, M. Weyermann et al., “p53 autoanti-bodies in patients with primary ovarian cancer are associatedwith higher age, advanced stage and a higher proportion ofp53-positive tumor cells,” Oncology, vol. 57, no. 4, pp. 324–329, 1999.

[25] M. Tschernatsch, O. Gross, N. Kneifel, M. Kaps, and F.Blaes, “SOX-1 autoantibodies in patients with paraneoplasticneurological syndromes,” Autoimmunity Reviews, vol. 8, no. 7,pp. 549–551, 2009.

[26] T. M. E. Davis, Z. Mehta, I. R. Mackay et al., “Autoantibodiesto the islet cell antigen SOX-13 are associated with durationbut not type of diabetes,” Diabetic Medicine, vol. 20, no. 3, pp.198–204, 2003.

[27] S. Li, Y. Deng, XI. Li, Q. P. Chen, X. C. Liao, and X. Qin,“Diagnostic value of Epstein-Barr virus capsid antigen-IgA innasopharyngeal carcinoma: a meta-analysis,” Chinese MedicalJournal, vol. 123, no. 9, pp. 1201–1205, 2010.

[28] I. Babel, R. Barderas, R. Dıaz-Uriarte, J. L. Martınez-Torrecuadrada, M. Sanchez-Carbayo, and J. I. Casal, “Iden-tification of tumor-associated autoantigens for the diagnosisof colorectal cancer in serum using high density proteinmicroarrays,” Molecular and Cellular Proteomics, vol. 8, no. 10,pp. 2382–2395, 2009.

[29] L. Zhong, K. Ge, J. C. Zu et al., “Autoantibodies as potentialbiomarkers for breast cancer,” Breast Cancer Research, vol. 10,no. 3, p. R40, 2008.

[30] C. Desmetz, C. Bascoul-Mollevi, P. Rochaix et al., “Identifica-tion of a new panel of serum autoantibodies associated withthe presence of in situ carcinoma of the breast in younger


women,” Clinical Cancer Research, vol. 15, no. 14, pp. 4733–4741, 2009.

[31] X. Chen, S. Fu, F. Chen, H. Chen, and Z. Chen, “Identificationof tumor-associated antigens in human hepatocellular carci-noma by autoantibodies,” Oncology Reports, vol. 20, no. 4, pp.979–985, 2008.

[32] C. C. Chan, C. W. Fan, Y. B. Kuo et al., “Multiple serologicalbiomarkers for colorectal cancer detection,” InternationalJournal of Cancer, vol. 126, no. 7, pp. 1683–1690, 2010.

[33] Y. Ran, H. Hu, Z. Zhou et al., “Profiling tumor-associatedautoantibodies for the detection of colon cancer,” ClinicalCancer Research, vol. 14, no. 9, pp. 2696–2700, 2008.

[34] Y. Chen, Y. Zhou, S. Qiu et al., “Autoantibodies totumor-associated antigens combined with abnormal alpha-fetoprotein enhance immunodiagnosis of hepatocellular car-cinoma,” Cancer Letters, vol. 289, no. 1, pp. 32–39, 2010.

[35] J. Y. Zhang, R. Megliorino, X. X. Peng, E. M. Tan, Y.Chen, and E. K. L. Chan, “Antibody detection using tumor-associated antigen mini-array in immunodiagnosing humanhepatocellular carcinoma,” Journal of Hepatology, vol. 46, no.1, pp. 107–114, 2007.

[36] HO. S. Lin, H. S. Talwar, A. L. Tarca et al., “Autoantibodyapproach for serum-based detection of head and neck cancer,”Cancer Epidemiology Biomarkers and Prevention, vol. 16, no.11, pp. 2396–2405, 2007.

[37] H. H. Wandall, O. Blixt, M. A. Tarp et al., “Cancerbiomarkers defined by autoantibody signatures to aberrantO-glycopeptide epitopes,” Cancer Research, vol. 70, no. 4, pp.1306–1313, 2010.

[38] D. D. Taylor, C. Gercel-Taylor, and L. P. Parker, “Patient-derived tumor-reactive antibodies as diagnostic markers forovarian cancer,” Gynecologic Oncology, vol. 115, no. 1, pp. 112–120, 2009.

[39] T. H. Patwa, C. Li, L. M. Poisson et al., “The identificationof phosphoglycerate kinase-1 and histone H4 autoantibodiesin pancreatic cancer patient serum using a natural proteinmicroarray,” Electrophoresis, vol. 30, no. 12, pp. 2215–2226,2009.

[40] X. Wang, J. Yu, A. Sreekumar et al., “Autoantibody signaturesin prostate cancer,” New England Journal of Medicine, vol. 353,no. 12, pp. 1224–1235, 2005.

[41] A. Murray, C. J. Chapman, G. Healey et al., “Technicalvalidation of an autoantibody test for lung cancer,” Annals ofOncology, vol. 21, no. 8, pp. 1687–1693, 2010.

[42] H. Honda, S. Kusumoto, K. Nishikawa et al., “Limitation ofCT in diagnosis of pancreatic cancer,” Radiation Medicine -Medical Imaging and Radiation Oncology, vol. 9, no. 2, pp. 61–67, 1991.

[43] U. Fagiolo, A. Amadori, R. Biselli et al., “Quantitative andqualitative analysis of anti-tetanus toxoid antibody responsein the elderly. Humoral immune response enhancement bythymostimulin,” Vaccine, vol. 11, no. 13, pp. 1336–1340, 1993.

[44] M. R. Barbouche, S. Romain, S. A. Avrameas, L. Piana, andP. M. Martin, “Prognostic significance of autoantibodies tolaminin in the sera of breast cancer patients: a preliminaryreport,” European Journal of Clinical Chemistry and ClinicalBiochemistry, vol. 32, no. 7, pp. 511–514, 1994.

[45] S. E. Conroy, G. B. Faulds, W. Williams, D. S. Latchman, andD. A. Isenberg, “Detection of autoantibodies to the 90 kDahead shock protein in systemic lupus erythematosus and otherautoimmune diseases,” British Journal of Rheumatology, vol.33, no. 10, pp. 923–926, 1994.

[46] T. Soussi, “p53 Antibodies in the sera of patients with varioustypes of cancer: a review,” Cancer Research, vol. 60, no. 7, pp.1777–1788, 2000.

[47] A. Fossa, A. Berner, S. D. Fossa, E. Hernes, G. Gaudernack, andE. B. Smeland, “NY-ESO-I protein expression and humoralimmune responses in prostate cancer,” Prostate, vol. 59, no. 4,pp. 440–447, 2004.

[48] B. Brankin, T. C. Skaar, M. Brotzman, B. Trock, and R.Clarke, “Autoantibodies to the nucleolar phosphoproteinnucleophosmin in breast cancer patients,” Cancer Epidemiol-ogy Biomarkers and Prevention, vol. 7, no. 12, pp. 1109–1115,1998.

[49] S. Gnjatic, E. Ritter, M. W. Buchler et al., “Seromic profilingof ovarian and pancreatic cancer,” Proceedings of the NationalAcademy of Sciences of the United States of America, vol. 107,no. 11, pp. 5088–5093, 2010.

[50] R. B. Parmigiani, F. Bettoni, D. M. Grosso et al., “Antibodiesagainst the cancer-testis antigen CTSP-1 are frequently foundin prostate cancer patients and are an independent prognosticfactor for biochemical-recurrence,” International Journal ofCancer, vol. 122, no. 10, pp. 2385–2390, 2008.

[51] R. Saffroy, J. C. Lelong, D. Azoulay et al., “Clinical significanceof circulating anti-p53 antibodies in European patients withhepatocellular carcinoma,” British Journal of Cancer, vol. 79,no. 3-4, pp. 604–610, 1999.

[52] B. Friedrichs, S. Siegel, M. Kloess et al., “Humoral immuneresponses against the immature laminin receptor protein showprognostic significance in patients with chronic lymphocyticleukemia,” Journal of Immunology, vol. 180, no. 9, pp. 6374–6384, 2008.

[53] X. F. Yang, C. J. Wu, S. McLaughlin et al., “CML66, a broadlyimmunogenic tumor antigen, elicits a humoral immuneresponse associated with remission of chronic myelogenousleukemia,” Proceedings of the National Academy of Sciences ofthe United States of America, vol. 98, no. 13, pp. 7492–7497,2001.

[54] C. P. Pallasch, A. K. Struss, A. Munnia et al., “Autoantibodiesagainst GLEA2 and PHF3 in glioblastoma: tumor-associatedautoantibodies correlated with prolonged survival,” Interna-tional Journal of Cancer, vol. 117, no. 3, pp. 456–459, 2005.

[55] E. R. Richards, P. L. Devine, R. J. Quin, J. D. Fontenot, B.G. Ward, and M. A. McGuckin, “Antibodies reactive withthe protein core of MUC1 mucin are present in ovariancancer patients and healthy women,” Cancer ImmunologyImmunotherapy, vol. 46, no. 5, pp. 245–252, 1998.

[56] B. Kocer, J. McKolanis, and A. Soran, “Humoral immuneresponse to MUC5AC in patients with colorectal polyps andcolorectal carcinoma,” BMC Gastroenterology, vol. 6, article 4,2006.

[57] B. Vural, L. C. Chen, P. Saip et al., “Frequency of SOX groupB (SOX1, 2, 3) and ZIC2 antibodies in Turkish patients withsmall cell lung carcinoma and their correlation with clinicalparameters,” Cancer, vol. 103, no. 12, pp. 2575–2583, 2005.

[58] D. Haidopoulos, M. M. Konstadoulakis, P. T. Antonakis et al.,“Circulating anti-CEA antibodies in the sera of patients withbreast cancer,” European Journal of Surgical Oncology, vol. 26,no. 8, pp. 742–746, 2000.

[59] N. Comtesse, A. Zippel, S. Walle et al., “Complex humoralimmune response against a benign tumor: frequent antibodyresponse against specific antigens as diagnostic targets,”Proceedings of the National Academy of Sciences of the UnitedStates of America, vol. 102, no. 27, pp. 9601–9606, 2005.

[60] A. Wadle, B. Kubuschok, J. Imig et al., “Serological immuneresponse to cancer testis antigens in patients with pancreatic


cancer,” International Journal of Cancer, vol. 119, no. 1, pp.117–125, 2006.

[61] E. Ersvaer, J. Y. Zhang, E. McCormack et al., “Cyclin B1 iscommonly expressed in the cytoplasm of primary humanacute myelogenous leukemia cells and serves as a leukemia-associated antigen associated with autoantibody response in asubset of patients,” European Journal of Haematology, vol. 79,no. 3, pp. 210–225, 2007.

[62] C. I. Vidal, P. J. Mintz, K. Lu et al., “An HSP90-mimic peptiderevealed by fingerprinting the pool of antibodies from ovariancancer patients,” Oncogene, vol. 23, no. 55, pp. 8859–8867,2004.

[63] L. Mussolin, P. Bonvini, K. Ait-Tahar et al., “Kinetics ofhumoral response to ALK and its relationship with minimalresidual disease in pediatric ALCL,” Leukemia, vol. 23, no. 2,pp. 400–402, 2009.

[64] I. Diesinger, C. Bauer, N. Brass et al., “Toward a more completerecognition of immunoreactive antigens in squamous cell lungcarcinoma,” International Journal of Cancer, vol. 102, no. 4, pp.372–378, 2002.

[65] B. B. J. Hermsen, R. H. M. Verheijen, F. H. Menko et al.,“Humoral immune responses to MUC1 in women with aBRCA1 or BRCA2 mutation,” European Journal of Cancer, vol.43, no. 10, pp. 1556–1563, 2007.

[66] V. Karanikas, S. Khalil, T. Kerenidi, K. I. Gourgoulianis, andA. E. Germenis, “Anti-survivin antibody responses in lungcancer,” Cancer Letters, vol. 282, no. 2, pp. 159–166, 2009.

[67] E. Jager, E. Srockert, Z. Zidianakis et al., “Humoral immuneresponses of cancer patients against ”cancer-testis” antigenNY-ESO-1: correlation with clinical events,” InternationalJournal of Cancer, vol. 84, no. 5, pp. 506–510, 1999.

[68] A. Akcakanat, T. Kanda, YU. Koyama et al., “NY-ESO-1expression and its serum immunoreactivity in esophagealcancer,” Cancer Chemotherapy and Pharmacology, vol. 54, no.1, pp. 95–100, 2004.

[69] E. Pavoni, A. Pucci, P. Vaccaro et al., “A study of the humoralimmune response of breast cancer patients to a panel ofhuman tumor antigens identified by phage display,” CancerDetection and Prevention, vol. 30, no. 3, pp. 248–256, 2006.

[70] J. Fensterle, J. C. Becker, T. Potapenko et al., “B-Raf specificantibody responses in melanoma patients,” BMC Cancer, vol.4, article 62, 2004.

[71] K. Angelopoulou, M. Stratis, and E. P. Diamandis, “Humoralimmune response against p53 protein in patients with colorec-tal carcinoma,” International Journal of Cancer, vol. 70, no. 1,pp. 46–51, 1997.

[72] P. Tangkijvanich, A. Janchai, N. Charuruks et al., “Clinicalassociations and prognostic significance of serum anti-p53antibodies in Thai patients with hepatocellular carcinoma,”Asian Pacific Journal of Allergy and Immunology, vol. 18, no.4, pp. 237–243, 2000.

[73] R. Tang, M. C. Ko, J. Y. Wang et al., “Humoral response top53 in human colorectal tumors: a prospective study of 1,209patients,” International Journal of Cancer, vol. 94, no. 6, pp.859–863, 2001.

[74] M. Maio, S. Coral, L. Sigalotti et al., “Analysis of cancer/testisantigens in sporadic medullary thyroid carcinoma: expressionand humoral response to NY-ESO-1,” Journal of ClinicalEndocrinology and Metabolism, vol. 88, no. 2, pp. 748–754,2003.

[75] F. Van Rhee, S. M. Szmania, F. Zhan et al., “NY-ESO-1 is highlyexpressed in poor-prognosis multiple myeloma and inducesspontaneous humoral and cellular immune responses,” Blood,vol. 105, no. 10, pp. 3939–3944, 2005.

[76] L. A. DiFronzo, R. K. Gupta, R. Essner et al., “Enhancedhumoral immune response correlates with improved disease-free and overall survival in American Joint Committee on can-cer stage II melanoma patients receiving adjuvant polyvalentvaccine,” Journal of Clinical Oncology, vol. 20, no. 15, pp. 3242–3248, 2002.

[77] K. Dickersin, “The existence of publication bias and riskfactors for its occurrence,” Journal of the American MedicalAssociation, vol. 263, no. 10, pp. 1385–1389, 1990.

[78] D. Valmori, N. E. Souleimanian, V. Tosello et al., “Vaccinationwith NY-ESO-1 protein and CpG in Montanide inducesintegrated antibody/Th1 responses and CD8 T cells throughcross-priming,” Proceedings of the National Academy of Sciencesof the United States of America, vol. 104, no. 21, pp. 8947–8952,2007.

[79] D. Lieberman, “Progress and challenges in colorectal cancerscreening and surveillance,” Gastroenterology, vol. 138, no. 6,pp. 2115–2126, 2010.

[80] M. O’Shaughnessy, B. Konety, and C. Warlick, “Prostate cancerscreening: issues and controversies,” Minnesota medicine, vol.93, no. 8, pp. 39–44, 2010.

[81] M. Kalager, M. Zelen, F. Langmark, and H.-O. Adami, “Effectof screening mammography on breast-cancer mortality inNorway,” New England Journal of Medicine, vol. 363, no. 13,pp. 1203–1210, 2010.

[82] A. C. Budd and C. J. Sturrock, “Cytology and cervical cancersurveillance in an era of human papillomavirus vaccination,”Sexual Health, vol. 7, no. 3, pp. 328–334, 2010.

[83] P. J. Marchionda, L. K. Krause, J. D. Jensen, and R. P. Dellavalle,“A North American perspective on dermoscopy: benefits,limitations, and grey zones,” Giornale Italiano di Dermatologiae Venereologia, vol. 145, no. 1, pp. 89–97, 2010.

[84] C. Ferte, N. Penel, J. Bonneterre, and A. Adenis, “Individuallife expectancy estimation using validated prognostic scoresfor patients with cancer of unknown primary,” Oncology, vol.78, no. 2, pp. 87–93, 2010.

[85] F. F. T. Barros, D. G. Powe, I. O. Ellis, and A. R. Green,“Understanding the HER family in breast cancer: interactionwith ligands, dimerization and treatments,” Histopathology,vol. 56, no. 5, pp. 560–572, 2010.

[86] C. C. Ton, N. Vartanian, X. Chai et al., “Gene expression arraytesting of FFPE archival breasttumor samples: an optimizedprotocol for WG-DASL sample preparation,” Breast CancerResearch and Treatment. In press.

[87] O. Tureci, U. Sahin, I. Schobert et al., “The SSX-2 gene, whichis involved in the t(X;18) translocation of synovial sarcomas,codes for the human tumor antigen HOM-MEL-40,” CancerResearch, vol. 56, no. 20, pp. 4766–4772, 1996.

[88] Y. T. Chen, A. O. Gure, S. Tsang et al., “Identificationof multiple cancer/testis antigens by allogeneic antibodyscreening of a melanoma cell line library,” Proceedings of theNational Academy of Sciences of the United States of America,vol. 95, no. 12, pp. 6919–6923, 1998.


Review Article

Focus on Adoptive T Cell Transfer Trials in Melanoma

Liat Hershkovitz,1 Jacob Schachter,1 Avraham J. Treves,2 and Michal J. Besser1, 3

1 Ella Institute of Melanoma, Sheba Medical Center, Tel-Hashomer 52621, Israel2 Sheba Cancer Research Center, Sheba Medical Center, Tel-Hashomer 52621, Israel3 Department of Human Microbiology and Immunology, Sackler School of Medicine, Tel Aviv University, Tel Aviv 69978, Israel

Correspondence should be addressed to Michal J. Besser, [email protected]

Received 29 July 2010; Accepted 8 November 2010

Academic Editor: Charles R. Rinaldo

Copyright © 2010 Liat Hershkovitz et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Adoptive Cell Transfer (ACT) of Tumor-Infiltrating Lymphocytes (TIL) in combination with lymphodepletion has proven to bean effective treatment for metastatic melanoma patients, with an objective response rate in 50%–70% of the patients. It is basedon the ex vivo expansion and activation of tumor-specific T lymphocytes extracted from the tumor and their administration backto the patient. Various TIL-ACT trials, which differ in their TIL generation procedures and patient preconditioning, have beenreported. In the latest clinical studies, genetically engineered peripheral T cells were utilized instead of TIL. Further improvementof adoptive T cell transfer depends on new investigations which seek higher TIL quality, increased durable response rates, and aimto treat more patients. Simplifying this therapy may encourage cancer centers worldwide to adopt this promising technology. Thispaper focuses on the latest progress regarding adoptive T cell transfer, comparing the currently available protocols and discussingtheir advantages, disadvantages, and implication in the future.

1. Introduction

Metastatic melanoma is a highly aggressive cancer. It is thesixth most common cancer type in men and the seventh inwoman and has a poor prognosis with a median survival of6–10 months and a 5 year survival rate of about 10% [1, 2].So far, only two drugs have been approved by the Food andDrug Administration (FDA) for the treatment of metastaticmelanoma patients, the chemotherapeutic agent dacarbazine(DTIC) and the immunomodulator Interleukin-2 (IL-2).DTIC is an antineoplastic chemotherapy drug used in thetreatment of various cancers [3, 4], which alkylates andcrosslinks DNA during all phases of the cell cycle, result-ing in disruption of DNA function, cell-cycle arrest, andapoptosis [5]. DTIC as a single agent can induce objectivetumor regression in 15%–20% of metastatic melanomapatients with median response duration of 5 to 6 monthsand a negligible complete response rate [6]. The admin-istration of IL-2 activates endogenous antitumor reactiveT cells and NK cells resulting in a 20%–30% objectiveresponse rate, with complete regression in only 5%–8% of

the melanoma patients [7–9]. Chemobiotherapy, combininglow-dose IL-2 and chemotherapy, demonstrates a similarresponse rate to that of high-dose IL-2 alone, with 7%of the patients achieving a durable complete response [10,11].

In the past decade, various therapeutic approaches formetastatic melanoma patients have been tested, most ofthem with unsatisfactory clinical results. The discovery ofmelanoma-specific tumor associated antigens paved the wayfor the use of immunotherapy [12]. Immunotherapy is aninnovative approach which includes: (1) cytokines, suchas IL-2 and IFNα, which stimulate immune effector cells,(2) monoclonal antibodies, against cytotoxic T lymphocyte-associated antigen 4 (CTLA-4) or programmed death 1(PD-1), (3) intratumoral gene transfer with genes encodingfor example for granulocyte-macrophage colony-stimulatingfactor (GM-CSF), (4) active immunization via vaccinationwith peptides, dendritic cells, recombinant viruses encod-ing tumor-associated antigens (TAA), and costimulatorymolecules or whole tumor cell vaccines, and (5) adoptive Tcell transfer.


Like IL-2, IFNα is FDA approved, but only for adjuvanttreatment of stage III melanoma patients and not for patientswith metastatic disease [11]. The monoclonal antibodyipilimumab, which blocks the inhibitory effects of CTLA-4 on T cells, was evaluated in a randomized phase IIIstudy and showed an improved overall survival comparedto glycoprotein 100- (gp100-) derived peptide vaccines.However, adverse events were sometimes severe or longlasting [13].

In the past few years, the field of intratumoral genetransfer has been established. A very encouraging phaseII study conducted on 43 melanoma patients reported analmost 30% objective response with very minor side effectsfollowing intratumoral injection of a viral vector encodingfor GM-CSF [14]. On the other hand, various cancervaccines trials have failed so far, as they demonstrated a verylow objective response rate of only 3%-4% [15].

Among the current immunotherapy approaches, ACTtherapy using autologous tumor infiltrating lymphocytes(TIL) has shown to be an effective treatment for patients withmetastatic melanoma and can mediate objective responserates between 50%–70% with manageable toxicity [16–18]. Furthermore, the ability to genetically engineer tumorantigen specific lymphocytes may enable the use of ACTto a wider patient population and for treatment of othermalignancies. This paper summarizes and compares thedifferent ACT approaches and discusses their advantages,disadvantages, and challenges.

2. Initial ACT Studies

Adoptive cell transfer immunotherapy relies on the transferof ex vivo expanded and activated tumor-specific lympho-cytes to the autologous tumor-bearing host. In the 1980s,the availability of recombinant IL-2 for in vitro use broughtthe breakthrough of ACT trials [19–21]. Using IL-2 as a Tcell growth factor in culture promoted the ability to expandhuman lymphocytes to larger scales and for longer periods.The first generation of ACT utilized lymphokine-activatedkiller (LAK) cells with high-dose IL-2 for the treatment ofpatients with advanced cancer but was not shown to besuperior to the treatment of IL-2 alone [22]. In vitro lysis orcytokine release assays revealed that expanded LAK cells didnot exhibit any anti-tumor activity [23].

The second generation of ACT employed tumor-infiltrating lymphocytes instead of LAK cells. TIL are Tcells present within the tumor of the patient, which havethe ability to specifically recognize tumor antigens andthereby eliminate malignant cells. However, in most cancerpatients, those naturally occurring TIL fail to destroy thetumor as they are outnumbered, not fully activated, orsuppressed. Therefore, the main objective of TIL-ACT is theproduction of large numbers of activated tumor-specific Tcells. Furthermore, patients receive IL-2 immediately afterTIL infusion to further activate the administered TIL. Inmost current protocols, high-dose intravenous bolus IL-2(720,0001 IU/kg) is administered intravenously every 8 hoursto tolerance or to a maximum of 15 doses.

An initial TIL-ACT trial for metastatic melanomapatients was published in 1988 and summarized in 1994by Rosenberg et al. at the National Cancer Institute (NCI),Bethesda [24, 25] (Table 1, row A). In this study, 86 con-secutive patients were treated with autologous TIL includinghigh-dose intravenous bolus IL-2. TIL single-cell suspensionwas generated by the enzymatic digestion of the entire tumorand expanded in a high concentration of IL-2. After 2-3weeks, the cultures were cleared of tumor cells and lympho-cytes were further expanded until they reached a numberof 109 cells. Then, the cells were transferred from tissueculture plates to cell culture bags and cultured until at least1011 TIL were generated. TIL and IL-2 were administered intwo cycles separated by approximately 2 weeks. Fifty-sevenof the 86 patients received an additional single low doseof cyclophosphamide (25 mg/kg) approximately 36 hoursbefore receiving the first TIL infusion. The overall objectiveresponse rate was 34% and comparable to studies withhigh-dose IL-2 alone (31%) or IL-2-based chemobiotherapy(35%). Only 5 of the patients experienced a completeresponse, and the median survival of all partial responderswas just 4 months. Analysis of the survival of the transferredcells in vivo using cells genetically labeled with neomycinphosphotransferase demonstrated a lack of persistence of theinfused cells. The persistence of the TIL in the circulation wasbarely 0.1% one week after administration [26]. Neverthe-less, this initial study demonstrated that TIL-ACT can indeedcause clinical remission although response rates were low atthat time.

A retrospective analysis comparing TIL characteristicsof responding versus nonresponding patients revealed thatresponders were treated with TIL that spent less time inculture, had shorter doubling times, and exhibited higherin vitro lysis against autologous tumor targets [27, 28].Additionally, patients who received TIL generated fromsubcutaneous tumor lesions had higher response rates (49%)compared to those receiving TIL from lymph nodes (17%),probably as lymph nodes contain many unspecific T cells[27].

The results of this study indicated the importance ofa suitable in vitro selection process which can detect anti-tumor specific TIL.

This pioneer trial paved the way for new clinical studiesdesigned to improve the TIL anti-tumor specificity, thegrowth and expansion conditions, and the creation ofyounger TIL culture with higher persistence and lowersenescence.

3. Current TIL Studies

3.1. The “Selected-TIL” Protocol. Earlier studies utilized bulkTIL cultures that contained heterogeneous populations oflymphocytes, only some of which showed anti-tumor activityin vitro [24, 25]. Studies of the T-cell receptor rearrange-ments in TIL revealed the variability of T cells present inthese cultures and emphasized the need to develop culturetechniques that specifically select anti-tumor reactive cellcultures without growth of other bystander cells that do


not contribute to the therapeutic effect [29]. It is knownthat T cells recognize tumor antigens in an MHC-restrictedmanner when the appropriate human leukocyte antigen(HLA) restriction element is present [30–32]. Upon antigenrecognition cytotoxic T cells secrete the cytokine IFNγ whichcan be measured by a simple enzyme-linked immunosorbentassay (ELISA). The integration of an IFNγ release-based invitro assay, which allows the selection of antigen-specific TILis the core element of the Selected-TIL protocol (Table 1,row B) [33–35].

The generation of Selected-TIL with specific reactivityagainst tumor antigens consists of three stages.

(1) Isolation and generation of TIL from the tumor.Several methods can be used for initiating TIL and

melanoma cultures from the resected tumor. TIL culturesare usually established by cutting the resected tumor spec-imens into 1–3 mm3 fragments from different areas of thetumor. Approximately 8–10 single fragments are placed intodifferent wells of a tissue culture plate, and each obtained TILculture is maintained independently. This method generatesmultiple TIL cultures and enables the in vitro screening ofevery individual culture. Besser et al. showed that for 97% ofthe patients at least one TIL culture can be generated whereasa primary melanoma cell culture can be established from70% of the resected tumor [36]. Generating TIL culturesfrom a single small fragment and expanding those cells toapproximately 50 × 106 cells, a suitable number to initiatelarge scale expansion, requires 21–36 days.

(2) In vitro TIL selection.Each individual TIL culture is screened for tumor specific

reactivity by coincubation of TIL with autologous melanomacells over night, followed by the measuring of IFNγ levels.Alternatively, a HLA-A2 matched melanoma cell line canbe used for TIL cultures from HLA-A2 positive patients.TIL pass the screening process, if IFNγ secretion exceeds200 pg/ml and double the level of a control in which the TILare coincubated with any HLA-mismatched melanoma cellline [37, 38]. Only individual TIL cultures complying withthis criterion can be used for further large-scale expansion.

(3) Large-scale expansion.

After reaching 50× 106 cells, TIL are massively expanded,resulting in an approximately 1000-fold expansion and afinal cell number of about 50 × 109 TIL. This rapid expan-sion process (REP) requires anti-CD3 antibody, irradiatedperipheral blood mononuclear feeder cells and IL-2. Thelarge-scale expansion process is a standardized procedurewhich requires 14 days [36, 37] and results in the productionof high numbers of activated TIL ready for administration.

Response rates are assessed using the Response Evalu-ation Criteria In Solid Tumors (RECIST) guidelines [39]and Objective Responses (OR) include Complete Remissions(CR) and Partial Responses (PR).

A major breakthrough in the field of ACT was the addi-tion of lymphodepleting NonMyeloablative Chemotherapy(NMC). NMC is given in order to suppress endogenousregulatory T cells and to provide the optimal environmentfor the infused TIL. Various mouse models have shownthe marked effect of lymphodepletion on the efficacy of

T cell transfer trials (reviewed in [40]). In 2002, Dudleyet al. conducted a Selected-TIL study in which all TILpatients received NMC starting 7 days prior TIL infusion(day 0). On days −7 and −6, patients were treated withcyclophosphamide (60 mg/kg) and on days −5 to −1 withfludarabine (25 mg/m2). Six of 13 (47%) patients achievedan objective response with significant shrinkage of thetumors (Table 1, row B) [41]. The addition of NMC beforeTIL administration resulted in the persistent repopulationof anti-tumor T cells in the patients, with proliferationof functionally active transferred cells in vivo and thetrafficking to tumor sites. A further study was conductedby Dudley et al. three years later including 35 patients,reported 18 (51%) patients that experienced OR includingthree ongoing complete remissions and 15 (42%) partialresponses with a mean progression-free survival of 11.5months [17]. The toxicities associated with this treatmentincluded the expected toxicities of high-dose IL-2 therapy,such as hypotension, pulmonary congestion, vascular leaksyndrome, and bone marrow suppression associated withlymphodepleting chemotherapy. Some patients exhibitedautoimmune manifestations, such as vitiligo and uveitis,caused by destruction of melanocytes in skin and eyes.Toxicities were mostly transient and readily managed bystandard supportive techniques with no treatment-relateddeaths.

The Selected-TIL protocol including lymphodepletingpreconditioning of the patient was indeed a major improve-ment to earlier TIL protocols. Though, the main disadvan-tage of this approach was the extremely high dropout rateof enrolled patients. In a report conducted in 2009, it wasshown that about two thirds of all enrolled patients had to beexcluded from the treatment [36]. This high dropout rate isthe result of two major reasons, directly related to the in vitroIFNγ screening process: (1) failure to establish an autologousmelanoma line for HLA-A2 negative patients which isessential to perform the screening (approximately 30%) and(2) absence of TIL cultures secreting IFNγ (approximately40%) [16, 36]. A few more patients were excluded dueto clinical deterioration or as TIL cultures could not beestablished [36].

Due to the high dropout rate, a different criterion whichrelates to clinical response was required. Interestingly, in1994, Schwartzentruber et al. could not correlate IFNγsecretion to clinical outcome [27]. On the other hand,characteristics that were repeatedly reported to have asignificant positive association to clinical response were shortTIL culture duration and telomere length [27, 28, 42–44]. Anin vitro study comparing younger TIL cultures to older IFNγ-Selected-TIL showed that TIL spending less time in culturehave longer telomeres and high levels of the costimulatorymolecules CD27 and CD28, which can lead to persistence invivo as well as objective responses [42–46]. These in vitro dataled to the development of the so called Young-TIL protocol[16, 36].

3.2. The “Young-TIL” Protocol. The Young-TIL protocolutilizes bulk unselected TIL, which spend minimal time inculture (Table 1, row C). In contrast to the Selected-TIL


Table 1: Comparison of published adoptive T cell transfer studies for melanoma patients.

T cell typeOrigin of Tcell

Culture initiation In vitro screeningLarge-scaleexpansion protocol

Patientconditioning

OR (%) Ref

A TILMelanomabiopsy

One single bulkculture

NoneCells grown withIL-2 till desirednumber obtained

Cyclophosphamide 35 [24, 25]

B Selected-TILMelanomabiopsy

Multiple cultures IFNγ ELISAanti-CD3, IL-2,IFC for 14 days

NMC 50 [17, 41]

C Young-TILMelanomabiopsy

One single bulkculture

Noneanti-CD3, IL-2,IFC for 14 days

NMC 50 [16]

D Selected-TILMelanomabiopsy


NMC + 2 Gy TBI 52 [18]

E Selected-TILMelanomabiopsy


NMC + 12 Gy TBI 72 [18]

FEngineere Tcells

LeukopheresisGene therapy withlow-affinityMART-1 TCR

IFNγ ELISA,Transductio efficiency

various protocols NMC 12 [56]

GEngineere Tcells

LeukopheresisGene therapy withhigh-affinityMART-1 TCR



HEngineere Tcells

LeukopheresisGene therapy withgp100 TCR



(TIL) Tumor infiltrating lymphocytes, (OR) objective response, (Ref) Reference, (IFC) irradiated feeder cells, (NMC) nonmyeloablative chemotherapy, and(TBI) total body irradiation.

protocol, Young-TIL are initiated from a single cell sus-pension obtained after enzymatic digestion of the resectedtumor and not from multiple small fragments. Consequently,the number of TIL is higher in the initial culture whichenables the generation of a TIL culture ready for expansionin only 10 to 18 days [45]. Young-TIL are consideredready for expansion when they reach a number of at least50 × 106 cells and all melanoma cells are eliminated.The process of growing a single TIL culture enormouslysimplifies the laboratory procedure and lowers costs. Ina study from 2010, Besser et al. showed that Young-TILcultures were successfully established for nearly 90% of thepatients (Table 1, row C). As no further selection process wasrequired, all established Young-TIL cultures were eligible fortreatment. This resulted in a dramatic improvement of theproportion of treated patients compared to the Selected-TILprotocol with a drop-out rate of only 26%, mainly causedby clinical deterioration of patients during TIL preparation[16].

TIL large-scale expansion, preconditioning of thepatients with NMC and administration of high-dose IL-2following cell transfer were identical to Selected-TIL ACT.IL-2 and chemotherapy-related toxicities as well as responserate were comparable to the Selected-TIL protocol. Ten of 20(50%) metastatic melanoma patients experienced a clinicalobjective response including two ongoing CR and eightPR [16]. The significant correlation between short cultureduration and clinical response was confirmed. All respondersreceived TIL cultures which were established within 19days, which then continued to the large-scale expansionprocess.

In conclusion, Young-TIL ACT as compared to Selected-TIL ACT has the major advantage of being less labor inten-sive, requires less laboratory expertise, enables the treatmentof most enrolled patients, and still results in an objectiveresponse rate of 50%.

This process simplification might be essential as it mayintegrate ACT to more cancer centers worldwide and thusexpose new patients to this effective therapy.

3.3. “Selected-TIL” ACT with Addition of Total Body Irradi-ation. Studies on murine ACT models have demonstratedthe need for lymphodepletion prior to TIL transfer in orderto eliminate suppressive CD4+CD25+ T-regulatory cells aswell as normal endogenous lymphocytes that compete withthe transferred cells for homeostatic cytokines such as IL-7and IL-15. These studies reported a significant correlationbetween the intensity of lymphodepletion and the in vivoantitumor effect of the infused cells [40, 47–50]. Non-myeloablative chemotherapy using cyclophosphamide andfludarabine are given to deplete the lymphoid compartmentof patient and to provide the optimal environment for theinfused anti-tumor lymphocytes. Addition of total bodyirradiation (TBI) can further augment lymphodepletion.

In order to validate if enhanced lymphdepletion canfurther increase response rate, a series of three nonran-domized clinical trials were conducted with a total of 93metastatic melanoma patients between the years 2002 and2008 [17, 18, 41]. In all three studies, TIL were establishedaccording to the Selected-TIL protocol. Until 2008 a totalof 43 patients were treated with the Selected-TIL protocolin combination with NMC at the NCI, Bethesda, of which


Table 2: Adoptive T cell transfer trials currently recruiting patients.

NCT ID Sponsors Phase Study design Pat. No. Transferred cells TBI NMC IL-2

00287131 Sheba Medical Center II NR 20 Selected or Young-TIL − + +

00338377M.D. AndersonCancer Center ChironCorporation

II R 83TIL versus TIL withdendritic cellimmunization

− + +

01118091National CancerInstitute

II R 135IL-2 versus IL-2 withenriched Young TIL CD8

− + +

00513604 National CancerInstitute

II NR 149 TIL +/− CD4+ depletion − + +

00910650

University ofCalifornia; CaliforniaInstitute of Technol.Univer. of SouthernCalifornia Universityof Connecticut

II NR 22

anti-MART-1TCR-geneengineered lymphocytesand MART-1 peptidepulsed dendritic cells

− + +


II NR 41

anti-MART-1 TCR-geneengineered lymphocytesand ALVAC virusimmunization

− + +


II NR 41

anti-gp100 TCR-geneengineered lymphocytesand ALVAC virusimmunization

− + +

00512889Dana-Farber CancerInstitute

I NR 20

MART1/Melan-A specificCTL +/− GM-CSF andirradiation of cutaneoustumor lesion

− − −

01082887Nantes UniversityHospital

I/II NR 12

TIL in combination withintra-tumoral injectionsof IFNγ adenovirus(Ad-IFNγ)

− − + (s.c)

00324623Centre HospitalierUniversitaire Vaudois

I NR 12Melan-A/MART-1antigen-specificCD8 Tlymphocytes

− + −


II R 95

anti-Mart-1 and peptidevaccines versusanti-gp100 TCR-geneengineered lymphocytesand peptide vaccines

+ + −


II NR 82anti-NY ESO-1TCR-Gene engineeredlymphocytes

− + −

01029873

Altor BioscienceCorporation;National CancerInstitute

I/II NR 58

Cisplatin with ALT-801(IL-2 fused to T-cellreceptor directed againstp53-derived peptides)

− − −


II NR 12anti-p53 TCR-geneengineered lymphocytes

− + +

00871481

Fred HutchinsonCancer ResearchCenter; NationalCancer Institute

I/II ∗∗∗ 30Autologous NY-ESO-1-melanoma-specific CD8+

T cells +/− ipilimumab− + + (s.c)

01106235Fred HutchinsonCancer ResearchCenter

I NR 10Autologous IL-21modulated CD8+

antigen-specific T cells− + + (s.c)

00925314 Cosmo Bioscience II NR 20

CB-10-01 (transgeniclymphocyteimmunization againsttelomerase)

− − −


Table 2: Continued.

NCT ID Sponsors Phase Study design Pat. No. Transferred cells TBI NMC IL-2


I/II NR 33 IL-2 gene-transduced TIL − + +


I/II NR 30Allogeneic tumor-reactivelymphocyte cell lineDMF5

− + +

All clinical trials are registered on http://www.clinicaltrials.gov/; (NCT ID) Clinical trial identifier, (Pat. No.) Number of treated patients, (R) Randomized,(NR) none randomized, (∗∗∗) not reported, (TIL) tumor infiltrating lymphocytes, (TBI) total body irradiation, (NMC) nonmyeloablative chemotherapy,(s.c) subcutaneous.

the result of the first 35 patients were discussed earlier [17](see Section 3.1, Table 1, row B). In a second trial, 25 patientsreceived 2-Gy TBI in addition to NMC one day beforecell administration (Table 1, row D). In the third study, 25patients received a total of 12-Gy TBI, 2-Gy was given twice aday for three consecutive days in addition to NMC (Table 1,row E) [18]. All TBI patients received autologous purifiedCD34+ hematopoietic stem cells one or two days after TILinfusion.

Objective responses were demonstrated in 49% (21 of43) of patients receiving NMC alone, 52% (13 of 25) of thepatients treated with NMC and 2-Gy TBI, and 72% (18 of25) among patients with NMC and 12-Gy TBI [18]. Thehighest rate of complete remissions was achieved in the 12-Gy TBI group with 28%, as compared to 8% and 6% in the 2-Gy TBI and the no TBI-treated patients, respectively. The 2-year survival rates were 42% and 30% in the trials with 2-GyTBI and no TBI, respectively, while followup was too shortin the 12-Gy TBI group. Patients treated with 12-Gy TBIdemonstrated a 1-2 days delay in marrow recovery comparedto all other protocol patients and required significantly moreblood product support than patients treated with 2-Gy TBI.Furthermore, several toxicities such as fatigue, anorexia,and weight loss were more prolonged in patients treatedwith 12-Gy TBI, with returning to normal daily routinesby 2-3 months, compared to one month for none TBIpatients.

It was confirmed that increased lymphodepletion wasassociated with increased circulating levels of homeostaticcytokines IL-7 and IL-15 with 12-Gy TBI achieving thehighest concentrations of the serum cytokines [18]. Inaddition, it was shown that OR was correlated with telomerelength of the transferred cells.

A major disadvantage of the TBI trials was again the highdrop-out rate. Alike the Selected-TIL protocol the majorityof patients did not have TIL applying to the IFNγ criterion.Additionally, transfusion of autologous CD34+ stem cells isrequired after TBI, but sufficient numbers of stem cells couldnot be isolated for every patient, thus increasing the drop-outrate even more.

Overall, using NMC was proven as a necessary andessential step in the ACT therapy and TBI was shownto augment lymphodepletion and thereby increases theresponse rate. However, this conclusion must be interpretedwith caution as this was not a randomized trial. Statistically,the 72% response rate achieved in 12-Gy TBI group wasnot significantly different from the 52% or 49% response

rate seen in the other groups, but induced severe additionaltoxicities. Therefore, this study should be continued withmore patients to verify the high response rate.

4. Genetically Engineered T Cells

One of the possibilities for improving ACT for metastaticmelanoma and other cancer patients is based on the transferof genetically modified peripheral T cells instead of TIL.Genetically engineered T cells may overcome several disad-vantages of the TIL protocol. The TIL protocol encountersseveral difficulties, for example, generating TIL cultures is atechnical challenge and requires high laboratory skills [36].Using this approach, patients have to undergo surgery inorder to resect tumor tissue for TIL isolation and TIL are aheterogenic cell mixture with unknown antigen specificity.In addition, the endogenous T cell repertoire is not alwaysresponsive to defined tumor-associated antigens such asself-antigen due to developed tolerance [51, 52]. Finally,the drop-out rates of enrolled patients can be quite high,as discussed earlier [36]. These disadvantages led to theinvestigation of an alternative method producing tumor-reactive T cells. The idea is to create T cells with desiredantigen specificity and thereby to enhance the effectiveness ofACT therapy [53, 54]. Genetic modification of T cells is basedon the generation of tumor-targeted T cells through thegenetic transfer of antigen-specific receptors, which consistof either MHC-restricted T cell receptors (TCR) or non-MHC-restricted synthetic chimeric antigen receptors (CAR)[55]. The genetic approach is more convenient to the patientsas it skips surgery and only requires patient’s leukopheresis toobtain peripheral T cells for genetic modification.

4.1. Artificial αβ T Cell Receptors. This approach utilizes thetransfer of cloned cDNAs encoding the α and β chains ofa tumor antigen-specific TCR into peripheral T cells by anintegrating vector and thus enables the production of largenumbers of tumor specific T cells [53, 54]. In an initialclinical trial, metastatic melanoma patients were treatedwith autologous T lymphocytes transduced with retrovirusesencoding TCR with low affinity for melanoma-associatedantigen MART-1 [56]. These lymphocytes were adminis-trated to patients preconditioned with NMC followed byhigh-dose bolus IL-2 (Table 1, row F). Two of 17 patients(12%) underwent regression of liver and lung metastases andexhibited sustained levels of circulating engineered cells one


year post-infusion. All the other patients exhibited durablecell engraftment of at least 2 months postinfusion, but didnot respond. This study demonstrates the feasibility of usinggenetically engineered T cells for ACT.

Although response rates were still low (12%) as opposedto 50%–70% in TIL-ACT therapy, this was the first evidencethat genetic-engineered T cells can indeed induce clinicalresponse in advanced cancer patients. This encouragingapproach is only at its beginning and still needs to beimproved. Some of the approaches that may increase theexpression and function of the transgene are being studied,including the use of different vectors, the introduction ofpowerful promoters specific to T cells, and the employmentof higher-affinity TCRs. Latest clinical trials are testing newTCRs that recognize a broad array of cancer antigens suchas p53, gp100, carcinoembryonic antigen and the NY-ESO-1 antigen of the cancer testis [57–59], which may enabletargeting of other cancer types in addition to melanoma.

The main disadvantage of using genes encoding TCR thattarget just one tumor antigen is that they are likely to beless effective. Tumors are highly heterogeneous and antigenexpression differs markedly between tumors not only amongindividual patients, but even within deposits of the samepatient. Investigators should also take into consideration thepotency of antigen/MHC downregulation or loss in tumors.Another limitation of this approach is its HLA-restriction.

In 2009, Johnson et al. conducted a clinical study with36 patients treated with high affinity human TCR to MART-1 (Table 1, row G) and mouse TCR to gp100 (Table 1, rowH) [60]. Objective cancer regressions were seen in 30%(6 of 20) and 19% (3 of 16) of patients who receivedTCR to MART-1 and gp100, respectively. TCR-transducedcells persisted at high levels in the blood of all patientsone month after treatment. However, patients exhibiteddestruction of normal melanocytes in the skin, eyes, andears and sometimes required local steroid administration inorder to treat uveitis and hearing loss [60]. This observationsuggests that great caution must be taken in the selection oftarget antigens to prevent severe toxicity of normal tissuesthat express the same antigen.

4.2. Chimeric Antigen Receptor (CAR). CARs combine anti-gen specificity and T cell activating properties in a singlefusion molecule. The CAR construct which is retrovirallyinserted into normal T lymphocytes, is generated by thefusion of the variable region of the heavy and light chainsof a monoclonal antibodies with the T cell signaling chainsderived from CD28, 41BB, or CD3ζ [61]. While, clinicalstudies in patients with various cancers treated with CARhave induced only modest responses [62–64], a diversearray of receptors with different functional properties arenow rapidly expanding and may provide a more significantassessment for this approach [65–67]. Since CARs bind thetumor antigen in an HLA-unrestricted manner, they areresistant to tumor escape mechanisms related HLA down-regulation and broadly applicable to patients irrespective oftheir HLA type [63, 64, 68].

To date, there are no published clinical studies using CARin ACT therapy for melanoma; nevertheless, in vitro studies

utilizing chimeric receptor against melanoma antigens suchas ganglioside GD2, GD3, and MAGE-A1 were reported [69–71]. Lately, a preclinical study using SCID mice showed thatthe overexpression of GD2 by human melanoma cells allowsthese cells to be targeted in vitro and in vivo by GD2 CAR-expressing T cells [69]. Moreover, this study demonstratedthat incorporation of endodomains from both CD28 andOX40 molecules mediate costimulation of T lymphocytes,inducing T cell activation, proliferation, and cytotoxicityagainst GD2-positive melanoma cells.

A major concern of the CAR approach is the toxicityprofile. T cells encoding CAR for tumor antigens maybe highly potent, but as most tumor-associated antigensare expressed on normal tissues as well, the potential fortoxicity is obvious. A clinical study with renal cell carcinomapatients had to be detained as three patients developedcholestasis due to the high expression of the target antigen,carbonic anhydrase on normal tissue [63]. A case report ona metastatic colon cancer patient using CAR targeting theERBB2 tumor antigen, described respiratory distress of thepatient within 15 minutes after cell infusion and displayed adramatic pulmonary infiltrate on chest X-ray [72].

In summary, T cells transduced with genes encoding forartificial TCR or CAR, enforce tumor antigen recognition,improve T cell survival, generate memory lymphocytes, andreduces T cell death, anergy and immune suppression [55].Another important advantage of these approaches is the useof patients’ circulating lymphocytes with no need of surgicalresection and the ability to produce any desired vector.However, this approach demands high laboratory skills andexpertise and requires longer culture durations resulting inhigh costs. The toxicity profile must be improved and moretrials need to be conducted to establish this technology.

5. Future Challenges of ACT

In order to improve ACT for the treatment of human cancerextensive numbers of studies are being performed. Clinicaltrials for different ACT studies that are currently recruitingpatients are summarized in Table 2. Perhaps the mostimportant trial on this list is a randomized phase II studycomparing CD8 enriched Young-TIL ACT to treatment withhigh-dose IL-2 alone (Table 2, NCT01118091). In fact, this isthe first randomized study which will finally reveal the truevalue of TIL ACT versus standard of care. Many ACT centersworldwide await the results of this study with great interest.

Many studies aim at identifying both patients andcell parameters that are associated with objective clinicalresponses. In this chapter, we will discuss some of them,including (1) the use of less differentiated T cells, (2)identification of the exact T cell subpopulation responsiblefor response, (3) combination of ACT with monoclonalantibodies, (4) genetic engineering of new vectors, and (5)reduction of toxicity.

The persistence of the transferred cells has been clearlycorrelated to clinical response in previous reports [42–46]. One of the factors that influence persistence is thedifferentiation state of the transferred T cells. Mouse modelsshowed that early effector T cells mediate better anti-tumor


response than intermediate and late effector T cells due totheir high proliferative and survival potential and their abilityto secrete IL-2 [73, 74]. Studies in humans also support thispreclinical finding showing that less differentiated T cells aremore suitable for ACT [42, 44, 75]. Long telomere length andhigh CD27 expression are markers of less differentiated cellswhich appear to correlate with persistence and anti-tumorresponse [42, 46, 76]. Addition of molecules inhibitingdifferentiation during T cell expansion, modification of thelarge-scale expansion protocol, transduction of TCR or CARinto less differentiated cells, or the use of other cytokinesbesides IL-2 may improve the quality of administrated cells.IL-2 has been shown to be an effective T cell growth factorbut also to promote the terminal differentiation of T cellsleading to activation-induced cell death [73, 77]. Therefore,the benefit of other cytokines such as IL-7, IL-15, or IL-21 administration alone or in addition to IL-2 should beexamined, as recent studies suggest that they may playimportant roles in manipulating the final T cell phenotypeand differentiation status [78–80].

Other efforts are aimed at identifying specific TILsubpopulations responsible for clinical effectiveness. To date,little is known about the underlying composition and cellularinteractions which determine the degree of TIL reactivityand consequently how to control TIL reactivity. Oved et al.preformed a study measuring the subpopulation frequenciesof TIL populations originating from different melanomapatients [81]. This study showed that the frequency ofa single subpopulation cannot predict the collective TILreactivity to cancer cells in vitro. However, by using a simplecomputational model, one can generate a set of mathematicalrules that accurately predict the degree of TIL reactivity interms of its subpopulation constituents with 91% sensitivity.In addition, Oved et al. controlled the in vitro reactivity ofthe TIL by manipulating their subpopulation composition,enabling to turn nonreactive TIL into reactive ones and viceversa, by simple depletion of specific subpopulation. Mea-suring the subpopulation frequencies of TIL from 45 ACTtreated patients is performed these days at our melanomaresearch center. Frequencies of hundreds of subpopulationswill be determined by flow cytometry and the correlationto clinical response will be evaluated by a computationalmodel. The result of this study will be of great interest asit may reveal the TIL characterization profile of respondingpatients.

The combination of T cell transfer with monoclonalantibodies blocking CTLA-4 or PD-1 function should alsobe evaluated in future studies. CTLA-4 and PD-1 are knownto hinder T cell function via blockade of costimulatorysignals. Therefore, addition of those antibodies could avoidthe inhibitory effects of CTLA-4 or PD-1 on transferred Tcells.

A lot of efforts have been invested in the field of geneticengineered lymphocytes utilizing genes encoding αβTCRand CAR to target TAA. Genetic-engineering also enables theproduction of T cells with other functions besides TAA rec-ognition. Modified T cells might constantly express costim-ulatory molecules like CD28, produce autocrine cytokinessuch as IL-2, express homing molecules (e.g., CD62L) or

prevent apoptosis by the expression of antiapoptotic genes(e.g., bcl-2). Genetic approaches provide a platform forfurther developments in immunotherapy and will directACT to a new era of antitumor therapy.

Another important aim is the reduction of the sideeffects related to the treatment. Establishing cells for ACT atthe laboratory is challenging, but great efforts are requiredfrom clinicians as well. Most grade III and IV toxicitiesare directly related to the intravenously administration ofhigh-dose bolus IL-2 (720,000 IU/kg, every 8 h) after celltransfer [16, 24]. Reducing the dose of IL-2 significantlydecreases toxicity [82, 83]. Many current trials (Table 2) aretherefore using low-dose subcutaneous IL-2 administrationin order to relieve the patients. Subcutaneous low-dose IL-2 administration also minimizes the costs of the treatmentas its administration does not require hospitalization. Acombined approach of Young-TIL ACT with subcutaneouslow-dose IL-2 should be evaluated in the future. Thiscombination drastically simplifies the laboratory and clinicalefforts and could allow the widespread applicability ofACT.

6. Conclusions

Melanoma has been widely studied as a target forimmunotherapy because it has been considered more sus-ceptible to immune attack than other tumors. Adoptive Tcell transfer has not only shown promising clinical results inthe last decade but also provides a platform for the future.In the coming years, the use of advanced protocols willsurely enhance the anti-tumor reactivity of transferred cells.Sophisticated genetic approaches are of great potential in thefuture, but so far TIL ACT is the most effective cell therapyfor metastatic melanoma patients, with objective responsesin more than 50% of the patients. Simplifying the TILproduction process, by using Young-TIL, may enable cancercenters worldwide to implement this effective approach andexpose more patients to ACT therapy.

Acknowledgments

The authors thank Haya and Nehemia Lemelbaum for theircontinuous and everlasting support as well as Gal Markel andDaphna Levy for their critical reading of this paper.


The authors indicate no potential conflict of interests.

References

[1] A. Jemal, R. Siegel, E. Ward et al., “Cancer statistics, 2008,” CACancer Journal for Clinicians, vol. 58, no. 2, pp. 71–96, 2008.

[2] A. A. Tarhini and S. S. A. Garwala, “Cutaneous melanoma:available therapy for metastatic disease,” Dermatologic Ther-apy, vol. 19, no. 1, pp. 19–25, 2006.

[3] P. Seam, J. E. Janik, D. L. Longo, and V. T. DeVita, “Role ofchemotherapy in Hodgkin’s lymphoma,” Cancer Journal, vol.15, no. 2, pp. 150–154, 2009.


[4] A. S. Yang and P. B. Chapman, “The history and future ofchemotherapy for melanoma,” Hematology/Oncology Clinics ofNorth America, vol. 23, no. 3, pp. 583–597, 2009.

[5] E. Bajetta, M. Del Vecchio, C. Bernard-Marty et al.,“Metastatic melanoma: chemotherapy,” Seminars in Oncology,vol. 29, no. 5, pp. 427–445, 2002.

[6] L. Serrone, M. Zeuli, F. M. Sega, and F. Cognetti,“Dacarbazine-based chemotherapy for metastatic melanoma:thirty-year experience overview,” Journal of Experimental andClinical Cancer Research, vol. 19, no. 1, pp. 21–34, 2000.

[7] M. T. Lotze, A. E. Chang, and C. A. Seipp, “High-doserecombinant interleukin 2 in the treatment of patients withdisseminated cancer. Responses, treatment-related morbidity,and histologic findings,” Journal of the American MedicalAssociation, vol. 256, no. 22, pp. 3117–3124, 1986.

[8] S. A. Rosenberg, M. T. Lotze, and L. M. Muul, “Observationson the systemic administration of autologous lymphokine-activated killer cells and recombinant interleukin-2 to patientswith metastatic cancer,” New England Journal of Medicine, vol.313, no. 23, pp. 1485–1492, 1985.

[9] S. A. Rosenberg, J. C. Yang, D. E. White, and S. M. Steinberg,“Durability of complete responses in patients with metastaticcancer treated with high-dose interleukin-2: identification ofthe antigens mediating response,” Annals of Surgery, vol. 228,no. 3, pp. 307–319, 1998.

[10] M. B. Atkins, L. Kunkel, M. Sznol, and S. A. Rosenberg, “High-dose recombinant interleukin-2 therapy in patients withmetastatic melanoma: long-term survival update,” CancerJournal from Scientific American, vol. 6, supplement 1, pp. S11–S14, 2000.

[11] H. Tsao, M. B. Atkins, and A. J. Sober, “Management ofcutaneous melanoma,” New England Journal of Medicine, vol.351, no. 10, pp. 998–1042, 2004.

[12] S. A. Rosenberg, “Progress in human tumour immunologyand immunotherapy,” Nature, vol. 411, no. 6835, pp. 380–384,2001.

[13] F. S. Hodi, S. J. O’Day, D. F. McDermott et al., “Improved sur-vival with ipilimumab in patients with metastatic melanoma,”New England Journal of Medicine, vol. 363, no. 8, pp. 711–723,2010.

[14] N. N. Senzer, H. L. Kaufman, T. Amatruda et al., “Phase IIclinical trial of a granulocyte-macrophage colony-stimulatingfactor-encoding, second-generation oncolytic herpesvirus inpatients with unresectable metastatic melanoma,” Journal ofClinical Oncology, vol. 27, no. 34, pp. 5763–5771, 2009.

[15] S. A. Rosenberg, J. C. Yang, and N. P. Restifo, “Cancerimmunotherapy: moving beyond current vaccines,” NatureMedicine, vol. 10, no. 9, pp. 909–915, 2004.

[16] M. J. Besser, R. Shapira-Frommer, A. J. Treves et al., “Clinicalresponses in a phase II study using adoptive transfer of short-term cultured tumor infiltration lymphocytes in metastaticmelanoma patients,” Clinical Cancer Research, vol. 16, no. 9,pp. 2646–2655, 2010.

[17] M. E. Dudley, J. R. Wunderlich, J. C. Yang et al., “Adoptivecell transfer therapy following non-myeloablative but lym-phodepleting chemotherapy for the treatment of patients withrefractory metastatic melanoma,” Journal of Clinical Oncology,vol. 23, no. 10, pp. 2346–2357, 2005.

[18] M. E. Dudley, J. C. Yang, R. Sherry et al., “Adoptive cell therapyfor patients with metastatic melanoma: evaluation of intensivemyeloablative chemoradiation preparative regimens,” Journalof Clinical Oncology, vol. 26, no. 32, pp. 5233–5239, 2008.

[19] J. J. Mule, S. Shu, S. L. Schwarz, and S. A. Rosenberg, “Adoptiveimmunotherapy of established pulmonary metastases withLAK cells and recombinant interleukin-2,” Science, vol. 225,no. 4669, pp. 1487–1489, 1984.

[20] L. M. Muul, K. Nason-Burchenal, and C. S. Carter, “Develop-ment of an automated closed system for generation of humanlymphokine-activated killer (LAK) cells for use in adoptiveimmunotherapy,” Journal of Immunological Methods, vol. 101,no. 2, pp. 171–181, 1987.

[21] L. Mesler Muul, K. Nason-Burchenal, C. Hyatt et al., “Stud-ies of serum-free culture medium in the generation oflymphokine activated killer cells,” Journal of ImmunologicalMethods, vol. 105, no. 2, pp. 183–192, 1987.

[22] S. A. Rosenberg, M. T. Lotze, J. C. Yang et al., “Prospectiverandomized trial of high-dose interleukin-2 alone or inconjunction with lymphokine-activated killer cells for thetreatment of patients with advanced cancer,” Journal of theNational Cancer Institute, vol. 85, no. 8, pp. 622–632, 1993.

[23] M. E. Dudley and S. A. Rosenberg, “Adoptive cell transfertherapy,” Seminars in Oncology, vol. 34, no. 6, pp. 524–531,2007.

[24] S. A. Rosenberg, J. R. Yannelli, J. C. Yang et al., “Treatment ofpatients with metastatic melanoma with autologous tumor-infiltrating lymphocytes and interleukin 2,” Journal of theNational Cancer Institute, vol. 86, no. 15, pp. 1159–1166, 1994.

[25] S. A. Rosenberg, B. S. Packard, P. M. Aebersold et al., “Useof tumor-infiltrating lymphocyts and interleukin-2 in theimmunotherapy of patients with metastatic melanoma. Apreliminary report,” New England Journal of Medicine, vol.319, no. 25, pp. 1676–1680, 1988.

[26] S. A. Rosenberg, P. Aebersold, K. Cornetta et al., “Gene trans-fer into humans - Immunotherapy of patients with advancedmelanoma, using tumor-infiltrating lymphocytes modifiedby retroviral gene transduction,” New England Journal ofMedicine, vol. 323, no. 9, pp. 570–578, 1990.

[27] D. J. Schwartzentruber, S. S. Hom, R. Dadmarz et al., “Invitro predictors of therapeutic response in melanoma patientsreceiving tumor-infiltrating lymphocytes and interleukin-2,”Journal of Clinical Oncology, vol. 12, no. 7, pp. 1475–1483,1994.

[28] P. Aebersold, C. Hyatt, S. Johnson et al., “Lysis of autologousmelanoma cells by tumor-infiltrating lymphocytes: associa-tion with clinical response,” Journal of the National CancerInstitute, vol. 83, no. 13, pp. 932–937, 1991.

[29] J. M. Kirkwood, A. A. Tarhini, M. C. Panelli et al., “Next gen-eration of immunotherapy for melanoma,” Journal of ClinicalOncology, vol. 26, no. 20, pp. 3445–3455, 2008.

[30] Y. Kawakami, R. Zakut, S. L. Topalian, H. Stotter, and S. A.Rosenberg, “Shared human melanoma antigens: recognitionby tumor-infiltrating lymphocytes in HLA-A2.1-transfectedmelanomas,” Journal of Immunology, vol. 148, no. 2, pp. 638–643, 1992.

[31] S. S. Hom, S. L. Topalian, T. Simonis, M. Mancini, and S.A. Rosenberg, “Common expression of melanoma tumor-associated antigens recognized by human tumor infiltratinglymphocytes: analysis by human lymphocyte antigen restric-tion,” Journal of Immunotherapy, vol. 10, no. 3, pp. 153–164,1991.

[32] T. L. Darrow, C. L. Slingluff, and H. F. Seigler, “The roleof HLA class I antigens in recognition of melanoma cells bytumor-specific cytotoxic T lymphocytes. Evidence for sharedtumor antigens,” Journal of Immunology, vol. 142, no. 9, pp.3329–3335, 1989.


[33] D. J. Schwartzentruber, S. L. Topalian, M. Mancini, and S.A. Rosenberg, “Specific release of granulocyte-macrophagecolony-stimulating factor, tumor necrosis factor-α, and IFN-γ by human tumor-infiltrating lymphocytes after autologoustumor stimulation,” Journal of Immunology, vol. 146, no. 10,pp. 3674–3681, 1991.

[34] D. J. Schwartzentruber, D. Solomon, S. A. Rosenberg, andS. L. Topalian, “Characterization of lymphocytes infiltratinghuman breast cancer: specific immune reactivity detected bymeasuring cytokine secretion,” Journal of Immunotherapy, vol.12, no. 1, pp. 1–12, 1992.

[35] S. S. Hom, D. J. Schwartzentruber, S. A. Rosenberg, and S. L.Topalian, “Specific release of cytokines by lymphocytes infil-trating human melanomas in response to shared melanomaantigens,” Journal of Immunotherapy, vol. 13, no. 1, pp. 18–30,1993.

[36] M. J. Besser, R. Shapira-Frommer, A. J. Treves et al.,“Minimally cultured or selected autologous tumor-infiltratinglymphocytes after a lympho-depleting chemotherapy regimenin metastatic melanoma patients,” Journal of Immunotherapy,vol. 32, no. 4, pp. 415–423, 2009.

[37] M. E. Dudley, J. R. Wunderlich, T. E. Shelton, J. Even, and S.A. Rosenberg, “Generation of tumor-infiltrating lymphocytecultures for use in adoptive transfer therapy for melanomapatients,” Journal of Immunotherapy, vol. 26, no. 4, pp. 332–342, 2003.

[38] M. J. Besser, A. J. Treves, O. Itzhaki et al., “Adoptive cell therapyfor metastatic melanoma patients: pre-clinical development atthe Sheba Medical Center,” Israel Medical Association Journal,vol. 8, no. 3, pp. 164–168, 2006.

[39] P. Therasse, S. G. Arbuck, E. A. Eisenhauer et al., “Newguidelines to evaluate the response to treatment in solidtumors. European Organization for Research and Treatmentof Cancer, National Cancer Institute of the United States,National Cancer Institute of Canada,” Journal of the NationalCancer Institute, vol. 92, no. 3, pp. 205–216, 2000.

[40] P. Muranski, A. Boni, C. Wrzesinski et al., “Increased intensitylymphodepletion and adoptive immunotherapy—how far canwe go?” Nature Clinical Practice Oncology, vol. 3, no. 12, pp.668–681, 2006.

[41] M. E. Dudley, J. R. Wunderlich, P. F. Robbins et al., “Cancerregression and autoimmunity in patients after clonal repop-ulation with antitumor lymphocytes,” Science, vol. 298, no.5594, pp. 850–854, 2002.

[42] J. Zhou, X. Shen, J. Huang, R. J. Hodes, S. A. Rosenberg, andP. F. Robbins, “Telomere length of transferred lymphocytescorrelates with in vivo persistence and tumor regression inmelanoma patients receiving cell transfer therapy,” Journal ofImmunology, vol. 175, no. 10, pp. 7046–7052, 2005.

[43] X. Shen, J. Zhou, K. S. Hathcock et al., “Persistence oftumor infiltrating lymphocytes in adoptive immunotherapycorrelates with telomere length,” Journal of Immunotherapy,vol. 30, no. 1, pp. 123–129, 2007.

[44] P. F. Robbins, M. E. Dudley, J. Wunderlich et al., “Cutting edge:persistence of transferred lymphocyte clonotypes correlateswith cancer regression in patients receiving cell transfertherapy,” Journal of Immunology, vol. 173, no. 12, pp. 7125–7130, 2004.

[45] K. Q. Tran, J. Zhou, K. H. Durflinger et al., “Minimally cul-tured tumor-infiltrating lymphocytes display optimal charac-teristics for adoptive cell therapy,” Journal of Immunotherapy,vol. 31, no. 8, pp. 742–751, 2008.

[46] D. J. Powell, M. E. Dudley, P. F. Robbins, and S. A. Rosenberg,“Transition of late-stage effector T cells to CD27 CD28 tumor-reactive effector memory T cells in humans after adoptive celltransfer therapy,” Blood, vol. 105, no. 1, pp. 241–250, 2005.

[47] P. A. Antony, C. A. Piccirillo, A. Akpinarli et al., “CD8 Tcell immunity against a tumor/self-antigen is augmented byCD4 T helper cells and hindered by naturally occurring Tregulatory cells,” Journal of Immunology, vol. 174, no. 5, pp.2591–2601, 2005.

[48] W. Dummer, A. G. Niethammer, R. Baccala et al., “T cellhomeostatic proliferation elicits effective antitumor autoim-munity,” Journal of Clinical Investigation, vol. 110, no. 2, pp.185–192, 2002.

[49] L. Gattinoni, S. E. Finkelstein, C. A. Klebanoff et al., “Removalof homeostatic cytokine sinks by lymphodepletion enhancesthe efficacy of adoptively transferred tumor-specific CD8 Tcells,” Journal of Experimental Medicine, vol. 202, no. 7, pp.907–912, 2005.

[50] C. Wrzesinski, C. M. Paulos, L. Gattinoni et al., “Hematopoi-etic stem cells promote the expansion and function ofadoptively transferred antitumor CD8 T cells,” Journal ofClinical Investigation, vol. 117, no. 2, pp. 492–501, 2007.

[51] M. Theobald, J. Biggs, J. Hernandez, J. Lustgarten, C. Labadie,and L. A. Sherman, “Tolerance to p53 by A2.1-restrictedcytotoxic T lymphocytes,” Journal of Experimental Medicine,vol. 185, no. 5, pp. 833–841, 1997.

[52] R. Romieu, M. Baratin, M. Kayibanda et al., “Passive but notactive CD8+ T cell-based immunotherapy interferes with livertumor progression in a transgenic mouse model,” Journal ofImmunology, vol. 161, no. 10, pp. 5133–5137, 1998.

[53] A. Murphy, J. A. Westwood, M. W. L. Teng, M. Moeller, P. K.Darcy, and M. H. Kershaw, “Gene modification strategies toinduce tumor immunity,” Immunity, vol. 22, no. 4, pp. 403–414, 2005.

[54] M. Sadelain, I. Riviere, and R. Brentjens, “Targeting tumourswith genetically enhanced T lymphocytes,” Nature ReviewsCancer, vol. 3, no. 1, pp. 35–45, 2003.

[55] M. Sadelain, R. Brentjens, and I. Riviere, “The promiseand potential pitfalls of chimeric antigen receptors,” CurrentOpinion in Immunology, vol. 21, no. 2, pp. 215–223, 2009.


[57] M. R. Theoret, C. J. Cohen, A. V. Nahvi et al., “Relationshipof p53 overexpression on cancers and recognition by anti-p53T cell receptor-transduced T cells,” Human Gene Therapy, vol.19, no. 11, pp. 1219–1231, 2008.

[58] P. F. Robbins, Y. F. Li, M. El-Gamil et al., “Single and dualamino acid substitutions in TCR CDRs can enhance antigen-specific T cell functions,” Journal of Immunology, vol. 180, no.9, pp. 6116–6131, 2008.

[59] M. R. Parkhurst, J. Joo, J. P. Riley et al., “Characterizationof genetically modified T-cell receptors that recognize theCEA:691-699 peptide in the context of HLA-A2.1 on humancolorectal cancer cells,” Clinical Cancer Research, vol. 15, no. 1,pp. 169–180, 2009.

[60] L. A. Johnson, R. A. Morgan, M. E. Dudley et al., “Genetherapy with human and mouse T-cell receptors mediates can-cer regression and targets normal tissues expressing cognateantigen,” Blood, vol. 114, no. 3, pp. 535–546, 2009.

[61] Z. Eshhar, T. Waks, G. Gross, and D. G. Schindler, “Spe-cific activation and targeting of cytotoxic lymphocytesthrough chimeric single chains consisting of antibody-bindingdomains and the γ or ζ subunits of the immunoglobulin


and T-cell receptors,” Proceedings of the National Academy ofSciences of the United States of America, vol. 90, no. 2, pp. 720–724, 1993.

[62] B. G. Till, M. C. Jensen, J. Wang et al., “Adoptive immunother-apy for indolent non-hodgkin lymphoma and mantle celllymphoma using genetically modified autologous CD20-specific T cells,” Blood, vol. 112, no. 6, pp. 2261–2271, 2008.

[63] C. H. Lamers, S. Sleijfer, A. G. Vulto et al., “Treatmentof metastatic renal cell carcinoma with autologous T-lymphocytes genetically retargeted against carbonic anhydraseIX: first clinical experience,” Journal of Clinical Oncology, vol.24, no. 13, pp. e20–e22, 2006.

[64] M. H. Kershaw, J. A. Westwood, L. L. Parker et al., “A phaseI study on adoptive immunotherapy using gene-modified Tcells for ovarian cancer,” Clinical Cancer Research, vol. 12, no.20, part 1, pp. 6106–6115, 2006.

[65] J. Maher, R. J. Brentjens, G. Gunset, I. Riviere, and M. Sadelain,“Human T-lymphocyte cytotoxicity and proliferation directedby a single chimeric TCRζ/CD28 receptor,” Nature Biotechnol-ogy, vol. 20, no. 1, pp. 70–75, 2002.

[66] A. Hombach, C. Heuser, T. Marquardt et al., “CD4+ T cellsengrafted with a recombinant immunoreceptor efficiently lysetarget cells in a MHC antigen- and Fas-independent fashion,”Journal of Immunology, vol. 167, no. 2, pp. 1090–1096, 2001.

[67] N. M. Haynes, J. A. Trapani, M. W. L. Teng et al., “Single-chain antigen recognition receptors that costimulate potentrejection of established experimental tumors,” Blood, vol. 100,no. 9, pp. 3155–3163, 2002.

[68] G. Gross, T. Waks, and Z. Eshhar, “Expression ofimmunoglobulin-T-cell receptor chimeric moleculesas functional receptors with antibody-type specificity,”Proceedings of the National Academy of Sciences of the UnitedStates of America, vol. 86, no. 24, pp. 10024–10028, 1989.

[69] E. Yvon, M. Del Vecchio, B. Savoldo et al., “Immunotherapyof metastatic melanoma using genetically engineered GD2-specific T cells,” Clinical Cancer Research, vol. 15, no. 18, pp.5852–5860, 2009.

[70] C. O. Yun, K. F. Nolan, E. J. Beecham, R. A. Reisfeld, and R.P. Junghans, “Targeting of T lymphocytes to melanoma cellsthrough chimeric anti-GD3 immunoglobulin t-cell receptors,”Neoplasia, vol. 2, no. 5, pp. 449–459, 2000.

[71] R. A. Willemsen, C. Ronteltap, P. Chames, R. Debets, andR. L. H. Bolhuis, “T cell retargeting with MHC classI-restricted antibodies: the CD28 costimulatory domainenhances antigen-specific cytotoxicity and cytokine produc-tion,” Journal of Immunology, vol. 174, no. 12, pp. 7853–7858,2005.

[72] R. A. Morgan, J. C. Yang, M. Kitano, M. E. Dudley, C. M. Lau-rencot, and S. A. Rosenberg, “Case report of a serious adverseevent following the administration of t cells transduced witha chimeric antigen receptor recognizing ERBB2,” MolecularTherapy, vol. 18, no. 4, pp. 843–851, 2010.

[73] L. Gattinoni, C. A. Klebanoff, D. C. Palmer et al., “Acquisitionof full effector function in vitro paradoxically impairs thein vivo antitumor efficacy of adoptively transferred CD8+T cells,” Journal of Clinical Investigation, vol. 115, no. 6, pp.1616–1626, 2005.

[74] C. A. Klebanoff, L. Gattinoni, P. Torabi-Parizi et al., “Centralmemory self/tumor-reactive CD8+ T cells confer superiorantitumor immunity compared with effector memory T cells,”Proceedings of the National Academy of Sciences of the UnitedStates of America, vol. 102, no. 27, pp. 9571–9576, 2005.

[75] J. Huang, H. T. Khong, M. E. Dudley et al., “Survival, persis-tence, and progressive differentiation of adoptively transferred

tumor-reactive T cells associated with tumor regression,”Journal of Immunotherapy, vol. 28, no. 3, pp. 258–267, 2005.

[76] J. Huang, K. W. Kerstann, M. Ahmadzadeh et al., “Modulationby IL-2 of CD70 and CD27 expression on CD8+ T cells:importance for the therapeutic effectiveness of cell transferimmunotherapy,” Journal of Immunology, vol. 176, no. 12, pp.7726–7735, 2006.

[77] Y. Refaeli, L. Van Parijs, C. A. London, J. Tschopp, and A.K. Abbas, “Biochemical mechanisms of IL-2-regulated Fas-mediated T cell apoptosis,” Immunity, vol. 8, no. 5, pp. 615–623, 1998.

[78] N. Pouw, E. Treffers-Westerlaken, J. Kraan et al., “Combina-tion of IL-21 and IL-15 enhances tumour-specific cytotoxicityand cytokine production of TCR-transduced primary T cells,”Cancer Immunology, Immunotherapy, vol. 59, no. 6, pp. 921–931, 2010.

[79] S. Jaleco, L. Swainson, V. Dardalhon, M. Burjanadze, S.Kinet, and N. Taylor, “Homeostasis of naive and memoryCD4 T cells: IL-2 and IL-7 differentially regulate the balancebetween proliferation and Fas-mediated apoptosis,” Journal ofImmunology, vol. 171, no. 1, pp. 61–68, 2003.

[80] C. S. Hinrichs, R. Spolski, C. M. Paulos et al., “IL-2 and IL-21confer opposing differentiation programs to CD8+ T cells foradoptive immunotherapy,” Blood, vol. 111, no. 11, pp. 5326–5333, 2008.

[81] K. Oved, E. Eden, M. Akerman et al., “Predicting andcontrolling the reactivity of immune cell populations againstcancer,” Molecular Systems Biology, vol. 5, article 265, 2009.

[82] C. Yee, J. A. Thompson, D. Byrd et al., “Adoptive T celltherapy using antigen-specific CD8+ T cell clones for thetreatment of patients with metastatic melanoma: in vivopersistence, migration, and antitumor effect of transferred Tcells,” Proceedings of the National Academy of Sciences of theUnited States of America, vol. 99, no. 25, pp. 16168–16173,2002.

[83] J. C. Yang, R. M. Sherry, S. M. Steinberg et al., “Randomizedstudy of high-dose and low-dose interleukin-2 in patients withmetastatic renal cancer,” Journal of Clinical Oncology, vol. 21,no. 16, pp. 3127–3132, 2003.


Review Article

T Cell-Tumor Interaction Directs the Development ofImmunotherapies in Head and Neck Cancer

A. E. Albers,1 L. Strauss,2 T. Liao,1 T. K. Hoffmann,3 and A. M. Kaufmann4

1 Department of Otolaryngology, Head and Neck Surgery, Charite-Universitatsmedizin Berlin, Campus Benjamin Franklin,12200 Berlin, Germany

2 Fondazione Humanitas per la Ricerca, 20089 Rozzano, Italy3 Department of Otolaryngology, Head and Neck Surgery, Universitat Essen, 45147 Essen, Germany4 Department of Gynecology, Charite-Universitatsmedizin Berlin, Campus Benjamin Franklin and Campus Mitte,12200 Berlin, Germany

Correspondence should be addressed to A. E. Albers, [email protected]

Received 17 July 2010; Accepted 16 October 2010

Academic Editor: Eiji Matsuura

Copyright © 2010 A. E. Albers et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The competent immune system controls disease effectively due to induction, function, and regulation of effector lymphocytes.Immunosurveillance is exerted mostly by cytotoxic T-lymphocytes (CTLs) while specific immune suppression is associatedwith tumor malignancy and progression. In squamous cell carcinoma of the head and neck, the presence, activity, but alsosuppression of tumor-specific CTL have been demonstrated. Functional CTL may exert a selection pressure on the tumor cellsthat consecutively escape by a combination of molecular and cellular evasion mechanisms. Certain of these mechanisms targetantitumor effector cells directly or indirectly by affecting cells that regulate CTL function. This results in the dysfunction orapoptosis of lymphocytes and dysregulated lymphocyte homeostasis. Another important tumor-escape mechanism is to avoidrecognition by dysregulation of antigen processing and presentation. Thus, both induction of functional CTL and susceptibility ofthe tumor and its microenvironment to become T cell targets should be considered in CTL-based immunotherapy.

1. Introduction

Squamous cell carcinomas of the head and neck (SCCHN)are the sixth most frequent type of malignancy worldwide.SCCHN accounts for approximately 6% of all cancercases and for about 650,000 new cases and 350,000 deathsworldwide each year [1–3]. While early-stage SCCHN can betreated relatively effectively, fewer than 40% of patients withadvanced, metastatic disease can be cured. Unfortunately,about two thirds of patients with SCCHN present withadvanced-stage disease, commonly involving regional lymphnodes. Distant metastases are found in about 10% of patientsat initial presentation. The 5-year survival for all stages isabout 60%. Despite significant improvements in surgery,radiation, and chemotherapy, long-term survival rates inpatients with advanced stage SCCHN have not significantlyincreased in the past 30 years [4–6].

Mortality from SCCHN remains high because of thedevelopment of distant metastases and the emergence

of therapy-resistant local and regional recurrences. It istherefore essential to develop a deeper understanding of thebiology of this disease for more effective alternative therapiessuch as immunotherapy. As basis for immunotherapeuticapproaches, interactions between tumors and the hostimmune system have been a subject of many studies. Ithas been shown that a naturally induced T-cell responserecognizing SCCHN exists that could potentially target andpossibly kill the tumor cells. Most cases in which this mayhave happened will remain obscure because they neverbecome visible. Those cases that become clinically apparentshow a different constellation. On one hand, antitumorimmune effects can be observed; on the other, a deleteriouseffect on immune cells is exerted by the tumor. Tumorprogression itself is therefore invariably linked to selectiveand pervasive impairment of immune cells.

The identification and characterization of a variety ofhuman tumor antigens with possible use for immunotherapyand immunomonitoring [7] and expectations triggered by


successful in vitro tests and therapeutic results in animalmodels have led to rapid translation of these experimentalfindings into clinical testing. This resulted in a ratherlarge number of patients with various types of malignantdiseases having been enrolled in clinical trials of T cell-based immunotherapy. In many cases tumor antigen-specificCTL immune responses could be achieved and successfullymonitored, but unfortunately these findings did not cor-relate with clinical responses [8]. True clinical responsesattributable to immunotherapy have been sparse so far. Thisdiscrepancy has initiated investigations into mechanismsunderlying the failure of tumor antigen-specific CTL tocontrol tumor growth in cancer patients, especially thosetreated with immunotherapies. The mechanisms responsiblefor this impairment may vary depending on the nature of thetumor milieu and manifest in the tumor microenvironmentas well as in the periphery [9, 10].

For the development of strategies to prevent or reversetumor-induced effects and to protect immune cells inthe tumor microenvironment, studies of both (a) nat-urally occurring immune cells that could be recruitedfor immunotherapeutic strategies targeting specifically thetumor cells and (b) direct and indirect mechanisms respon-sible for dysfunction and death of immune cells in SCCHNare essential [11, 12].

This review focuses on the mechanisms of tumor-mediated interference with the host immune system andCTL in particular concerning SCCHN. We describe tumor-escape mechanisms from the immune system at the tumorsite and in the periphery and describe strategies to redirectthe immune system to a more effective antitumor response.

2. Distinct Etiologies of Head and Neck Cancer

Two distinct causes of SCCHN are known. SCCHN is eithercaused by the spontaneous accumulation of multiple geneticalterations modulated by genetic predisposition and chronicinflammation, enhanced by environmental influences suchas tobacco and alcohol abuse or by infection with oncogenichuman papillomavirus (HPV). Carcinogens are regardedthe most important factors. Thus, two main etiologies canbe defined: tumors induced by toxic substances or by theactivity of the viral oncogenes of HPV. Both etiologies involvea multistep process and result in alterations affecting twolarge groups of genes: oncogenes and tumor suppressorgenes. However, it has been demonstrated that another set ofgenes is also altered. These are related to immune responsemechanisms like the MHC complex. This observation isimportant because it shows that immune recognition isrelevant and probably a force driving selection in the tumortowards immune-resistant variants (see below).

HPV-associated SCCHN defines a distinct subgroupthat could specifically be targeted by novel preventive ortherapeutic measures. More than 100 human papillomavirus(HPV) subtypes are known to date. Of these, 15 have beenshown to be oncogenic in humans. HPV type 16 and 18 seemto play the major role in the etiology of HPV an associatedSCCHN, particularly those arising in the oropharynx [13,14]. HPV infection has been detected in 20% to 30% of

tumors located in all head and neck anatomic subsites and inabout 50% of tonsil squamous cancers. For laryngeal cancer,the role of HPV is less clear. Data on the prevalence of high-risk HPV-associated SCCHN vary and multicenter studieshave not been performed yet [15, 16].

The lower rate of carcinogenic-risk factors and p53mutations and a younger patient population suggests thatfactors, currently unknown, are associated with viral entry,propagation/transformation, and immune evasion in HPV-associated SCCHN patients [14, 17]. Failure to clear HPVinfection leaves host cells under the influence of theviral oncogenes. Persistently infected persons can developclinically or histologically recognizable precancers that canpersist and may develop over time into invasive cancer.These oncogenes are vital to the tumor cell survival andproliferation and therefore provide a suitable target for anti-tumor vaccination.

Therefore, with these two different etiologies, differenttreatment options according to the genesis of their malig-nancy may be developed for future patients. Immunogenictumors in patients mounting specific immune responsesmay be treated by induction or enhancement of specificCTL responses. Such responses may have clinical impact forprimary therapy or in adjuvant settings when tumor burdenhas been reduced.

3. Immunoresponses in SCCHN

3.1. Tumor-Specific T Cells. Effective antitumor responses inindividuals with SCCHN depend on the presence and func-tion of immune cells that are able to recognize and eliminatetumor cells. These tumor antigen-specific T cells are knownto be present in the peripheral circulation and tissues ofpatients with cancer. They can be monitored by multicolorflow-cytometry directly using tetrameric peptide-MHC classI and II complexes, so called tetramers. These moleculesbind to the cognate T cell receptor. These stainings arecombined with T cell markers (e.g., CD3, 4, 8) and if desiredwith makers for the differentiation (e.g., CCR7, CD45RA)and the functional (e.g., CD107a, perforin, granzyme B) ordysfunctional (e.g., annexin, 7-AAD, CD3-zeta-chain) statusof these cells.

A number of studies have investigated the frequency oftumor-specific T cells in the peripheral circulation of patientsand in the tumor [9, 10, 18–24]. In most of these studies,T cells reactive against p53-derived epitopes by HLA-class-I and II were described. The special interest in p53-derivedepitopes can be explained by the fact that the majority ofhuman cancers, including head and neck cancer, seem to“overexpress” this protein. Other studies have focused on Tcells specific to HPV-derived epitopes [25].

Wild-type (wt) sequence p53 peptides like other tumorepitopes are processed and presented to the host immunecells either directly by the tumor cells or by professionalantigen presenting cells (APCs) such as dendritic cells(DC). This results in an increased number of wt p53peptide-specific T cells and, in some instances, p53-specificantibodies [21, 26, 27].


It could be demonstrated that the frequency of tetramer+

CD8+ T cells specific for the HLA-A2.1-restricted wtp53264−272 peptide is significantly higher in the peripheralcirculation of HLA-A2.1+ patients with head and neckcancer than that in normal donors [27]. In a subsequentstudy, the frequency of wt p53 epitope-specific T cells fortwo distinct epitopes was determined in the peripheralcirculation and in tumor infiltrating lymphocytes (TIL)isolated from SCCHN. Wt p53 epitope-specific T cells werefound to be significantly enriched in the TIL, demonstratinga preferential localization at the tumor site or in tumor-involved lymph nodes [9]. Interestingly, the presence andfrequency of wt p53 epitope-specific effector T cells amongTIL did not correlate with tumor stage. This implies thatthe frequency of tetramer+CD8+ effector cells alone has noeffect on tumor progression. In this study, two patients withsufficient numbers of TIL were available to test in vitroresponsiveness after polyclonal stimulation with anti-CD3mAb. Only a low IFN-γ expression of the CD3+CD8+ T cellscould be measured indicating a poor responsiveness to thisstimulus. At the same time, a significantly increased numberof regulatory T cells (Treg) were found at the tumor sitecompared to the periphery [9]. It has been well acceptedthat the presence and accumulation of Treg inhibits T-cell responses in vivo and may be responsible in part, fordownregulation of antitumor immune responses in patientswith head and neck cancer [28]. Treg are likely to mediatesuppressive effects directed at self-reactive T cells (see belowin detail) [29]. This immunosuppressive mechanism may beparticularly relevant to T cells that recognize self-peptideslike wt p53 epitopes and thus are likely to be tolerized,especially at the sites of their accumulation in tumortissues. Other data also confirm the depressed functionalityor even spontaneous apoptosis of CD8+ tumor-specificT lymphocytes [10]. To date, the molecular mechanismsdriving spontaneous apoptosis of circulating epitope-specificT cells in patients with cancer are unknown. Differentand from each other independent factors and mechanismslike the Fas/FasL pathway, propriocidal lymphocyte death[30], or tumor-derived factors responsible for disruption ofsignal transduction pathways may play a role. The recentdiscovery of FasL+ microvesicles in the circulation of patientswith head and neck and other cancers also suggests apotential mechanism for systemic elimination of CD95+

activated CD8+ T cells. However, neither the tumor originof microvesicles present in patients’ sera nor their in vivoparticipation in T cell apoptosis has been confirmed.

A subgroup of head and neck cancer is associatedwith human papilloma virus. Virus-derived antigens areconsidered superior targets for T cells than tumor-associatedself-antigens because they have higher affinity to MHC andare more immunogenic. As opposed to tumor-associatedself-antigens, thymic negative selection of T cells recognizingthese virus derived antigens has not reduced the pool fromwhich tumor reactive T cells can be recruited. Therefore,HPV-encoded oncogenic proteins, such as E6 and E7, arepromising tumor-specific antigens in addition to the factthat they are considered obligatory for tumor growth.Surprisingly, few studies have characterized endogenous T

cell immunity to HPV-encoded oncogenes in SCCHN as aprerequisite for immunotherapeutic targeting of these anti-gens. Two studies show an increased frequency of CD8+ Tlymphocytes directed against HPV E7 epitopes documentinga natural immune response [18, 22]. These HPV-specificT cells were able to recognize and kill a naturally HPV-16transformed SCCHN cell line after IFN-γ treatment thatenhanced antigen processing and presentation by the tumorcells. Further phenotypic characterization of the HPV-specific T cells revealed an increase in terminally differenti-ated/lytic T cells (CD8+CD45RA+CCR7−). This populationwas also characterized by a high frequency of staining forthe degranulation marker CD107a in E7 tetramer+ T cells,compared with bulk CD8+ T cells, consistent with theirterminally differentiated lytic, degranulated status. Thesecells may account for the unsuccessful antiviral immuneresponse [31] to these tumors, indicating that incompleteactivation of tumors-pecific T cells or suboptimal targetrecognition may enable tumor progression in vivo. On theother hand, if T cells from this reservoir could be adequatelyactivated and expanded, these cells could provide a valuablereservoir of effectors for cancer vaccination.

Several current studies suggest that cancer vaccineshave increased efficacy if they incorporate tumor-specificcytotoxic as well as helper epitopes. Although CTL areconsidered to play the primary role in tumor eradication, itis also hypothesized that the participation of tumor antigen-specific CD4+ T-helper lymphocytes may be required foroptimal antitumor effects by generating and maintainingantitumor immune responses through interactions with CTLand other cells [32, 33]. As a result, efforts have been madeto define class II HLA-restricted tumor peptides for usein cancer vaccines. As for CTL defined epitopes, wild-typesequence (wt) p53 peptides also provide a source for CD4+ T-helper cells [20, 34]. By ex vivo experiments, performed in anautologous human system, the ability of anti-wt p53110−−124

CD4+ T cells to enhance the generation and antitumorfunction of CD8 effector cells was demonstrated. The resultsemphasize the crucial role of T helper-defined epitopesin shaping the immune response to multiepitope cancervaccines targeting p53. This suggests that future vaccinationstrategies targeting tumor cells should incorporate helperand cytotoxic T cell-defined epitopes [34].

4. Mechanisms of Tumor Evasion

4.1. Suppression of T Cells in the Cancer-Bearing Host.Homeostasis of lymphocytes, that circulates through thetissues and the blood, is maintained by refreshing the poolvia the thymic output of naive lymphocytes and expansion ofantigen-specific lymphocytes upon adequate stimulation andcontraction of the lymphocyte pool by death of lymphocytesin the periphery that have completed their functions [35–37].

One of several mechanisms by which tumors escapefrom the host immune system is induction of apoptosis ineffector T cells [12]. It was shown that a proportion ofCD3+, Fas+ T cells in the peripheral circulation were inthe process of apoptosis. This indicates that the Fas/FasLpathway is involved in spontaneous apoptosis of circulating


CD95 (Fas+) T lymphocytes. Interestingly, these cells showeddecreased expression of CD3-zeta-chain. Expression of Tcell receptor- or Fc-gamma receptor III-associated signal-transducing zeta chain is important for the functionalintegrity of immune cells [31]. Fas/FasL interactions mightlead to increased turnover of T cells in the circulation and,consequently, to reduced immune competence in patientswith SCCHN [38]. This may be explained by an imbalance inthe absolute counts of T-lymphocyte subsets and an overalldecreased absolute T cell count in patients not treated withcytotoxic agents [39, 40]. The rapid turnover affects mostlyT cells with effector phenotype [41] that also show defectsin signaling [31]. Preferentially tumor-specific T cells areaffected by apoptosis indicating a tumor-related effect [10].This observation can be explained further by the analysisof TCR Vbeta profiles of CD8+ T cells in patients withSCCHN that were altered relative to normal controls. Thismay reflect increased apoptosis of expanded or tumor-contracted CD8+ T cells, which define the TCR Vbeta profileof antigen-responsive T-cell populations in patients withcancer [19]. Reports on T-cell apoptosis at the tumor siteand in the peripheral circulation [42, 43] support theseobservations and suggest that death of tumor-infiltratinglymphocytes (TIL), generally considered to represent tumor-associated antigen-specific effector cells, is driven by thetumor or tumor-derived factors. Recent studies in SCCHNalso demonstrate that T regulatory cells (Treg) expresshigh levels of Fas and selectively kill CD8+ T effector cellsvia Fas/FasL [44]. FasL is upregulated exclusively on Tregisolated from patients with no evidence of disease afterreceiving cancer therapy [44]. These FasL-expressing Tregare resistant to apoptosis themselves but strongly suppressand kill CD8+ effector cells, adverting the cancer communitythat traditional cancer therapy might contribute to tumorprogression by collaborating with the peripheral toleranceprocess. In addition, Treg in patients with SCCHN kill CD4+

T effector cells via granzyme B in the presence of IL-2 [44].Signaling defects in the TCR as well as NF-B activationpathways in TIL have been described in comparison to T cellsinfiltrating inflammatory noncancerous sites. These defectsappear to be responsible for their loss of function [45].Patients with tumors infiltrated by TIL expressing normallevels of CD3-zeta-chain were found to have a better 5-year survival than those showing loss of CD3-zeta-chainexpression [8, 46]. A high rate of apoptosis in TIL isconsidered to be a factor for poor prognosis [47].

Taken together it appears that apoptosis of lymphocytesin the periphery as well as at the tumor site leads to rapid andselective tumor-specific lymphocyte turnover followed by aloss of effector cells and thus failure to control tumor growthin cancer patients.

4.2. Role of Regulatory T Cells. Because of the identificationof the forkhead box transcription factor Foxp3 as an essentialtranscription factor in CD4+ regulatory T cells (Treg), Treghave been well characterized as a distinct lineage of T cells[48]. Their thymic origin (denominated naturally occurringTreg or nTreg) as well as their importance for the main-tenance of peripheral tolerance under noninflammatory

conditions throughout the life span of an individual, hasbeen demonstrated in men and mice [49, 50]. When Foxp3+

Treg are depleted in an adult individual, fatal multiorganautoimmunity finally results [51] and the phenotype of thisdisease is virtually indistinguishable from genetic deficiencyof Foxp3 that is characterized by a massive lymphopro-liferative syndrome [52, 53]. However, the conditions andmechanisms required to generate Foxp3+ Treg de novo in theperipheral immune compartment (denominated inducibleTreg or iTreg) or to selectively expand Treg in peripheralblood and lymphoid organs are less clear. Essential rolesof IL-2, TGF-β, and TCR signaling in iTreg generation,expansion, function, and survival have been established[54–56]. However, it is unclear which particular arm ineach of these pathways is important in FoxP3 regulation.Foxp3 expression is required for suppressor capacity of Tregin men and mice [57]. In addition, regulation of Foxp3expression in nTreg as well as iTreg in response to acuteand chronic inflammation remains an unresolved issue.Elucidating the signaling pathways that command Foxp3induction and expression in inflammation constitutes achallenge of particular interest as Treg induction, expansion,survival, and function are significantly altered in diseases inwhich inflammation is a key regulator of the pathology (i.e.,autoimmune disease and cancer).

It has become increasingly clear that malignant trans-formation and cancer progression are immunologicallyrecognizable events and the immunologic status of thehost as well as the inflammatory conditions of the tumormicroenvironment influences significantly the outcome ofthe patient [58]. In early stages, inflammation at thetumor site induces chemotaxis of immune cells from theperiphery to the tumor and immunologic recognitionmay exert selective pressure, braking tumor growth ofthe emerging cancer [59]. However, when the immunesystem is confronted with persistent exposure to tumorantigens the establishment of tolerance [60] and in con-sequence immunosuppression of the patient are favored.Much like tolerance to normal self-antigens, tolerance totumor antigens (TA) can arise from a failure to encounterantigen or the deletion or functional inactivation of tumor-specific T cells. It has been demonstrated during the last10 years that Treg frequency increases in peripheral blood,lymph nodes, and tumors of patients with several typesof cancer [61], including patients with HNSCC [62, 63].It also correlates with tumor progression and outcome[64]. Suppressor capacity and suppressor phenotype ofTreg isolated from SCCHN cancer patients are significantlyincreased in comparison to Treg isolated from healthysubjects [62, 63], suggesting that enhanced function andsurvival of suppressor cells might constitute one of themechanisms that are responsible for immunosuppressionof adaptive and innate immunity in these patients. Indeed,in several models, tumor immunosurveillance is augmentedwhen CD4+CD25+ Tregs are depleted [65, 66]. RemovingTreg has also been shown to increase tumor immunityelicited by vaccination [67].

Thus, one therapeutic possibility for restoration of anti-tumor immunity in patients with cancer is to eliminate


tumor antigen (TA)-specific Treg and to boost simultane-ously TA-specific T helper and CTL responses. The fact thatTreg and activated T effector cells share receptors and com-mon metabolites in their differentiation, function, survival,and expansion (i.e., IL-2) suggests that regulation of theeffector and suppressor compartment is dichotomic. Thus,one new challenge in modern immunotherapy is to under-stand the signaling pathways that command the interplay ofeffector and suppressor responses in physiologic conditionsand in inflammation. A detailed knowledge of these pathwaysmight enable us to design immunotherapeutic strategiesthat selectively promote expansion, survival, and function ofeffector or Treg responses in pathologies where one of the twocompartments is in disadvantage resulting in autoreactivekiller responses in the absence of Treg or in immunosup-pression in the case of an excessive Treg response. Revertimmunosuppression in cancer to anti-tumor immunity isessential to increase the quality and success of traditionalcancer therapy as well as the response to tumor vaccines.

Clinical trials for tumor vaccines using TA antigens andantigen presenting cells (APCs) or DC are now under way forthe treatment of different types of cancer. To date, more than1000 tumor vaccines have been reported [68]. The collectiveresults are encouraging but not satisfying. The study ofChakraborty et al. shows that the MAGE-specific CTLresponse in patients with melanoma contracts in circulationby day 28 after vaccination [69]. Multiple factors may explainthe decline of TA-specific CTL in circulation of patientsafter receiving a tumor vaccine. For example, the declinemight be a result of the activated CTL leaving the circulationand homing into extravascular sites. The contraction mightalso be a reflection of programmed cell death or activation-induced cell death as physiologic homeostatic process [69].However, the work of Zhou et al. suggests that contractionmight also result from a Treg cell regulated process. Theydemonstrate that vaccination of the tumor-bearing hostexpands Treg, blunting the expansion of naıve tumor-specificT cells and blocking the execution of effector function in vitroand in vivo [70]. The reports of Zhou et al. and Chakrabortyet al. suggest that the development of treatment paradigmsthat seek to not only increase the frequency of tumor-specificT cells but to do so in conjunction with strategies that selec-tively inactivate or remove suppressor T cells is a must. Thisobjective has become even more complex with the recentidentification of Th17 cells. Collectively, the amendmentof the Th1/Th2 paradigm by new subsets of T cells hasresolved some inconsistencies of the Th1/Th2 paradigm buthas also made the understanding of the pathogenic processof cancer inflammation more complex. The fact that Th17cell induction and differentiation is mediated by metabolitesthat are also essential in Treg development and functionsuggests that Treg-Th17 cell interactions might influencethe induction process of peripheral tolerance. Indeed, somerecent studies report that Th17 cells increase simultaneouslywith Treg in cancer subjects and this dichotomic increase cor-relates with cancer progression [71]. Thus, a new challenge incancer immunology has become to elucidate the populationdynamics and kinetics of Th17 and Th1 cells. Their interplayand susceptibility towards regulatory mechanisms at the site

of inflammation and in the periphery possibly defines phase-specific approaches for therapeutic interventions in order toprevent cancer–inflammation-mediated tolerance.

First studies investigating the dynamics of Treg andeffector responses indicate that molecular signaling pathwaysin response to IL-2, TGF-β1, and the TCR are determinantin regulating homeostasis of suppressor cell and effector cellpopulations [54–56, 79]. Therefore, a detailed knowledge ofthese pathways might provide valuable insights on how Tregmight be regulated to support tumorrejection.

IL-2 plays a dominant role in vivo, in the maintenanceof immune system homeostasis and self-tolerance. The latterfunction is emphasized by the finding that mice deficient inIL-2 or components of the IL-2 receptor (IL-2Rα or IL-2Rβ)succumb to lymphoproliferative autoimmune syndrome,with the effect of IL-2Rβ lack being more severe [80, 81].TGF-β1 is a pluripotent cytokine that has pronounced effectson T cell-mediated immune suppression as well as on thecontrol of autoimmunity. The role of TGF-β is in influencingthe constitutive expression of Foxp3. IL-2, the transcriptionfactor NFκB, might be essential in regulating Treg and Teffector cell homeostasis in conditions of inflammation.Elucidating the role of TGFβ and NF-κB in Treg and Teffector cell dynamics is promising because both are up-regulated during cancer inflammation and are commonregulators of Treg and Th17 cell differentiation. The designof studies that elucidate these signaling pathways in patientswith cancer constitutes one step towards immunotherapeuticstrategies that enforce immunogenicity of tumor vaccines.

Taken together, these complex immunoregulatory mech-anisms lead to an immunoediting of the T cell and inparticular the CTL response by the tumor in order to avoidelimination by the immune system. However, also the tumorcell immunogenicity is edited by the immunoresponse.

4.3. Tumor-Immune Escape. Tumor-T cell interaction leadsto a negative selection pressure on tumor cells that arebeing recognized by the immune system. T cell recognitionis therefore reduced in tumor cells by downregulation ofmolecules important for antigen processing and presentationor costimulation [82].

Several observations demonstrate an immune selectionof SCCHN tumor cells as discussed below. This selectionprocess and the resulting immune escape variants in thetumor indicate that an effective CTL response must havetaken place during the development of the malignancy. TheCTL-mediated cytolysis of immunogenic tumor cells is thedriving force of the selection process towards non-CTL-susceptible tumor cell variants. The immune-evaded tumorcells have several features making them resistant to furthernatural CTL attack.

(a) Reduced expression of costimulating molecules onthe cell surface leads to inadequate T-cell activation,even if the tumor is recognized and in turn totolerance induction [83].

(b) Expression of MHC class I is altered in up to 50% ofSCCHN [84–86]. Both, total loss of HLA-class I andmore selective downregulation of the HLA-A, B, or Clocus expression has been shown [87].


Table 1: Tumor escape and potential reversal strategies by adjuvant treatment options.

Tumor evasion mechanism Desired effect Potential reversal strategy

Loss of intracellular proteasomal antigenprocessing, transport (TAP deletion) andMHC-loading (beta2-microglobulindeletion)

Restoration of antigen-processing andMHC-loading, for sufficient tumor-antigenpresentation

Interferon-γ treatment of the tumor

Silencing of MHC genes Restoration of MHC-expressionInterferon-γ treatment, application ofhypomethylating agents

Loss of T cell costimulation (e.g., CD80/86,and CD54, CD58)

Restoration of costimulatory moleculesToll-like receptor stimulation, interferontreatment

Unfavourable microenvironment forCTL-response

proinflammatory microenvironment forCTL-response

Application of immune response modifiers,suitable vaccine adjuvants, and induction ofCD4 T helper cells

Role of T cells

Too few tumor-specific T cellsInduction of more CTL with lytic activity,broader T cell response including CD4T-helper cells

Specific CTL stimulation and expansion.Vaccination with single or multiepitopevaccines including MHC class I and IIpeptides. Induction and expansion of CD4T-helper cells.

Loss of immunodominant tumor antigenDirection of the immune response to otherantigens or epitopes

Identification of optimal MHC-class I and IIepitopes. Reexpression of the tumor-antigen

Suppressive Treg effects Inhibition of deleterious T-cell effects

Modulation/reduction of Treg bypretherapeutic treatment with antibodies orpreferentially Treg targetingchemotherapeutic agents

Tumor-induced T cell apoptosis Rescue of apoptotic T cells

T cell protection by:(i) reversal of redox potential(ii) treatment with anti-apoptotic drugs(iii) blocking of proapoptotic molecules(e.g., CD95)

(c) Expression of components of the antigen processingmachinery (APM), namely, LMP2, LMP7 and TAP1is frequently downregulated or completely lost intumor lesions as compared to surrounding tissue aswell as in cell lines obtained from SCCHN cancerpatients. As a consequence of the downregulatedAPM components, fewer antigen is processed andloaded onto MHC-complexes that are by themselvesdecreased in number or dysfunctional.

In combination, these three mechanisms compromiserecognition of the progressed tumor by tumor-specific Tcells. In cell lines, the expression of APM-componentscould be restored by incubation of with IFN-g [18, 88].Furthermore, LMP2, LMP7, TAP1, TAP2, and HLA class Iantigen expression rates in primary SCCHN lesions werefound to predict overall survival [88]. Since expression ofAPM-components could be functionally restored, structuralabnormalities such as genetic alterations are unlikely. Withregard to the two different etiologies of SCCHN (onebeing alcohol and tobacco abuse and the other oncogenichuman papillomavirus), more detailed studies are neededto investigate if this dysregulation can be observed intumors with both etiologies. Whether this is a generalphenomenon, as has been reported in other tumor typeswithout viral etiologies, or is due to HPV specific factors,

as has been suggested in HPV-6- and HPV-11-associatedlaryngeal papilloma [89], remains to be clarified. So far, onlyin one of our studies HPV-typing was carried out [18].

Tumors can also interfere with the immune system byproducing and releasing numerous factors that modulatefunctions of immune cells or directly induce apoptosis. Thesefactors take action in the tumor microenvironment andbeyond.

5. Vaccination Strategies Aiming atInduction of Cytotoxic CTL and ReversingImmune-Escape Mechanisms

For effective vaccination, a successful stimulation of theimmune system and an effective modulation of suppressiveeffects exerted by the tumor cells are necessary (Table 1).

5.1. Role of APM and Abnormal MHC Class 1 Expression.Downregulation of dysfunction of APM components by thetumor may disturb both the induction of tumor-specific Tcells in the initial phase of the immune response and sub-sequently during the effector-phase the proper recognitionof the tumor. This effect is augmented by absent or reducedpresence of MHC class-I molecules on the cellular surface.These cells are considered to have a more aggressive pheno-type [90] which may also be the result of immunoselection of


geneticalteration

HPVinfection

Heterogeneousvariants

Metastasis HLA loss

Mutagens

Selection pressure

HLA negative

Escape variantsEffective lysis!

Proliferation,- Immunity- Immune evasion

Regrowth withHLA loss,

APC-dysfunction

T-cellAPC-dysregulation

Mutagen (e.g., tobacco) induced genetic alterations that accumulate result in tumor initiation and progression over time

(e.g., tobacco)

HPV persistence, together with genetic alterations, results in tumor initiation and progression over time

Figure 1: Immune escape variants of cells in the tumor indicate an effective T cell response. HPV infection leads to unregulated cellproliferation and accumulation of chromosomal aberration. Cumulative genetic alterations in tumor cell subclones lead to the emergence oftumor cell variants with divergent characteristics, for example, loss of HLA expression. Selection pressure is exerted by the microenvironmentof the tumor and immune response mechanisms. Over time, susceptible cells will be eliminated and resistant cells will regrow, to form atumor consisting of predominantly immunoresistant cells and compromising immunotherapeutic strategies.

tumor cells able to evade the immunosurveillance (Figure 1).The result can be seen by a low number of tumor infiltratinglymphocytes and ineffective generation, activation, or evenenhanced apoptosis of tumor-specific T cells [9, 10, 18, 91].

In experimental systems, incubation of SCCHN cell lineswith IFN-γ was able to restore T cell recognition and killing[18, 88]. These preliminary data should inspire more basicand clinical research to better understand and further refineand develop these adjuvant strategies for clinical application.From the current point of view, it seems indispensable tocombine APM- and MHC-class-I restoration with inductionof tumor-specific immune responses.

5.2. Apoptosis of T Cells. Tumor-induced increased apopto-sis of lymphocytes, including NK-cells and tumor-specificCD8+ T cells in the peripheral circulation and at thetumor site, has been observed [10, 38, 41–43, 92]. Thisphenomenon may lead to depletion of the lymphocyte poolfrom which tumor-specific effector cells could be recruitedor expanded by immunotherapy. Therapeutic approachesshould, therefore, aim to reverse this effect by restoring anormal lymphocyte turnover or protection of CD8+ T cellsfrom apoptosis [93].

5.3. Monitoring and Targeting Treg Responses in Patients withCancer: Therapeutic Relevance. The observation that Tregare increased in peripheral blood and at the tumor site inpatients with cancer suggests that the development of treat-ment paradigms that seek to not only increase the frequencyof tumor-specific T cells but to do so in conjunction withstrategies that selectively inactivate or remove suppressor Tcells is a must. However, the monitoring of Treg frequencyand function in patients with cancer is not only requiredto improve the success of cancer vaccines and traditionalcancer therapy but might also provide a broader basis for the

development of more reliable prognostic factors, improvinga tumor-free outcome after therapy. On the other hand,selective depletion of Treg in patients that have receivedtraditional cancer therapy might be essential to avoid ordecrease recurrent disease. We have shown that hema-tologic recovery in response to oncologic therapy resultsin expansion of the CD4+CD25+ compartment, includingCD4+CD25highFoxp3+ [62]. Importantly, not only was theproportion of CD4+CD25+ T and CD4+CD25highFoxp3+

cells increased in the patients after receiving oncologictherapy, but suppressor function and survival of thesecells were significantly elevated. The observed expansion ofTreg expressing Foxp3 correlated with increased suppressionmediated by these cells in patients with non evident disease[62]. The effect of oncologic therapy on Treg might berelated to two phenomena: (i) homeostatic regulation afterradio/chemotherapy-induced lymphopenia resulting in theexpansion of a total lymphocyte pool, and (ii) activation andexpansion of de novo induced Treg and T responder cells byinflammatory cytokines derived from the strong inflamma-tory response that usually accompanies radio/chemotherapy.The immune system controls the level and the activationstate of each cellular compartment through homeostaticregulation, a process that is triggered during developmentand after the induction of a lymphopenic state by externalstimuli [94]. Radio/chemotherapy may have profound effectson the peripheral blood cell count, due to an increasein the availability of homeostatic cytokines and increasedinteractions of T cells with APC. It has been proposedthat lymphodepletion removes endogenous cellular elementsthat act as sinks for cytokines, which are responsible foraugmenting the activity of tumor-reactive T cells [95].Thus, T-cells surviving after therapy receive very strongstimuli (cytokines and enhanced APC-T cell interactions)that trigger T-cell activation and expansion. Several groups


CSC

CSC

CSC

CSC

CSC

CSC

Surgery

Chemotherapy Radiotherapy

CTL

CTLCancer stem cell Cytotoxic T-cell

derived from CSC, hESCs, iPS,

-epitope loss

-downregulation

-of MHC class I and II

-antigen processing machinery components

Possible negative modulation of the

immune response by cellular factors

microenvironment

Possible induction of apoptosis in

effector T-cells

Tumor recurrence

Tumor regression

Conventional

therapies

CSC-targeted

immunotherapy

TGF-β, IL-10 · · · ) in the tumor

“Bulk” tumor cells

CTL target CSC-specific epitopes

stemness markers and signaling pathways

Possible immune escape by

-costimulatory molecules

(e.g., Treg) and soluble factors (e.g.,

Figure 2: Comparison of the effects of a failed conventional therapy and the outcome of a hypothetical CSC-targeted immunotherapy.Currently applied conventional therapies target bulk tumor cells that are less resistant than CSC. This leads to initial shrinking of the tumormass but eventually regrowth from residual CSC. An immunotherapeutic approach targeting CSC directly would cut off the rejuvenatingsupply of CSC and ultimately lead to tumor regression.

Table 2: Examples for studies targeting CSC with CTL.

Target Tissue Ref

Dendritic cells loaded with CSC as antigen source glioblastoma [72]

CD8 defined ALDH1-specific T-cell epitope HNSCC [73]

Vaccination with murine prostate stem cell antigen encoding cDNA Murine prostate cancer [74]

Dendritic cells loaded with neurospheres from brain glioma cells Murine glioma [75]

Identification of 2 CD8 defined prostate stem cell antigen-specific T-cell epitopes Prostate cancer [76]

Vaccination with defined human embryonic stem cells (hESCs) or induced pluripotent stem (iPS) cells Colon cancer [5]

CD8 defined SOX2-specific T-cell epitopes Glioma [77, 78]

have reported that antitumor responses seen after adoptivetransfer of tumor-reactive T cells into lymphodepletedhosts are significantly increased [96, 97]. In this context,it is reasonable to suggest that a similar mechanism (s)is involved in activation and expansion of the activated Tresponder and memory T-cell compartments in SCCHNpatients who receive oncologic therapy. Concomitantly, theTreg subset is significantly expanded. These findings suggestthat monitoring of Treg frequency and function as well asdepletion of Treg after oncologic therapy might be crucial

to allow the development of an effective antitumor T-cellresponse able to eliminate secondary tumors.

5.4. Single- and Multi-T-Cell Epitope Vaccines. p53 may serveas a model antigen for the development of broadly applicableantitumor vaccines in SCCHN. A number of p53-derivedepitopes that can be used for the design of vaccines havebeen identified [98, 99]. Since mutations in the p53 sequenceare frequent [100], epitopes incorporating these mutationswould have to be tailored specifically to each patient.


Therefore, epitopes composed of the wild-type (wt) sequenceare especially attractive, since they are shared among thesame HLA type and are therefore not patient specific.

In vitro stimulation of CD8+ T cells with wt p53 peptide-pulsed autologous dendritic cells can be used to induceeither HLA-A2-restricted, wt p53149−157 and/or wt p53264−272

peptide-specific responses from epitope-specific precursors.Interestingly, using these single epitopes, wt p53 peptide-specific CD8+ T cells were generated in only a third of healthydonors or subjects with cancer [27]. Others have reportedcomparable findings [101, 102]. The limited responsivenessof healthy donors may be explained by negative thymicselection of T cells with receptors specific for self-antigens.It can be expected that especially T cells with high-affinityreceptors are eradicated. The observed limited responsive-ness to HLA class I-restricted wt p53 peptides among HLAclass I-compatible healthy donors and subjects with cancersuggests that multiple wt p53 peptides are needed in order tomaximize donor responsiveness. The underlying causes canonly be suspected and may partly be due to the mechanismsof tumor immune evasion discussed above.

Since it may prove difficult to determine in an individualcase the responsiveness prevaccination, a vaccine consistingof more than one epitope may be the more promisingapproach.

5.5. Immune Intervention Approaches For HPV-AssociatedSCCHN. Oncogenic HPV genotypes, particularly the HPVtypes 16 and 18, are found in a subset of SCCHN mostlylacking the risk factors of alcohol and tobacco abuse. Ittherefore seems to be a suitable malignancy for vaccinationagainst HPV-associated targets. Thus, these HPV-associatedSCCHN may be preventing by (a) the existing prophylacticHPV-vaccines or (b) treated by vaccines designed to inducean appropriate antitumor immune response against HPV-specific tumor antigens [18, 22]. Lately, several authorshave advocated a combined approach of prophylactic andtherapeutic HPV-vaccination in patients with dysplasia andrisk for reinfection with oncogenic HPV-types [103, 104].

(a) Prophylactic HPV Vaccines. Neutralizing antibodies spe-cific for the viral capsid proteins may prevent infection byHPV. Prophylactic HPV vaccines have been developed andapproved. As antigen, they contain capsid protein L1 ofthe most prevalent HPV types. They were introduced in2006 and have the estimated potential to reduce the burdenof cervical cancer, the tumor entity they were originallydeveloped for, by approximately 70% [105, 106] and thevaccine type-related precancers. Estimations for SCCHN arecurrently not available. Prophylactic HPV vaccines have,however, no therapeutic potential due to their mechanismof action via virus-neutralizing antibodies targeting the L1capsid antigen. Unfortunately, this capsid protein is notexpressed in persistently HPV-infected basal epithelial cellsand transformed cells in infected mucosa and is thereforeuseless for therapeutic vaccination. Accordingly, prophylacticvaccines have not shown any therapeutic activity [107].

However, they may be of benefit in posttherapeutic sit-uations where infected lesions have been removed surgically

to prevent formation of new lesions due to reinfection [104].These settings are currently under investigation.

(b) Therapeutic HPV Vaccines. The rationale for vaccinestargeting HPV-associated SCCHN is that virus-relatedoncogenes are obligatory to tumor growth. Vaccines withtherapeutic potential must target the two HPV oncogenicproteins, E6 and E7 as antigens that are important inthe induction and maintenance of cellular transformationand are coexpressed in the majority of HPV-associatedcarcinomas. Two studies have investigated if an endogenousT-cell immunity to HPV-encoded oncogenes E6 and E7 inSCCHN patients exists [18, 22]. This group of T cells wouldhave the potential to specifically identify and target the tumorupon appropriate activation. Therefore, these cells are acritical prerequisite for the development of vaccine-basedstrategies for enhancing antitumor immunity in patientswith HPV+ tumors. Indeed, in both studies it was found thatinfection with HPV-16 (as compared to uninfected controlindividuals) significantly alters the frequency and functionalcapacity of virus-specific T cells in SCCHN patients. Inaddition to the presence of HPV-specific effector T cells, suc-cessful tumor elimination requires that HPV-infected tumorcells function as appropriate targets for cytotoxic T lympho-cyte recognition and elimination. Immunohistochemistryof HPV-16+ SCCHN tumors showed that the antigen-processing machinery components are downregulated intumors compared to adjacent normal squamous epithelium[18]. Thus, immunity to HPV-16 E7 is associated with thepresence of HPV-16 infection and presentation of E7-derivedpeptides on SCCHN cells, which shows evidence of immuneescape comparable to cervical cancers [108]. These findingssuggest that development of E7-specific immunotherapy inHPV-associated SCCHN should be combined with strategiesto enhance the antigen processing machinery componentexpression and function [18].

6. T-Cell Therapies Directed toCancer Stem Cells

Tumors consist of heterogeneous populations of cells.According to the cancer stem cells (CSCs) hypothesis, CSCsare a subpopulation of the tumor more capable than othercells to self-propagate, initiate new tumors differentiateinto bulk tumor, and therefore sustain tumor growth.Because of these properties, CSCs have been moved into thefocus of targeted therapies. The current knowledge of theexistence of CSCs begins to lead to studies of their specificelimination (Figure 2 and Table 2). It is being envisioned thatthe targeting of CSC in combination with the establishedtherapeutic modalities such as radiation and chemotherapythat due to a relative resistance of the CSC more preferentiallykill the bulk of the tumor may decrease the frequency ofrecurrences and enhance the patient’s long-term survival.Therefore, the development of strategies that target the CSCpopulation directly is highly desirable. Eliminating CSC leadsto an abrogation of the replenishing pool of cancer cells andultimately leads to petering out the tumor growth, as hasbeen documented in animal experiments where removal of


CSC and transplantation of only the non-CSC tumor cellsdid not lead to sustained tumor growth.

The development of CSC targeted therapy has to over-come three major hindrances, that relatively to the bulktumor population are increased (i) chemoresistance, (ii) resi-stance to radiotherapy, and (iii) immunescape mechanisms.

Since radio- and chemotherapies have already beenoptimized towards the limits of clinical benefit and yettolerable side effects, a very attractive alternative approachof specifically targeting CSC is to develop antitumor T cellvaccines. One of the possible reasons that these therapieslacked efficacy in past clinical trials could be attributedto the fact that rather bulk tumor than CSC have beentargeted. This may change with the identification of tumor-specific epitopes derived from CSC markers. One such asCSC model-target for head and neck cancer and othersis the recently described CD8 defined T-cell epitope ofaldehyde dehydrogenase-1 (ALDH1) [73]. Examples of othersuch CD8 defined T-cell epitopes are available for prostatestem cell antigen [76]. Less well-defined approaches includethe development of a CSC-dendritic cell vaccine [72].Recent studies using animal models for prostate cancer andmalignant glioma demonstrated the potential of differentvaccination strategies (dendritic cells, gene-gun and virus)targeting CSC in cancer immunotherapy [74, 75]. It wassuggested recently that stemness related proteins expressedin CSC might also be a source for tumor antigens. Tumortypes most dependent on CSC for their growth kinetics werenamed to be the best suited for approaches targeting stem cellgenes [109].

In several studies, the efficacy of potential therapiesdirected against stem cell-associated signaling pathways aretested. For example, T cell immunity against embryonicstem cells antigen SOX2 and SOX6 has been explored inglioma stem cells [77, 78]. Since the expression of stemness-related genes is a common feature of stem cells and CSC,the question of vaccine induced autoimmunity to stem cellswill have to be addressed by scientists following this path.One example is the vaccination with embryonic stem cells(ES) or induced pluripotent stem cells (iPS) that has beenshown to induce protective immunity in colon carcinoma[110]. Another group used dendritic cells (DC) generatedfrom mouse and human ES or iPS as a means for anti-cancerimmunotherapy [111].

Success of these potential therapies will depend on howwell immunological responses to CSC can be modulatedfor example by vaccine adjuvants upregulating antigenprocessing and presentation. Recently, a reduced activity ofthe 26S proteasome in breast cancer cells and in gliomaswas observed as a feature of CSC [112]. Proteasomes arethought to play an important role in antigen processingand presentation of antigens in association with HLA classI [113]. This may result in reduced antigen processing andpresentation of peptides presented to the immune systemon major histocompatibility complex -I molecules. Reducedproteasomal activity was also used as explanation for the highexpression levels of known stem cell markers like BMI-1 andnestin in CSC, which are both substrates of the proteasome[114, 115].

7. Significance of CSC for FutureTreatment Strategies

The classification of conclusive CSC markers followed by theidentification of defined T cell-recognized CSC epitopes inthe future may lead to the clinical application of anti-CSCvaccination strategies. Several approaches are currently beingevaluated (Table 2). Whether targeted therapies directedagainst stem cell-associated signaling pathways, which maybe activated in stem cells and in CSC, will be of clinical useor be limited by undesirable side effects in vivo remains sofar unresolved.

8. Conclusion

Immunotherapy is considered an attractive treatment moda-lity because it specifically targets the cancer avoiding or mini-mizing side effects. Ideally, its therapeutic effects will alsoreach distant metastasis and will be sustainable lasting beyo-nd the presence of the cancer due to immunologic memory.

Therapeutic strategies should consider that SCCHN hasat least two distinct etiologies. One is chronic alcohol andtobacco abuse and the other is related to oncogenic humanpapillomavirus infection and transformation. Both etiologieswill differ significantly in the antigenic make up of thetumor cells based on presence of self or viral antigens,respectively. In both cases, however, immunotherapeuticapproaches should aim at induction of adequate antigenprocessing and presentation by the tumor cells to becomevisible for the immune system as target. Furthermore, tumorinduced immune dysregulation should be redirected in favorof tumor rejection and finally an adequate stimulation ofeffector T cells capable of in vivo expansion and survivalin the tumor-microenvironment is thought critical forimproving clinical results.

Abbreviations

APC: Antigen presenting cellDC: Dendritic cellIFN-g: Interferon-gammaSCCHN: Squamous cell carcinomas of the head and

neckHLA: Human leukocyte antigenMHC: Multihistocompatibility complex.

References

[1] A. Mashberg, N. S. Russell, H. Bartelink et al., “Head andneck cancer,” The New England Journal of Medicine, vol. 328,no. 24, pp. 1783–1784, 1993.

[2] E. E. Vokes, R. R. Weichselbaum, S. M. Lippman, and W. K.Hong, “Medical progress: head and neck cancer,” The NewEngland Journal of Medicine, vol. 328, no. 3, pp. 184–194,1993.

[3] D. M. Parkin, F. Bray, J. Ferlay, and P. Pisani, “Global cancerstatistics, 2002,” Ca: A Cancer Journal for Clinicians, vol. 55,no. 2, pp. 74–108, 2005.

[4] R. T. Greenlee, M. B. Hill-Harmon, T. Murray, and M. Thun,“Cancer statistics, 2001,” Ca: A Cancer Journal for Clinicians,vol. 51, no. 1, pp. 15–36, 2001.


[5] A. S. Jones, P. Morar, D. E. Phillips, J. K. Field, D. Husband,and T. R. Helliwell, “Second primary tumors in patients withhead and neck squamous cell carcinoma,” Cancer, vol. 75, no.6, pp. 1343–1353, 1995.

[6] C. R. Leemans, R. Tiwari, J. J. P. Nauta, I. Van der Waal, andG. B. Snow, “Recurrence at the primary site in head and neckcancer and the significance of neck lymph node metastases asa prognostic factor,” Cancer, vol. 73, no. 1, pp. 187–190, 1994.

[7] L. Novellino, C. Castelli, and G. Parmiani, “A listing ofhuman tumor antigens recognized by T cells: March 2004update,” Cancer Immunology, Immunotherapy, vol. 54, no. 3,pp. 187–207, 2005.


[9] A. E. Albers, R. L. Ferris, G. G. Kim, K. Chikamatsu, A. B.DeLeo, and T. L. Whiteside, “Immune responses to p53 inpatients with cancer:enrichment in tetramer+ p53 peptide-specific T cells and regulatory T cells at tumor sites,” CancerImmunology, Immunotherapy, vol. 54, no. 11, pp. 1072–1081,2005.

[10] A. E. Albers, C. Schaefer, C. Visus, W. Gooding, A. B. DeLeo,and T. L. Whiteside, “Spontaneous apoptosis of tumor-specific tetramer+ CD8+ T lymphocytes in the peripheralcirculation of patients with head and neck cancer,” Head andNeck, vol. 31, no. 6, pp. 773–781, 2009.

[11] B. Seliger, T. Cabrera, F. Garrido, and S. Ferrone, “HLA classI antigen abnormalities and immune escape by malignantcells,” Seminars in Cancer Biology, vol. 12, no. 1, pp. 3–13,2002.

[12] T. L. Whiteside, “Tumor-induced death of immune cells: itsmechanisms and consequences,” Seminars in Cancer Biology,vol. 12, no. 1, pp. 43–50, 2002.

[13] M. L. Gillison and K. V. Shah, “Human papillomavirus-associated head and neck squamous cell carcinoma: mount-ing evidence for an etiologic role for human papillomavirusin a subset of head and neck cancers,” Current Opinion inOncology, vol. 13, no. 3, pp. 183–188, 2001.

[14] M. L. Gillison, W. M. Koch, R. B. Capone et al., “Evidencefor a causal association between human papillomavirus anda subset of head and neck cancers,” Journal of the NationalCancer Institute, vol. 92, no. 9, pp. 709–720, 2000.

[15] E. M. Smith, K. F. Summersgill, J. Allen et al., “Humanpapillomavirus and risk of laryngeal cancer,” Annals ofOtology, Rhinology and Laryngology, vol. 109, no. 11, pp.1069–1076, 2000.

[16] C. G. L. Hobbs, J. A. C. Sterne, M. Bailey, R. S. Heyderman,M. A. Birchall, and S. J. Thomas, “Human papillomavirusand head and neck cancer: a systematic review and meta-analysis,” Clinical Otolaryngology, vol. 31, no. 4, pp. 259–266,2006.

[17] H. zur Hausen, “Papillomaviruses and cancer: from basicstudies to clinical application,” Nature Reviews Cancer, vol.2, no. 5, pp. 342–350, 2002.

[18] A. Albers, K. Abe, J. Hunt et al., “Antitumor activity of humanpapillomavirus type 16 E7-specific T cells against virallyinfected squamous cell carcinoma of the head and neck,”Cancer Research, vol. 65, no. 23, pp. 11146–11155, 2005.

[19] A. E. Albers, C. Visus, T. Tsukishiro et al., “Alterations in theT-cell receptor variable β gene—restricted profile of CD8+

T lymphocytes in the peripheral circulation of patients withsquamous cell carcinoma of the head and neck,” ClinicalCancer Research, vol. 12, no. 8, pp. 2394–2403, 2006.

[20] K. Chikamatsu, A. Albers, J. Stanson et al., “p53110-124-specific human CD4+ T-helper cells enhance in vitrogeneration and antitumor function of tumor-reactive CD8+

T cells,” Cancer Research, vol. 63, no. 13, pp. 3675–3681, 2003.[21] K. Chikamatsu, K. Nakano, W. J. Storkus et al., “Generation

of anti-p53 cytotoxic T lymphocytes from human peripheralblood using autologous dendritic cells,” Clinical CancerResearch, vol. 5, no. 6, pp. 1281–1288, 1999.

[22] T. K. Hoffmann, C. Arsov, K. Schirlau et al., “T cellsspecific for HPV16 E7 epitopes in patients with squamouscell carcinoma of the oropharynx,” International Journal ofCancer, vol. 118, no. 8, pp. 1984–1991, 2006.

[23] T. K. Hoffmann, A. D. Donnenberg, S. D. Finkelstein et al.,“Frequencies of tetramer+ T cells specific for the wild-typesequence p53264-272 peptide in the circulation of patientswith head and neck cancer,” Cancer Research, vol. 62, no. 12,pp. 3521–3529, 2002.

[24] T. K. Hoffmann, D. J. Loftus, K. Nakano et al., “The abilityof variant peptides to reverse the nonresponsiveness of Tlymphocytes to the wild-type sequence p53264-272 epitope,”Journal of Immunology, vol. 168, no. 3, pp. 1338–1347, 2002.

[25] A. E. Albers and A. M. Kaufmann, “Therapeutic humanpapillomavirus vaccination,” Public Health Genomics, vol. 12,no. 5-6, pp. 331–342, 2009.

[26] T. Soussi, “The humoral response to the tumor-suppressorgene-product p53 in human cancer: implications for diagno-sis and therapy,” Immunology Today, vol. 17, no. 8, pp. 354–356, 1996.

[27] T. K. Hoffmann, K. Nakano, E. M. Elder et al., “Generationof T cells specific for the wild-type sequence p53264-272peptide in cancer patients: implications for immunoselectionof epitope loss variants,” Journal of Immunology, vol. 165, no.10, pp. 5938–5944, 2000.

[28] C. Schaefer, G. G. Kim, A. Albers, K. Hoermann, E. N.Myers, and T. L. Whiteside, “Characteristics of CD4+CD25+

regulatory T cells in the peripheral circulation of patientswith head and neck cancer,” British Journal of Cancer, vol. 92,no. 5, pp. 913–920, 2005.

[29] J. Shimizu, S. Yamazaki, and S. Sakaguchi, “Induction oftumor immunity by removing CD4+CD25+ T cells: a com-mon basis between tumor immunity and autoimmunity,”Journal of Immunology, vol. 163, no. 10, pp. 5211–5218, 1999.

[30] T. L. Whiteside, “The role of death receptor ligands in shap-ing tumor microenvironment,” Immunological Investigations,vol. 36, no. 1, pp. 25–46, 2007.

[31] I. Kuss, A. D. Donnenberg, W. Gooding, and T. L. Whiteside,“Effector CD8+CD45RO−CD27− T cells have signallingdefects in patients with squamous cell carcinoma of the headand neck,” British Journal of Cancer, vol. 88, no. 2, pp. 223–230, 2003.

[32] R. E. M. Toes, F. Ossendorp, R. Offringa, and C. J. M. Melief,“CD4 T cells and their role in antitumor immune responses,”Journal of Experimental Medicine, vol. 189, no. 5, pp. 753–756, 1999.

[33] K. Hung, R. Hayashi, A. Lafond-Walker, C. Lowenstein, D.Pardoll, and H. Levitsky, “The central role of CD4+ T cellsin the antitumor immune response,” Journal of ExperimentalMedicine, vol. 188, no. 12, pp. 2357–2368, 1998.

[34] D. Ito, A. Albers, Y. X. Zhao et al., “The wild-type sequence(wt) p5325-35 peptide induces HLA-DR7 and HLA-DR11-restricted CD4+ Th cells capable of enhancing the ex vivoexpansion and function of anti-wt p53264-272 peptide CD8+

T cells,” Journal of Immunology, vol. 177, no. 10, pp. 6795–6803, 2006.


[35] J. N. Blattman, D. J. D. Sourdive, K. Murali-Krishna, R.Ahmed, and J. D. Altman, “Evolution of the T cell repertoireduring primary, memory, and recall responses to viralinfection,” Journal of Immunology, vol. 165, no. 11, pp. 6081–6090, 2000.

[36] M. Y. Lin, L. K. Selin, and R. M. Welsh, “Evolution of the CD8T-cell repertoire during infections,” Microbes and Infection,vol. 2, no. 9, pp. 1025–1039, 2000.

[37] J. W. J. van Heijst, C. Gerlach, E. Swart et al., “Recruitmentof antigen-specific CD8+ Tcells in response to infection ismarkedly efficient,” Science, vol. 325, no. 5945, pp. 1265–1269, 2009.

[38] T. K. Hoffmann, G. Dworacki, T. Tsukihiro et al., “Sponta-neous apoptosis of circulating T lymphocytes in patients withhead and neck cancer and its clinical importance,” ClinicalCancer Research, vol. 8, no. 8, pp. 2553–2562, 2002.

[39] I. Kuss, B. Hathaway, R. L. Ferris, W. Gooding, and T. L.Whiteside, “Imbalance in absolute counts of T lymphocytesubsets in patients with head and neck cancer and its relationto disease,” Advances in Oto-Rhino-Laryngology, vol. 62, pp.161–172, 2005.

[40] I. Kuss, B. Hathaway, R. L. Ferris, W. Gooding, and T.L. Whiteside, “Decreased absolute counts of T lymphocytesubsets and their relation to disease in squamous cellcarcinoma of the head and neck,” Clinical Cancer Research,vol. 10, no. 11, pp. 3755–3762, 2004.

[41] T. Tsukishiro, A. D. Donnenberg, and T. L. Whiteside, “Rapidturnover of the CD8+CD28− T-cell subset of effector cells inthe circulation of patients with head and neck cancer,” CancerImmunology, Immunotherapy, vol. 52, no. 10, pp. 599–607,2003.

[42] T. E. Reichert, L. Strauss, E. M. Wagner, W. Gooding, andT. L. Whiteside, “Signaling abnormalities, apoptosis, andreduced proliferation of circulating and tumor-infiltratinglymphocytes in patients with oral carcinoma,” ClinicalCancer Research, vol. 8, no. 10, pp. 3137–3145, 2002.

[43] T. E. Reichert, H. Rabinowich, J. T. Johnson, and T.L. Whiteside, “Mechanisms responsible for signaling andfunctional defects,” Journal of Immunotherapy, vol. 21, no. 4,pp. 295–306, 1998.

[44] L. Strauss, C. Bergmann, and T. L. Whiteside, “Humancirculating CD4+CD25highFoxp3+ regulatory T cells killautologous CD8+ but not CD4+ responder cells by Fas-mediated apoptosis,” Journal of Immunology, vol. 182, no. 3,pp. 1469–1480, 2009.

[45] T. L. Whiteside, “Signaling defects in T lymphocytes ofpatients with malignancy,” Cancer Immunology Immunother-apy, vol. 48, no. 7, pp. 346–352, 1999.

[46] T. E. Reichert, R. Day, E. M. Wagner, and T. L. Whiteside,“Absent or low expression of the ζ chain in T cells at thetumor site correlates with poor survival in patients with oralcarcinoma,” Cancer Research, vol. 58, no. 23, pp. 5344–5347,1998.

[47] K. Okada, K. Komuta, S. Hashimoto, S. Matsuzaki, T.Kanematsu, and T. Koji, “Frequency of apoptosis of tumor-infiltrating lymphocytes induced by Fas counterattack inhuman colorectal carcinoma and its correlation with prog-nosis,” Clinical Cancer Research, vol. 6, no. 9, pp. 3560–3564,2000.

[48] J. D. Fontenot, J. P. Rasmussen, M. A. Gavin, and A. Y.Rudensky, “A function for interleukin 2 in Foxp3-expressingregulatory T cells,” Nature Immunology, vol. 6, no. 11, pp.1142–1151, 2005.

[49] M. A. Gavin, J. P. Rasmussen, J. D. Fontenot et al., “Foxp3-dependent programme of regulatory T-cell differentiation,”Nature, vol. 445, no. 7129, pp. 771–775, 2007.

[50] K. Lahl, C. Loddenkemper, C. Drouin et al., “Selectivedepletion of Foxp3+ regulatory T cells induces a scurfy-likedisease,” Journal of Experimental Medicine, vol. 204, no. 1, pp.57–63, 2007.

[51] J. M. Kim, J. P. Rasmussen, and A. Y. Rudensky, “RegulatoryT cells prevent catastrophic autoimmunity throughout thelifespan of mice,” Nature Immunology, vol. 8, no. 2, pp. 191–197, 2007.

[52] C. L. Bennett, J. Christie, F. Ramsdell et al., “The immunedysregulation, polyendocrinopathy, enteropathy, X-linkedsyndrome (IPEX) is caused by mutations of FOXP3,” NatureGenetics, vol. 27, no. 1, pp. 20–21, 2001.

[53] V. L. Godfrey, J. E. Wilkinson, and L. B. Russell, “X-linkedlymphoreticular disease in the scurfy (sf) mutant mouse,”American Journal of Pathology, vol. 138, no. 6, pp. 1379–1387,1991.

[54] W. Chen, W. Jin, N. Hardegen et al., “Conversion of periph-eral CD4+CD25− naive T cells to CD4+CD25+ regulatoryT cells by TGF-β induction of transcription factor Foxp3,”Journal of Experimental Medicine, vol. 198, no. 12, pp. 1875–1886, 2003.

[55] T. S. Davidson, R. J. DiPaolo, J. Andersson, and E. M.Shevach, “Cutting edge: IL-2 is essential for TGF-β-mediatedinduction of Foxp3+ T regulatory cells,” Journal of Immunol-ogy, vol. 178, no. 7, pp. 4022–4026, 2007.

[56] S. G. Zheng, J. Wang, P. Wang, J. D. Gray, and D. A. Horwitz,“IL-2 is essential for TGF-β to convert naive CD4+CD25−

cells to CD25+Foxp3+ regulatory T cells and for expansion ofthese cells,” Journal of Immunology, vol. 178, no. 4, pp. 2018–2027, 2007.

[57] A. Marson, K. Kretschmer, G. M. Frampton et al., “Foxp3occupancy and regulation of key target genes during T-cellstimulation,” Nature, vol. 445, no. 7130, pp. 931–935, 2007.

[58] A. Mantovani, P. Allavena, A. Sica, and F. Balkwill, “Cancer-related inflammation,” Nature, vol. 454, no. 7203, pp. 436–444, 2008.

[59] G. P. Dunn, L. J. Old, and R. D. Schreiber, “The immuno-biology of cancer immunosurveillance and immunoediting,”Immunity, vol. 21, no. 2, pp. 137–148, 2004.

[60] E. M. Sotomayor, I. Borrello, and H. I. Levitsky, “Toleranceand cancer: a critical issue in tumor immunology,” CriticalReviews in Oncogenesis, vol. 7, no. 5-6, pp. 433–456, 1996.

[61] U. K. Liyanage, T. T. Moore, H.-G. Joo et al., “Prevalenceof regulatory T cells is increased in peripheral blood andtumor microenvironment of patients with pancreas or breastadenocarcinoma,” Journal of Immunology, vol. 169, no. 5, pp.2756–2761, 2002.

[62] L. Strauss, C. Bergmann, W. Gooding, J. T. Johnson, andT. L. Whiteside, “The frequency and suppressor function ofCD4+CD25highFoxp3+ T cells in the circulation of patientswith squamous cell carcinoma of the head and neck,” ClinicalCancer Research, vol. 13, no. 21, pp. 6301–6311, 2007.

[63] L. Strauss, C. Bergmann, M. Szczepanski, W. Gooding,J. T. Johnson, and T. L. Whiteside, “A unique subset ofCD4+CD25highFoxp3+ T cells secreting interleukin-10 andtransforming growth factor-β1 mediates suppression in thetumor microenvironment,” Clinical Cancer Research, vol. 13,no. 15, part 1, pp. 4345–4354, 2007.

[64] T. J. Curiel, G. Coukos, L. Zou et al., “Specific recruitmentof regulatory T cells in ovarian carcinoma fosters immune


privilege and predicts reduced survival,” Nature Medicine,vol. 10, no. 9, pp. 942–949, 2004.

[65] D. Golgher, E. Jones, F. Powrie, T. Elliott, and A. Gallimore,“Depletion of CD25+ regulatory cells uncovers immuneresponses to shared murine tumor rejection antigens,”European Journal of Immunology, vol. 32, no. 11, pp. 3267–3275, 2002.

[66] E. Jones, M. Dahm-Vicker, A. K. Simon et al., “Depletion ofCD25+ regulatory cells results in suppression of melanomagrowth and induction of autoreactivity in mice,” CancerImmun, vol. 2, p. 1, 2002.

[67] R. P. M. Sutmuller, L. M. Van Duivenvoorde, A. Van Elsas etal., “Synergism of cytotoxic T lymphocyte-associated antigen4 blockade and depletion of CD25+ regulatory T cells in anti-tumor therapy reveals alternative pathways for suppressionof autoreactive cytotoxic T lymphocyte responses,” Journal ofExperimental Medicine, vol. 194, no. 6, pp. 823–832, 2001.

[68] D. Ridgway, “The first 1000 dendritic cell vaccinees,” CancerInvestigation, vol. 21, no. 6, pp. 873–886, 2003.

[69] N. G. Chakraborty, S. Chattopadhyay, S. Mehrotra, A.Chhabra, and B. Mukherji, “Regulatory T-cell response andtumor vaccine-induced cytotoxic T lymphocytes in humanmelanoma,” Human Immunology, vol. 65, no. 8, pp. 794–802,2004.

[70] G. Zhou, C. G. Drake, and H. I. Levitsky, “Amplificationof tumor-specific regulatory T cells following therapeuticcancer vaccines,” Blood, vol. 107, no. 2, pp. 628–636, 2006.

[71] V. Bronte, “Th17 and cancer: friends or foes?” Blood, vol. 112,no. 2, p. 214, 2008.

[72] Q. Xu, G. Liu, X. Yuan et al., “Antigen-specific T-cell responsefrom dendritic cell vaccination using cancer stem-like cell-associated antigens,” Stem Cells, vol. 27, no. 8, pp. 1734–1740,2009.

[73] C. Visus, D. Ito, A. Amoscato et al., “Identification of humanaldehyde dehydrogenase 1 family member a1 as a novel CD8+

T-cell-defined tumor antigen in squamous cell carcinomaof the head and neck,” Cancer Research, vol. 67, no. 21, pp.10538–10545, 2007.

[74] M. D. L. L. Garcia-Hernandez, A. Gray, B. Hubby, O.J. Klinger, and W. M. Kast, “Prostate stem cell antigenvaccination induces a long-term protective immune responseagainst prostate cancer in the absence of autoimmunity,”Cancer Research, vol. 68, no. 3, pp. 861–869, 2008.

[75] S. Pellegatta, P. L. Poliani, D. Corno et al., “Neurospheresenriched in cancer stem-like cells are highly effective ineliciting a dendritic cell-mediated immune response againstmalignant gliomas,” Cancer Research, vol. 66, no. 21, pp.10247–10252, 2006.

[76] A. Kiessling, M. Schmitz, S. Stevanovic et al., “Prostatestem cell antigen: identification of immunogenic peptidesand assessment of reactive CD8+ T cells in prostate cancerpatients,” International Journal of Cancer, vol. 102, no. 4, pp.390–397, 2002.

[77] R. Ueda, K. Ohkusu-Tsukada, N. Fusaki et al., “Identificationof HLA-A2- and A24-restricted T-cell epitopes derived fromSOX6 expressed in glioma stem cells for immunotherapy,”International Journal of Cancer, vol. 126, no. 4, pp. 919–929,2010.

[78] M. Schmitz, A. Temme, V. Senner et al., “Identification ofSOX2 as a novel glioma-associated antigen and potentialtarget for T cell-based immunotherapy,” British Journal ofCancer, vol. 96, no. 8, pp. 1293–1301, 2007.

[79] R. K. Selvaraj and T. L. Geiger, “A kinetic and dynamicanalysis of Foxp3 induced in T cells by TGF-beta,” Journalof Immunology, vol. 179, no. 2, p. 11, 2007, following 1390.

[80] A. R. M. Almeida, N. Legrand, M. Papiernik, and A. A.Freitas, “Homeostasis of peripheral CD4+ T cells: IL-2Rαand IL-2 shape a population of regulatory cells that controlsCD4+ T cell numbers,” Journal of Immunology, vol. 169, no.9, pp. 4850–4860, 2002.

[81] T. R. Malek, A. Yu, V. Vincek, P. Scibelli, and L. Kong, “CD4regulatory T cells prevent lethal autoimmunity in IL-2Rβ-deficient mice: implications for the nonredundant functionof IL-2,” Immunity, vol. 17, no. 2, pp. 167–178, 2002.

[82] R. L. Ferris, T. L. Whiteside, and S. Ferrone, “Immune escapeassociated with functional defects in antigen-processingmachinery in head and neck cancer,” Clinical CancerResearch, vol. 12, no. 13, pp. 3890–3895, 2006.

[83] S. Wang and L. Chen, “Co-signaling molecules of theB7-CD28 family in positive and negative regulation of Tlymphocyte responses,” Microbes and Infection, vol. 6, no. 8,pp. 759–766, 2004.

[84] F. Esteban, A. Concha, C. Huelin et al., “Histocompatibilityantigens in primary and metastatic squamous cell carcinomaof the larynx,” International Journal of Cancer, vol. 43, no. 3,pp. 436–442, 1989.

[85] J. R. Houck, F. M. Sexton, and G. Zajdel, “HLA class I andclass II antigen expression on squamous cell carcinoma of thehead and neck,” Archives of Otolaryngology—Head and NeckSurgery, vol. 116, no. 10, pp. 1181–1185, 1990.

[86] J. R. Grandis, D. M. Falkner, M. F. Melhem, W. E. Gooding,S. D. Drenning, and P. A. Morel, “Human leukocyte antigenclass I allelic and haplotype loss in squamous cell carcinomaof the head and neck: clinical and immunogenetic conse-quences,” Clinical Cancer Research, vol. 6, no. 7, pp. 2794–2802, 2000.

[87] F. M. Marincola, E. M. Jaffee, D. J. Hickljn, and S. Fer-rone, “Escape of human solid tumors from t-cell recogni-tion: molecular mechanisms and functional significance,”Advances in Immunology, no. 74, pp. 181–273, 2000.

[88] M. Meissner, T. E. Reichert, M. Kunkel et al., “Defectsin the human leukocyte antigen class I antigen-processingmachinery in head and neck squamous cell carcinoma:association with clinical outcome,” Clinical Cancer Research,vol. 11, no. 7, pp. 2552–2560, 2005.

[89] A. Vambutas, J. DeVoti, W. Pinn, B. M. Steinberg, and V.R. Bonagura, “Interaction of human papillomavirus type11 E7 protein with TAP-1 results in the reduction of ATP-dependent peptide transport,” Clinical Immunology, vol. 101,no. 1, pp. 94–99, 2001.

[90] B. Seliger, M. J. Maeurer, and S. Ferrone, “Antigen-processingmachinery breakdown and tumor growth,” ImmunologyToday, vol. 21, no. 9, pp. 455–464, 2000.

[91] I. Cruz, C. J. L. M. Meijer, J. M. M. Walboomers, P. J. F.Snijders, and I. van der Waal, “Lack of MHC class I surfaceexpression on neoplastic cells and poor activation of thesecretory pathway of cytotoxic cells in oral squamous cellcarcinomas,” British Journal of Cancer, vol. 81, no. 5, pp. 881–889, 1999.

[92] T. Bauernhofer, I. Kuss, B. Henderson, A. S. Baum, and T.L. Whiteside, “Preferential apoptosis of CD56dim naturalkiller cell subset in patients with cancer,” European Journalof Immunology, vol. 33, no. 1, pp. 119–124, 2003.


[93] T. L. Whiteside, “Immune suppression in cancer: effects onimmune cells, mechanisms and future therapeutic interven-tion,” Seminars in Cancer Biology, vol. 16, no. 1, pp. 3–15,2006.

[94] B. K. Cho, V. P. Rao, Q. Ge, H. N. Eisen, and J. Chen,“Homeostasis-stimulated proliferation drives naive T cellsto differentiate directly into memory T cells,” Journal ofExperimental Medicine, vol. 192, no. 4, pp. 549–556, 2000.

[95] L. Gattinoni, S. E. Finkelstein, C. A. Klebanoff et al.,“Removal of homeostatic cytokine sinks by lymphodepletionenhances the efficacy of adoptively transferred tumor-specificCD8+ T cells,” Journal of Experimental Medicine, vol. 202, no.7, pp. 907–912, 2005.

[96] A. N. Theofilopoulos, W. Dummer, and D. H. Kono, “T cellhomeostasis and systemic autoimmunity,” Journal of ClinicalInvestigation, vol. 108, no. 3, pp. 335–340, 2001.

[97] R. Dummer, U. Dobbeling, R. Geertsen, J. Willers, G. Burg,and J. Pavlovic, “Interferon resistance of cutaneous T-celllymphoma-derived clonal T-helper 2 cells allows selectiveviral replication,” Blood, vol. 97, no. 2, pp. 523–527, 2001.

[98] A. B. DeLeo and T. L. Whiteside, “Development of multi-epitope vaccines targeting wild-type-sequence p53 peptides,”Expert Review of Vaccines, vol. 7, no. 7, pp. 1031–1040, 2008.

[99] K. Sakakura, K. Chikamatsu, N. Furuya, E. Appella, T. L.Whiteside, and A. B. DeLeo, “Toward the development ofmulti-epitope p53 cancer vaccines: an in vitro assessment ofCD8+ T cell responses to HLA class I-restricted wild-typesequence p53 peptides,” Clinical Immunology, vol. 125, no.1, pp. 43–51, 2007.

[100] V. Balz, K. Scheckenbach, K. Gotte, U. Bockmuhl, I. Petersen,and H. Bier, “Is the p53 inactivation frequency in squamouscell carcinomas of the head and neck underestimated?Analysis of p53 exons 2-11 and human papillomavirus 16/18E6 transcripts in 123 unselected tumor specimens,” CancerResearch, vol. 63, no. 6, pp. 1188–1191, 2003.

[101] E. Y. Nikitina, J. I. Clark, J. Van Beynen et al., “Dendriticcells transduced with full-length wild-type p53 generateantitumor cytotoxic T lymphocytes from peripheral blood ofcancer patients,” Clinical Cancer Research, vol. 7, no. 1, pp.127–135, 2001.

[102] K. Azuma, S. Shichijo, Y. Maeda et al., “Mutated p53 geneencodes a nonmutated epitope recognized by HLA-B∗4601-restricted and tumor cell-reactive CTLs at tumor site,” CancerResearch, vol. 63, no. 4, pp. 854–858, 2003.

[103] K. Devaraj, M. L. Gillison, and T.-C. Wu, “Developmentof HPV vaccines for HPV-associated head and neck squa-mous cell carcinoma,” Critical Reviews in Oral Biology andMedicine, vol. 14, no. 5, pp. 345–362, 2003.

[104] A. E. Albers and A. M. Kaufmann, “Therapeutic humanpapillomavirus vaccination,” Public Health Genomics, vol. 12,no. 5-6, pp. 331–342, 2009.

[105] J. Paavonen, P. Naud, J. Salmeron et al., “Efficacy ofhuman papillomavirus (HPV)-16/18 AS04-adjuvanted vac-cine against cervical infection and precancer caused byoncogenic HPV types (PATRICIA): final analysis of a double-blind, randomised study in young women,” The Lancet, vol.374, no. 9686, pp. 301–314, 2009.

[106] L. L. Villa, G. Perez, S. K. Kjaer et al., “Quadrivalent vaccineagainst human papillomavirus to prevent high-grade cervicallesions,” The New England Journal of Medicine, vol. 356, no.19, pp. 1915–1927, 2007.

[107] A. Hildesheim, R. Herrero, S. Wacholder et al., “Effect ofhuman papillomavirus 16/18 L1 viruslike particle vaccineamong young women with preexisting infection: a random-ized trial,” Journal of the American Medical Association, vol.298, no. 7, pp. 743–753, 2007.

[108] U. Ritz, F. Momburg, H. Pilch, C. Huber, M. J. Maeurer,and B. Seliger, “Deficient expression of components of theMHC class I antigen processing machinery in human cervicalcarcinoma,” International Journal of Oncology, vol. 19, no. 6,pp. 1211–1220, 2001.

[109] M. V. Dhodapkar, “Immunity to stemness genes in humancancer,” Current Opinion in Immunology, vol. 22, no. 2, pp.245–250, 2010.

[110] Y. Li, H. Zeng, R.-H. Xu, B. Liu, and Z. Li, “Vaccination withhuman pluripotent stem cells generates a broad spectrum ofimmunological and clinical responses against colon cancer,”Stem Cells, vol. 27, no. 12, pp. 3103–3111, 2009.

[111] S. Senju, S. Hirata, Y. Motomura et al., “Pluripotent stemcells as source of dendritic cells for immune therapy,”International Journal of Hematology, vol. 91, no. 3, pp. 392–400, 2010.

[112] E. Vlashi, K. Kim, C. Lagadec et al., “In vivo imaging,tracking, and targeting of cancer stem cells,” Journal of theNational Cancer Institute, vol. 101, no. 5, pp. 350–359, 2009.

[113] M. T. Michalek, E. P. Grant, C. Gramm, A. L. Goldberg, andK. L. Rock, “A role for the ubiquitin-dependent proteolyticpathway in MHC class I-restricted antigen presentation,”Nature, vol. 363, no. 6429, pp. 552–554, 1993.

[114] K. Mellodew, R. Suhr, D. A. Uwanogho et al., “Nestinexpression is lost in a neural stem cell line through amechanism involving the proteasome and Notch signalling,”Developmental Brain Research, vol. 151, no. 1-2, pp. 13–23,2004.

[115] R. Ben-Saadon, D. Zaaroor, T. Ziv, and A. Ciechanover,“The polycomb protein Ring1B generates self atypical mixedubiquitin chains required for its in vitro histone H2A ligaseactivity,” Molecular Cell, vol. 24, no. 5, pp. 701–711, 2006.


Review Article

Immunotherapy for Renal Cell Carcinoma

Momoe Itsumi and Katsunori Tatsugami

Department of Urology, Graduate School of Medical Sciences, Kyushu University, 3-1-1 Maidashi, Higashi-ku,Fukuoka 812-8582, Japan

Correspondence should be addressed to Katsunori Tatsugami, [email protected]

Received 2 July 2010; Accepted 29 November 2010

Academic Editor: Eiji Matsuura

Copyright © 2010 M. Itsumi and K. Tatsugami. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Immunotherapy plays a significant role in the management of renal cell carcinoma (RCC) patients with metastatic diseasebecause RCC is highly resistant to both chemotherapy and radiation therapy. Many reports illustrate various approaches to thetreatment of RCC, such as cytokine-, antigen- or dendritic cell- (DC-) based immunotherapy, and the safety and effectivenessof immunotherapy have been highlighted by multiple clinical trials. Although antitumor immune responses and clinicallysignificant outcomes have been achieved in these trials, the response rate is still low, and very few patients show long-term clinicalimprovement. Recently, the importance of immune regulation by antigen-presenting cells (APC) and regulatory T cells (Treg cells)has also been discussed. The authors outline the principles of cell-mediated tumor immunotherapy and discuss clinical trials ofimmunotherapy for RCC.

1. Introduction

Renal cell carcinoma (RCC), a glandular carcinoma, ac-counts for approximately 85%–95% of adult malignantkidney cancer cases [1]. Patients with advanced or metastaticdisease have a poor prognosis, with a 5-year survival rateof less than 15%. Surgical treatment is effective, even inpatients with advanced or metastatic RCC, because of itshigh resistance to chemotherapy and radiation therapy.Immunotherapy using interferon (IFN)-α and/or interleukin(IL)-2 has shown promising anti-tumor activity in RCC[2–4]. However, these cytokines have a positive effect inonly 10%–20% of cases [5]. Like melanoma, RCC is classedas an immunogenic tumor based on its response rate toimmunotherapy, the incidence of spontaneous regression,and the high level of tumor T cell infiltration. Despite itsimmunogenicity, only a few CD8+ cytotoxic-T-lymphocytes(CTLs), which can efficiently eliminate RCC cells, have beenisolated [6]. This is in line with the small number of RCC-associated antigens that have so far been identified, therebylimiting the trials of candidate vaccines in these patients[7, 8].

Recently, tumor immunotherapy using DC has beenshown to have therapeutic potential for malignant tu-

mors. Moreover, nonmyeloablative stem cell transplantation(NST), which was developed for the treatment of leukemia,is effective against RCC [9, 10] and other solid tumors [11].In this review, we discuss the current status of cell-mediatedtumor-specific and nonspecific immunotherapy for RCC.

2. Tumor-Specific and Non-SpecificImmunotherapy

In vivo studies show that cellular immunity mediated by Tcells, natural killer (NK) cells or NK T cells plays a centralrole in the eradication of tumors. Since 1980, many attemptshave been made to administer anti-tumor cells to cancerpatients. In the late 1980s, human tumor antigens wereidentified and tumor-specific cellular immunity mediated viathese tumor antigens received a lot of attention. Also, theadministration of cytokines that activate cellular anti-tumorresponses, including those mediated by T cells and NK cells,has been the subject of much research. It is thought thatIFN-α induces Th1 cytokine production, thus promotinganti-tumor activity by cells that elicit cytotoxicity by actingdirectly on the tumor [12].


Table 1: Immunotherapy using inactivated tumor cells and a gene modified tumor vaccine (GMTV).

Authors Vaccine Adjuvant Patients Duration of PFS/RFS Results

Galligioni Auto irrad tumor BCG 120 13 mo5-year DFS 63%

(control 72%) P = .21

Schwaab Auto irrad tumor BCG, IFN-α, IFN-γ 14 — 3 MR, 5 SD, l PD

Dillman Auto irrad tumorBCG, IFN-α, IFN-β

25 2.4 momedian survival 33.4 mo,

GM-CSF, Cy 5-year survival 43%

Jocham Auto lysate None 379 47.8 mo5-year PFS 77.4%

(control 67.8%) P = .02

Dudek Auto LMI None, Cy, Cy+IL-2 31 2.8 moNone: 5 SD, Cy: 4 SD,

Cy+ IL-2: 1PR 3 SD

May Auto lysate None 495 —5 year,10 year OS: 80.6, 68.9%

(control 79.2, 62.1%) P = .066

SimonsAuto irrad tumor

None 16 — 1 PR+ GM-CSF

WittigAuto irrad tumor

Oligonucleotides 10 — 1 CR, 1 PR, 1 MR, 2 SD, 5 PD+ GM-CSF, IL-7

AntoniaAuto irrad tumor

IL-2 15 — 2 PR, 2 SD+ B7.1 gene

TaniAuto irrad tumor

None 6 — 1 SD, l MR+ GM-CSF

PizzaAuto irrad tumor

None 30 170.5 dy 1 CR, 4 PR, 9 SD+ IL-2

MoiseyenkoAuto irrad tumor

None 3 mo 1 SD, l MR+ tag7/PGPR-S gene 4

FishmanAuto irrad tumor

IL-2 39 — 1 CR, 2 PR, 24 SD+ B7.1 gene

BuchnerAuto irrad tumor

None 12 5.3 mo PFS 5.3 mo, OS 15.6 mo+ B7.1, IL-2 gene

LMI: large multivalent immunogen, Cy: cyclophosphamide, DFS: disease-free survival, Os: overall survival, PR: partial response, MR: mixed response, SD:stable disease, PD: progressive disease, PFS: progression-free survival, RFS: recurrence-free survival.

IL-2 is a growth/differentiation factor for NK cells andT cells, which induces and maintains the cytotoxicity, boththese cell types [13]. Because cytokine treatment inducesnonspecific anti-tumor activity, it is known as nonspecificimmunotherapy.

In 1984, Mule et al. reported lymphokine-activated killer(LAK) cell treatment of tumors using inducible cultured cells[14]. Culturing immune cells isolated from a cancer patient’speripheral blood, or excised tumor tissue, with IL-2 causesthem to differentiate into LAK cells. Since the second halfof the 1980s, treatment using LAK cells has been attemptedin several facilities [15, 16]. However, because the treatmentmethod causes severe side effects, it was never establishedas an effective treatment method. LAK cells have no tumorspecificity because they are induced in culture in response toIL-2 alone and not by tumor antigens. Thus, it was thoughtthat the adoptive transfer of LAK cells might result in damageto normal host cells in vivo.

Since Van Der Bruggen et al. identified tumor antigensthat were specifically recognized by T cells in a melanoma-bearing patient [17], research became more focused ontumor-specific immunotherapy. Though LAK cells, CTLs,macrophages, NK cells and NKT cells are all involved in host

immune response against tumors, CTLs are now thought tobe one of the most important factors responsible for anti-tumor immunity.

3. Immunotherapy Using InactivatedTumor Cells and Gene Modified TumorVaccines (GMTV)

Immunotherapy using inactivated tumor cells or tumorlysates is based on the idea that tumor cells express anti-gens that induce anti-tumor immune responses [18–22](Table 1). Because immunotherapy using tumor cells isrelatively straightforward, Jocham et al. undertook a large-scale randomized controlled trial and reported that the “non-replaced phase” after surgery for kidney cancer was extendedby an autologous tumor vaccine [20]. The percentage ofvaccinated patients showing no disease progression 5 yearsafter treatment was 77.4% compared with 67.8% of thecontrols.

Both cytokines and antigen-presenting cells are impor-tant for the induction of effective immune responses[23]. Thus, GMTV was used to introduce virus-expressing


CTLCD8T cell

CD4T cell

MHC class ICostimulatory

interaction

Costimulatorymolecule

MHC class II

DC

Phagocytosis

Processing

MHC class Ipeptide

Tumorlysate

Virusvector

Tumor antigen

MHC class II peptideKLH

Helper protein

Tumor

Figure 1: CTL induction by Apcs. Antigens are taken up and degraded into peptide fragments by antigen presenting cells (APC), such asimmature DC. At some point on their path to the cell surface, newly synthesized MHC class II or I molecules bind the peptide antigenfragments and transport the peptides to the cell surface. CD8+ T cells recognizing the antigen expressed by weakly costimulatory cells becomeactivated only in the presence of CD4+ T cells bound to the same APC. This happens via CD4+ T cells recognizing antigens presented byAPCs and being triggered to induce increased levels of costimulatory activity by the antigen-presenting cell. The CD4+ T cells also produceincreased amounts of IL-2, which drives CD8+ T cell proliferation. CD8+ T cells then become cytotoxic T lymphocytes (CTL).

cytokines, or costimulatory molecules, into tumor cells(Table 1) [18–22, 24–33]. GMTV-immunotherapy introduc-ing cytokine transgene, such as GM-CSF or IL-2, or costim-ulatory molecule transgene such as B7-1 into autologous ir-radiated tumors, has been carried out. However, these studieswere disappointing in terms of a significant clinical response,such as tumor regression. Though the use of multiple tumorantigens should induce a greater immune response, onecannot rule out the possibility of unintentionally inhibitinganti-tumor immunity or of eliciting non-specific immuneresponses.

4. Peptide-Based Immunotherapy

Since the development of the SEREX method, which enablesthe identification of tumor antigens from cDNA libraries,many peptide-based vaccination studies have been under-taken. Because the effective induction of anti-tumor immu-nity using single peptides is difficult, MHC class II peptideshave been used along with adjuvants (Table 2) [34–40].HSPPC-96 (vitespen) is a heat shock protein. It is a peptidecomplex, in which the heat shock protein plays the role ofan adjuvant. However, a recent randomized phase III studysuggested that this complex did not improve recurrence-free survival rates [41]. Further studies are required to seewhether antigen-specific T cells homogeneously induced bya single tumor antigen can be effective against a diversepopulation of tumor cells.

5. DC-Based Immunotherapy

Antigens processed within the proteasome of tumor cells arepresented on major histocompatibility antigen (MHC) classI molecules of tumor cell as tumor antigen peptides thatCTLs recognize, thus triggering CTL-mediated cytotoxicity.However, CTLs are not activated by direct recognitionof the antigens expressed by tumor cells; they need helpfrom dendritic cells (DCs) and CD4+ helper T cells. Toactivate a CD8+ T cell to become a CTL, engagement ofthe T cell receptor with a peptide antigen presented by anMHC class I molecule is not enough. The T cell must alsorecognize a costimulatory molecule (e.g. CD80 or CD86)(Figure 1). Moreover, antigen presenting cells (APCs) areactivated through their interaction with CD4+ T cells, andthen they express various costimulatory molecules. DCs arethe most well-known and efficient APCs and are presentin various tissues, including lymphoid and nonlymphoidorgans and the blood, where they take up both particulateand soluble antigens before migrating to the lymph nodesto induce immune responses. Subsequently, DCs presentantigen to T cells in the lymph nodes and induce antigen-specific immune responses, including the induction of CTLs.DCs also present antigen to other cells, including NK cells.

Clinical trials of DC therapy are listed in Table 3 [23,35, 36, 42–59]. Although immunotherapy using DCs andnonautologous tumor cells seems to induce host immunecells to recognize tumor cells, there is still the possibility ofalloreactive immune responses induced by nonself-antigens.


Table 2: Peptide-based immunotherapy.

Authors Stage Vaccine Adjuvant PatientsDuration of

PFS/RFSResults

Uemura mRCC CA9-derived peptideIncomplete Freund’s

adjuvant23 12.2 mo 3 PR, 6SD

Iiyama mRCC WT 1-peptideIncomplete Freund’s

adjuvant3 — 2 SD

Suekane mRCC 4 different peptides None, IFN-α, IL-2 10 23 wk 6 SD

Wood cT1b-T4N0M0 or Tant N1-2 M0

HSPPC-96 (vitespen) None 728 1.9 yrNo difference in

recurrence-free survival

Jonasch mRCC HSPPC-96 (vitespen) None 60 65 dy 2 CR, 2 PR, 7 SD

mRCC: metastatic RCC, PADRE: pan-MHC class II binding peptide, Auto mDC: autologous mature DC, CR: complete response, PR: partial response, SD:stable disease, PFS: progression-free survival, RFS: recurrence-free survival.

Because nonautologous DCs (allo-DCs) may be attacked bythe host immune system, immunotherapy using autologous-DCs (auto DCs) might be more effective in vivo. To date, allreports regarding DC treatment are of phase I/II trials incor-porating different methodologies. Although delayed-typehypersensitivity reactions in response to tumor cell lysatesor keyhole limpet hemocyanin (KLH) and the productionof IFN-γ by antigen-specific lymphocytes were observed, thenumber of patients showing a positive clinical response wasstill low.

We also used IFN-α as an adjunctive agent for DCtherapy. As previously noted, IFN-α induced an environmentconducive to DC activation and enhanced migratory com-petence [60, 61]. We evaluated the efficacy of DC-therapyin combination with IFN-α in patients with advanced RCC.After 4 months of vaccinations, five patients had stabledisease and two had progressive disease. In six patients,the time-to-progression was prolonged compared with thatseen after previous cytokine treatment. Because cytokinecombination therapy induces the proliferation and mainte-nance of DC-activated T cells, combination therapy usingIL-2 is reasonable. However, Oosterwijk et al. reported thatcombination therapy with IL-2 plus DCs was no moreeffective than DCs alone [44]. Recently, it was reported thatIL-2 participates in the maintenance of regulatory T cells(Tregs), which suppress immune responses [62]. Furtherstudy of the role of IL-2 in immunotherapy is required.

6. Nonmyeloablative Stem CellTransplantation (NST)

Though NST was developed for the treatment of leukemia,it began to gain attention as a treatment for solid tumors.In 2000, Childs et al. performed NST on 19 renal carcinomapatients and reported a success rate 53%; three patients werein complete remission and seven patents were in partialremission. Previous reports have highlighted the importantrole played by cellular anti-tumor immunity, including thatmediated by donor T cells in graft versus host disease(GVHD) and the graft versus tumor effect (GVT); the ap-pearance of GVHD induced by transplantation of donor Tcells is inversely correlated with the rate of tumor recurrence.Recurrence is especially high in T cell-depleted stem cell

transplants, and the administration of donor lymphocyteseffectively reduces the incidence of recurrence [63, 64].Donor T cells induce GVHD/GVT against recipient anti-gens, including MHC molecules, minor histocompatibilityantigens and tumor cell-specific antigens. An effective GVTresponse can be induced if the antigen distribution betweennormal cells and tumor cells can be identified, and if donorT cell responses against normal cells can be controlled. Thus,in NST, the mechanism by which tumor specific immunity isinduced is very important, and a recent study attempted toaddress the question of how this response was activated [65].

When the patient receives immunosuppressive treatmentfor GVHD, it might also cause suppression of the associatedanti-tumor effects. In these patients, the differentiation ofmononuclear cells into DCs is inhibited in vitro [66]. There-fore, when treating a patient with NST, one should bear inmind possible aggravation of the neoplasm by immunosup-pressive therapy directed against GVHD.

7. Regulatory CD4+ T Cells and the Tumor

Recent research shows that CD4+ T cells constitutively ex-pressing the IL-2 receptor α-chain (CD25) act in a regulatorycapacity by suppressing the activation and function of otherT cells [67]. Their physiological role is to protect the hostagainst the development of autoimmunity by regulatingimmune responses against antigens expressed by normaltissues [68, 69]. Since tumor antigens are largely self-antigens, these so-called Treg cells may also prevent thetumor-bearing host from mounting an effective antitumorimmune response. Previous studies have shown that elevatednumbers of CD4+CD25+ Tregs can be found in patients withadvanced cancer [70] and that high Treg frequencies areassociated with reduced survival [71]. In our experimentsinto cytokine therapy for RCC patients, the number ofCD4+ and FoxP3+Treg cells was significantly decreasedafter IFN-α treatment, and Treg cell levels before treatmentcorrelated with the clinical response [72]. The importantrole of CD4+CD25+ Tregs in controlling tumor growth wasfurther highlighted by the demonstration that depletion ofTregs using anti-CD25 antibodies evokes effective antitumorimmunity in mice [73, 74]. Dannull et al. used a recom-binant IL-2:diphtheria toxin conjugate (DAB389IL-2; also


Table 3: DC-based immunotherapy.

Authors Antigen DC Adjuvant PatientsDuration of

PFS/RFSResults

Oosterwijk-Wakka Auto lysate Auto imDC KLH/IL-2 12 — 8 SD, 4 PD

Marten Auto lysate Auto mDC KLH 15 — 1 PR, 7 SD, 7 PD

Holtl Auto & Allo lysate Auto mDC KLH 27 20.4 mo2 CR, 1 PR, 7

SD, 17 PD

Azuma Auto lysate Auto imDC KLH 3 — 1 NC, 2 PD

Marten DC/auto tumorfusion

Allo mDC — 12 — 4 SD, 8 PD

Su tumor RNA Auto imDC — 10 — not evaluated

Gitliz Auto lysate Auto imDC — 12 — 1 PR, 3 SD, 8 PD

Barbuto DC/auto tumorfusion

Allo mDC — 19 5.7 mo30 R, 14 SD, 2

PD

Avigan DC/auto tumorfusion

Auto imDC KLH 13 4.2 mo 5 SD, 8 PD

Pandha Allo lysate Auto imDC KLH 5 — 2 SD

Arroyo Auto lysate Auto mDC KLH 5 9.6 mo (5–16) 3 SD

Holtl Auto & Allo TuLy Allo mDC KLH/Cy 20 22.3 mo2 MR, 3 SD, 15

PD

Wierecky MUC-1 peptide Auto mDC PADRE 20 10.8 mo (4–24)1 CR, 2 MR, 2

PR, 5 SD, 10 PD

Bleumer CA9 peptide Auto mDCKLH CA9 class

II peptide6 — 6 PD

Wei DC/auto tumorfusion

Auto mDC IL-2 10 7 mo (5–12) 1 PR, 3 SD, 6 PD

Matsumoto Auto lysate Auto mDC KLH 3 — 1 SD, 2 PD

Kim Auto lysate Auto mDC KLH 9 5.2 mo 1 PR, 5 SD, 3 PD

BerntsenLysate or surviving

and telomerasepeptides

Auto mDC IL-2 27 2.7 mo 13 SD, 14 PD

Tatsugami Auto TuLy Auto mDC IFN-α 7 7.8 mo 5 SD, 2 PD

Zbou DC/auto tumorfusion

Allo mDC — 10 — 1 PR, 6 SD, 3 PD

Cy: cyclophosphamide, PADRE: pan-MHC class II binding peptide, Auto mDC: autologous mature DC, Allo imDC: allogeneic immature DC CR: completeresponse, PR: partial response, MR: mixed response, SD: stable disease, OR: objective response, PD: progressive disease, PFS: progression-free survival, RFS:recurrence-free survival.

known as denileukin diftitox and ONTAK) to eliminateCD25-expressing Tregs in metastatic RCC patients, and re-ported that depletion of Tregs in RCC patients followed byvaccination with tumor RNA-transfected DCs led to im-proved stimulation of tumor-specific T cells compared withvaccination alone [75]. It will be critical to collect accurateinformation regarding Tregs to address the clinical efficacyof such strategies in cancer patients.

8. Conclusions

The use of immunotherapy using cultured cells, such asDCs, to treat large numbers of patients, and the conductionof large-scale studies are difficult because of the problemsassociated with the need for adequate culture facilities andappropriate culture techniques. Because of the complexity ofthe immune responses involved, it is difficult to evaluate the

efficacy of immunotherapy compared with other treatments.However, as it is clear that the immune system plays asignificant role in the control of tumors, continued analysisof the mechanisms involved in tumor immunity and thedevelopment of new immunotherapies are vital.

References

[1] P. F. A. Mulders, A. H. Brouwers, C. A. Hulsbergen-van DerKaa, E. N. J. TH. Van Lin, S. Osanto, and P. H. M. De Mulder,“Guideline ’Renal cell carcinoma,” Nederlands Tijdschrift voorGeneeskunde, vol. 152, no. 7, pp. 376–380, 2008.

[2] D. T. Alexandrescu and C. A. Dasanu, “Kidney cancertherapy: new perspectives and avenues,” Expert Opinion onPharmacotherapy, vol. 7, no. 18, pp. 2481–2493, 2006.

[3] J. C. Yang and R. Childs, “Immunotherapy for renal cellcancer,” Journal of Clinical Oncology, vol. 24, no. 35, pp. 5576–5583, 2006.


[4] R. J. Motzer and R. M. Bukowski, “Targeted therapy formetastatic renal cell carcinoma,” Journal of Clinical Oncology,vol. 24, no. 35, pp. 5601–5608, 2006.

[5] B. C. Leibovich, K. R. Han, M. H. T. Bui et al., “Scor-ing algorithm to predict survival after nephrectomy andimmunotherapy in patients with metastatic renal cell car-cinoma: a stratification tool for prospective clinical trials,”Cancer, vol. 98, no. 12, pp. 2566–2575, 2003.

[6] J. H. Finke, P. Rayman, M. Edinger et al., “Characterizationof a human renal cell carcinoma specific cytotoxic CD8 T cellline,” Journal of Immunotherapy, vol. 11, no. 1, pp. 1–11, 1992.

[7] N. Brouwenstijn, B. Gaugler, K. M. Kruse et al., “Renal-cell carcinoma-specific lysis by cytotoxic T-lymphocyte clonesisolated from peripheral blood lymphocytes and tumor-infiltrating lymphocytes,” International Journal of Cancer, vol.68, no. 2, pp. 177–182, 1996.

[8] C. Gaudin, F. Kremer, E. Angevin, V. Scott, and F. Triebel, “Ahsp70-2 mutation recognized by CTL on a human renal cellcarcinoma,” Journal of Immunology, vol. 162, no. 3, pp. 1730–1738, 1999.

[9] R. Childs, A. Chernoff, N. Contentin et al., “Regressionof metastatic renal-cell carcinoma after nonmyeloablativeallogeneic peripheral-blood stem-cell transplantation,” NewEngland Journal of Medicine, vol. 343, no. 11, pp. 750–758,2000.

[10] B. I. Rini, T. Zimmerman, W. M. Stadler, T. F. Gajewski, andN. J. Vogelzang, “Allogeneic stem-cell transplantation of renalcell cancer after nonmyeloablative chemotherapy: feasibility,engraftment, and clinical results,” Journal of Clinical Oncology,vol. 20, no. 8, pp. 2017–2024, 2002.

[11] S. Slavin, A. Nagler, E. Naparstek et al., “Nonmyeloablativestem cell transplantation and cell therapy as an alternative toconventional bone marrow transplantation with lethal cytore-duction for the treatment of malignant and nonmalignanthematologic diseases,” Blood, vol. 91, no. 3, pp. 756–763, 1998.

[12] F. Belardelli, M. Ferrantini, E. Proietti, and J. M. Kirkwood,“Interferon-alpha in tumor immunity and immunotherapy,”Cytokine and Growth Factor Reviews, vol. 13, no. 2, pp. 119–134, 2002.

[13] K. A. Margolin, “Interleukin-2 in the treatment of renalcancer,” Seminars in Oncology, vol. 27, no. 2, pp. 194–203,2000.

[14] J. J. Mule, S. Shu, S. L. Schwarz, and S. A. Rosenberg, “Adoptiveimmunotherapy of established pulmonary metastases withLAK cells and recombinant interleukin-2,” Science, vol. 225,no. 4669, pp. 1487–1489, 1984.

[15] S. Negrier, T. Philip, G. Stoter et al., “Interleukin-2 with orwithout LAK cells in metastatic renal cell carcinoma: a reportof a European multicentre study,” European Journal of Cancerand Clinical Oncology, vol. 25, supplement 3, pp. S21–S28,1989.

[16] J. P. Kuebler, R. P. Whitehead, D. L. Ward, G. P. Hemstreet,E. C. Bradley, and J. B. DeKernion, “Treatment of metastaticrenal cell carcinoma with recombinant interleukin- 2 incombination with vinblastine or lymphokine-activated killercells,” Journal of Urology, vol. 150, no. 3, pp. 814–820, 1993.

[17] P. Van Der Bruggen, C. Traversari, P. Chomez et al., “A geneencoding an antigen recognized by cytolytic T lymphocytes ona human melanoma,” Science, vol. 254, no. 5038, pp. 1643–1647, 1991.

[18] E. Galligioni, M. Quaia, A. Merlo et al., “Adjuvant im-munotherapy treatment of renal carcinoma patients with

autologous tumor cells and bacillus Calmette-Guerin: five-year results of a prospective randomized study,” Cancer, vol.77, no. 12, pp. 2560–2566, 1996.

[19] T. Schwaab, J. A. Heaney, A. R. Schned et al., “A randomizedphase II trial comparing two different sequence combinationsof autologous vaccine and human recombinant interferon γand human recombinant interferon α2B therapy in patientswith metastatic renal cell carcinoma: clinical outcome andanalysis of immunological parameters,” Journal of Urology, vol.163, no. 4, pp. 1322–1327, 2000.

[20] D. Jocham, A. Richter, L. Hoffmann et al., “Adjuvantautologous renal tumour cell vaccine and risk of tumourprogression in patients with renal-cell carcinoma after radicalnephrectomy: phase III, randomised controlled trial,” Lancet,vol. 363, no. 9409, pp. 594–599, 2004.

[21] A. Z. Dudek, M. F. Mescher, I. Okazaki et al., “Autologous largemultivalent immunogen vaccine in patients with metastaticmelanoma and renal cell carcinoma,” American Journal ofClinical Oncology: Cancer Clinical Trials, vol. 31, no. 2, pp.173–181, 2008.

[22] R. Dillman, N. Barth, L. VanderMolen et al., “Autologoustumor cell line-derived vaccine for patient-specific treatmentof advanced renal cell carcinoma,” Cancer Biotherapy andRadiopharmaceuticals, vol. 19, no. 5, pp. 570–580, 2004.

[23] K. Tatsugami, M. Eto, M. Harano et al., “Dendritic celltherapy in combination with interferon-α for the treatmentof metastatic renal cell carcinoma,” International Journal ofUrology, vol. 15, no. 8, pp. 694–698, 2008.

[24] S. J. Antonia, J. Seigne, J. Diaz et al., “Phase I trial of aB7-1 (CD80) gene modified autologous tumor cell vaccinein combination with systemic interleukin-2 in patients withmetastatic renal cell carcinoma,” Journal of Urology, vol. 167,no. 5, pp. 1995–2000, 2002.

[25] M. May, F. Kendel, B. Hoschke et al., “Adjuvant autologoustumour cell vaccination in patients with renal cell carci-noma: overall survival analysis with a follow-up period inexcess of more than 10 yearsAdjuvante autologe Tumor-vakzine beim Nierenzellkarzinom: gesamtuberlebensanalysemit einem Nachbeobachtungszeitraum von mehr als 10Jahren,” Urologe A, vol. 48, no. 9, pp. 1075–1083, 2009.

[26] A. Buchner, H. Pohla, G. Willimsky et al., “Phase 1 trial of allo-geneic gene-modified tumor cell vaccine RCC-26/CD80/IL-2in patients with metastatic renal cell carcinoma,” Human GeneTherapy, vol. 21, no. 3, pp. 285–297, 2010.

[27] R. O. Dillman, N. M. Barth, L. A. VanderMolen et al.,“Treatment of kidney cancer with autologous tumor cellvaccines of short-term cell lines derived from renal cellcarcinoma,” Cancer Biotherapy and Radiopharmaceuticals, vol.16, no. 1, pp. 47–54, 2001.

[28] J. W. Simons and B. Mikhak, “Ex vivo gene therapy usingcytokine-transduced tumor vaccines: molecular and clinicalpharmacology,” Seminars in Oncology, vol. 25, no. 6, pp. 661–676, 1998.

[29] B. Wittig, A. Marten, T. Dorbic et al., “Therapeutic vaccinationagainst metastatic carcinoma by expression-modulated andimmunomodified autologous tumor cells: a first clinical phaseI/II trial,” Human Gene Therapy, vol. 12, no. 3, pp. 267–278,2001.

[30] M. Fishman, T. B. Hunter, H. Soliman et al., “Phase II trialof B7-1 (CD-86) transduced, cultured autologous tumor cellvaccine plus subcutaneous interleukin-2 for treatment of stageIV renal cell carcinoma,” Journal of Immunotherapy, vol. 31,no. 1, pp. 72–80, 2008.


[31] K. Tani, M. Azuma, Y. Nakazaki et al., “Phase I study ofautologous tumor vaccines transduced with the GM-CSF genein four patients with stage IV renal cell cancer in Japan: clinicaland immunological findings,” Molecular Therapy, vol. 10, no.4, pp. 799–816, 2004.

[32] G. Pizza, C. De Vinci, G. Lo Conte et al., “Allogeneic gene-modified tumour cells in metastatic kidney cancer. Report II,”Folia Biologica, vol. 50, no. 6, pp. 175–183, 2004.

[33] V. M. Moiseyenko, A. O. Danilov, I. A. Baldueva et al., “PhaseI/II trial of gene therapy with autologous tumor cells modifiedwith tag7/PGRP-S gene in patients with disseminated solidtumors,” Annals of Oncology, vol. 16, no. 1, pp. 162–168, 2005.

[34] H. Uemura, K. Fujimoto, M. Tanaka et al., “A phase Itrial of vaccination of CA9-derived peptides for HLA-A24-positive patients with cytokine-refractory metastatic renal cellcarcinoma,” Clinical Cancer Research, vol. 12, no. 6, pp. 1768–1775, 2006.

[35] J. Wierecky, M. Mueller, and P. Brossart, “Dendritic cell-based cancer immunotherapy targeting MUC-1,” CancerImmunology, Immunotherapy, vol. 55, no. 1, pp. 63–67, 2006.

[36] I. Bleumer, D. M. Tiemessen, J. C. Oosterwijk-Wakka etal., “Preliminary analysis of patients with progressive renalcell carcinoma vaccinated with CA9-peptide-pulsed maturedendritic cells,” Journal of Immunotherapy, vol. 30, no. 1, pp.116–122, 2007.

[37] T. Iiyama, K. Udaka, S. Takeda et al., “WT1 (Wilms’ tumor 1)peptide immunotherapy for renal cell carcinoma,” Microbiol-ogy and Immunology, vol. 51, no. 5, pp. 519–530, 2007.

[38] S. Suekane, M. Nishitani, M. Noguchi et al., “Phase I trialof personalized peptide vaccination for cytokine-refractorymetastatic renal cell carcinoma patients,” Cancer Science, vol.98, no. 12, pp. 1965–1968, 2007.

[39] C. Wood, P. Srivastava, R. Bukowski et al., “An adjuvantautologous therapeutic vaccine (HSPPC-96; vitespen) versusobservation alone for patients at high risk of recurrence afternephrectomy for renal cell carcinoma: a multicentre, open-label, randomised phase III trial,” The Lancet, vol. 372, no.9633, pp. 145–154, 2008.

[40] E. Jonasch, C. Wood, P. Tamboli et al., “Vaccination ofmetastatic renal cell carcinoma patients with autologoustumour-derived vitespen vaccine: clinical findings,” BritishJournal of Cancer, vol. 98, no. 8, pp. 1336–1341, 2008.

[41] H. Van Poppel, S. Joniau, and S. W. Van Gool, “Vaccinetherapy in patients with renal cell carcinoma,” EuropeanUrology, vol. 55, no. 6, pp. 1333–1344, 2009.

[42] Y. C. Wei, R. P. Sticca, J. Li et al., “Combined treatment ofdendritoma vaccine and low-dose interleukin-2 in stage IVrenal cell carcinoma patients induced clinical response: a pilotstudy,” Oncology Reports, vol. 18, no. 3, pp. 665–671, 2007.

[43] J. H. Kim, Y. Lee, Y. S. Bae et al., “Phase I/II study ofimmunotherapy using autologous tumor lysate-pulsed den-dritic cells in patients with metastatic renal cell carcinoma,”Clinical Immunology, vol. 125, no. 3, pp. 257–267, 2007.

[44] J. C. Oosterwijk-Wakka, D. M. Tiemessen, I. Bleumer et al.,“Vaccination of patients with metastatic renal cell carcinomawith autologous dendritic cells pulsed with autologous tumorantigens in combination with interleukin-2: a phase 1 study,”Journal of Immunotherapy, vol. 25, no. 6, pp. 500–508, 2002.

[45] A. Marten, D. Flieger, S. Renoth et al., “Therapeutic vacci-nation against metastatic renal cell carcinoma by autologousdendritic cells: preclinical results and outcome of a first clinicalphase I/II trial,” Cancer Immunology, Immunotherapy, vol. 51,no. 11-12, pp. 637–644, 2002.

[46] L. Holtl, C. Zelle-Rieser, H. Gander et al., “Immunotherapyof metastatic renal cell carcinoma with tumor lysate-pulsedautologous dendritic cells,” Clinical Cancer Research, vol. 8, no.11, pp. 3369–3376, 2002.

[47] T. Azuma, S. Horie, K. Tomita et al., “Dendritic cell im-munotherapy for patients with metastatic renal cell carci-noma: university of Tokyo experience,” International Journalof Urology, vol. 9, no. 6, pp. 340–346, 2002.

[48] A. Marten, S. Renoth, T. Heinicke et al., “Allogeneic dendriticcells fused with tumor cells: preclinical results and outcome ofa clinical phase I/II trial in patients with metastatic renal cellcarcinoma,” Human Gene Therapy, vol. 14, no. 5, pp. 483–494,2003.

[49] Z. Su, J. Dannull, A. Heiser et al., “Immunological and clinicalresponses in metastatic renal cancer patients vaccinated withtumor RNA-transfected dendritic cells,” Cancer Research, vol.63, no. 9, pp. 2127–2133, 2003.

[50] B. J. Gitlitz, A. S. Belldegrun, A. Zisman et al., “A pilot trialof tumor lysate-loaded dendritic cells for the treatment ofmetastatic renal cell carcinoma,” Journal of Immunotherapy,vol. 26, no. 5, pp. 412–419, 2003.

[51] J. A. M. Barbuto, L. F. C. Ensina, A. R. Neves et al., “Dendriticcell-tumor cell hybrid vaccination for metastatic cancer,”Cancer Immunology, Immunotherapy, vol. 53, no. 12, pp. 1111–1118, 2004.

[52] H. S. Pandha, R. J. John, J. Hutchinson et al., “Dendritic cellimmunotherapy for urological cancers using cryopreservedallogeneic tumour lysate-pulsed cells: a phase I/II study,” BJUInternational, vol. 94, no. 3, pp. 412–418, 2004.

[53] J. C. Arroyo, F. Gabilondo, L. Llorente, M. A. Meraz-Rıos, andC. Sanchez-Torres, “Immune response induced in Vitro byCD16 and CD16 monocyte-derived dendritic cells in patientswith metastatic renal cell carcinoma treated with dendritic cellvaccines,” Journal of Clinical Immunology, vol. 24, no. 1, pp.86–96, 2004.

[54] J. Wierecky, M. R. Muller, S. Wirths et al., “Immunologicand clinical responses after vaccinations with peptide-pulseddendritic cells in metastatic renal cancer patients,” CancerResearch, vol. 66, no. 11, pp. 5910–5918, 2006.

[55] A. Matsumoto, K. Haraguchi, T. Takahashi et al., “Im-munotherapy against metastatic renal cell carcinoma withmature dendritic cells,” International Journal of Urology, vol.14, no. 4, pp. 277–283, 2007.

[56] A. Berntsen, R. Trepiakas, L. Wenandy et al., “Therapeuticdendritic cell vaccination of patients with metastatic renal cellcarcinoma: a clinical phase 1/2 trial,” Journal of Immunother-apy, vol. 31, no. 8, pp. 771–780, 2008.

[57] J. Zhou, D. Weng, F. Zhou et al., “Patient-derived renal cellcarcinoma cells fused with allogeneic dendritic cells elicit anti-tumor activity: in vitro results and clinical responses,” CancerImmunology, immunotherapy , vol. 58, no. 10, pp. 1587–1597,2009.

[58] L. Holtl, R. Ramoner, C. Zelle-Rieser et al., “Allogeneic den-dritic cell vaccination against metastatic renal cell carcinomawith or without cyclophosphamide,” Cancer Immunology,Immunotherapy, vol. 54, no. 7, pp. 663–670, 2005.

[59] D. Avigan, B. Vasir, J. Gong et al., “Fusion cell vaccina-tion of patients with metastatic breast and renal cancerinduces immunological and clinical responses,” Clinical Can-cer Research, vol. 10, no. 14, pp. 4699–4708, 2004.

[60] T. Luft, P. Luetjens, H. Hochrein et al., “IFN-α enhances CD40ligand-mediated activation of immature monocyte-deriveddendritic cells,” International Immunology, vol. 14, no. 4, pp.367–380, 2002.


[61] E. Padovan, G. C. Spagnoli, M. Ferrantini, and M. Heberer,“IFN-α2a induces IP-10/CXCL10 and MIG/CXCL9 produc-tion in monocyte-derived dendritic cells and enhances theircapacity to attract and stimulate CD8 effector T cells,” Journalof Leukocyte Biology, vol. 71, no. 4, pp. 669–676, 2002.

[62] H. Zhang, K. S. Chua, M. Guimond et al., “Lymphopeniaand interleukin-2 therapy alter homeostasis of CD4 CD25regulatory T cells,” Nature Medicine, vol. 11, no. 11, pp. 1238–1243, 2005.

[63] M. M. Horowitz, R. P. Gale, P. M. Sondel et al., “Graft-versus-leukemia reactions after bone marrow transplantation,” Blood,vol. 75, no. 3, pp. 555–562, 1990.

[64] H. J. Kolb, J. Mittermuller, C. Clemm et al., “Donor leukocytetransfusions for treatment of recurrent chronic myelogenousleukemia in marrow transplant patients,” Blood, vol. 76, no.12, pp. 2462–2465, 1990.

[65] M. Harano, M. Eto, T. Iwai et al., “Renal cancer treatment withlow levels of mixed chimerism induced by nonmyeloablativeregimen using cyclophosphamide in mice,” Cancer Research,vol. 65, no. 21, pp. 10032–10040, 2005.

[66] K. Tatsugami, M. Eto, M. Harano et al., “Dendritic-celltherapy after non-myeloablative stem-cell transplantation forrenal-cell carcinoma,” Lancet Oncology, vol. 5, no. 12, pp. 750–752, 2004.

[67] E. M. Shevach, “Certified professionals: CD4CD25 suppressorT cells,” Journal of Experimental Medicine, vol. 193, no. 11, pp.F41–F45, 2001.

[68] H. Jonuleit, E. Schmitt, G. Schuler, J. Knop, and A. H.Enk, “Induction of interleukin 10-producing, nonprolifer-ating CD4 T cells with regulatory properties by repetitivestimulation with allogeneic immature human dendritic cells,”Journal of Experimental Medicine, vol. 192, no. 9, pp. 1213–1222, 2000.

[69] S. Read and F. Powrie, “CD4 regulatory T cells,” CurrentOpinion in Immunology, vol. 13, no. 6, pp. 644–649, 2001.

[70] E. Y. Woo, H. Yeh, C. S. Chu et al., “Cutting edge: regulatoryT cells from lung cancer patients directly inhibit autologous Tcell proliferation,” Journal of Immunology, vol. 168, no. 9, pp.4272–4276, 2002.

[71] T. J. Curiel, G. Coukos, L. Zou et al., “Specific recruitmentof regulatory T cells in ovarian carcinoma fosters immuneprivilege and predicts reduced survival,” Nature Medicine, vol.10, no. 9, pp. 942–949, 2004.

[72] K. Tatsugami, M. Eto, and S. Naito, “Influence of im-munotherapy with interferon-α on regulatory T cells in renalcell carcinoma patients,” Journal of Interferon and CytokineResearch, vol. 30, no. 1, pp. 43–48, 2010.

[73] J. Shimizu, S. Yamazaki, and S. Sakaguchi, “Induction oftumor immunity by removing CD25+CD4+ T cells: a com-mon basis between tumor immunity and autoimmunity,”Journal of Immunology, vol. 163, no. 10, pp. 5211–5218, 1999.

[74] S. Onizuka, I. Tawara, J. Shimizu, S. Sakaguchi, T. Fujita, andE. Nakayama, “Tumor rejection by in vivo administration ofanti-CD25 (interleukin-2 receptor α) monoclonal antibody,”Cancer Research, vol. 59, no. 13, pp. 3128–3133, 1999.

[75] J. Dannull, Z. Su, D. Rizzieri et al., “Enhancement of vaccine-mediated antitumor immunity in cancer patients after deple-tion of regulatory T cells,” Journal of Clinical Investigation, vol.115, no. 12, pp. 3623–3633, 2005.


Research Article

Long-Term Follow-Up of HLA-A2+ Patients with High-Risk,Hormone-Sensitive Prostate Cancer Vaccinated with the ProstateSpecific Antigen Peptide Homologue (PSA146-154)

Supriya Perambakam,1 Hui Xie,2 Seby Edassery,3 and David J. Peace1

1 Section of Hematology and Oncology, Department of Medicine, College of Medicine Research Building, University of Illinois at Chicago,909 South Wolcott Avenue, Chicago, IL 60612, USA

2 Department of Epidemiology and Biostatistics, University of Illinois at Chicago, Chicago, IL 60612, USA3 Proteomics Core Research Facility, Rush University Medical Center, Chicago, IL 60612, USA

Correspondence should be addressed to Supriya Perambakam, [email protected]

Received 1 July 2010; Revised 7 December 2010; Accepted 14 December 2010

Academic Editor: Bartholomew Akanmori

Copyright © 2010 Supriya Perambakam et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Twenty-eight HLA-A2+ patients with high-risk, locally advanced or metastatic, hormone-sensitive prostate cancer wereimmunized with a peptide homologue of prostate-specific antigen, PSA146-154, between July 2002 and September 2004 andmonitored for clinical and immune responses. Fifty percent of the patients developed strong PSA146-154-peptide-specific delayed-type hypersensitivity skin responses, tetramer and/or IFN-γ responses within one year. Thirteen patients had stable or decliningserum levels of PSA one year post-vaccination. A decreased risk of biochemical progression was observed in patients who developedaugmented tetramer responses at six months compared to pre-vaccination levels (P = .02). Thirteen patients have died while 15patients remain alive with a mean overall survival of 60 months (95% CI, 51 to 68 months) per Kaplan-Meier analysis. Atrend towards greater overall survival was detected in men with high-risk, hormone-sensitive CaP who developed specific T-cellimmunity following vaccination with PSA146-154 peptide.

1. Introduction

Prostate cancer (CaP) is the second leading cause of cancer-related mortality in the United States. There were approx-imately 27,360 deaths caused by CaP in 2009 [1]. Patientswho recur after primary ablative therapy respond transientlyto androgen deprivation therapy (ADT) but subsequentlyprogress to hormone-refractory disease for which curativesystemic therapies are lacking [2]. Recent studies havedemonstrated that overall survival (OS) of patients withhormone refractory CaP can be modestly extended byvaccination with autologous dendritic cells (DC) loaded withrecombinant proteins consisting of granulocyte macrophagecolony stimulating factor and prostatic acid phosphatase[3]. It is widely assumed that improved outcomes might

be achieved by vaccinating patients at earlier points in thedevelopment of their disease at a time when host immuneeffector mechanisms remain robust.

Prostate-specific antigen (PSA) contains an HLA-A2-restricted epitope, PSA146-154, amino acid sequenceKLQCVDLHV, that is an attractive candidate for specific im-munotherapy of HLA-A2+ patients with CaP [4, 5]. Thesafety and immunogenicity of PSA146-154 peptide vacci-nation in patients with metastatic, hormone-sensitive CaP,or a disease that is at high risk of recurrence on the basisof tumor stage, serum PSA levels, and Gleason score havebeen previously reported [6]. Herein, we report the clinicaloutcome of patients up to eight years following vaccinationand correlate patients’ survival with their immunologicalresponses to the PSA146-154 vaccine.


Specific T-cell responses, defined by PSA146-154 pep-tide-tetramer staining and IFN-γ release assays, were quan-tified in pre- and postvaccine peripheral blood mononuclearcells (PBMC) and correlated with clinical parameters includ-ing biochemical progression and OS. In addition, microarraywhole human gene expression analysis was conducted toidentify differentially expressed genes and gene pathwaysin pre-vaccination PBMC that distinguish strong immuneresponders from nonresponders.


2.1. Patient Characteristics. Long-term follow-up of all pa-tients previously enrolled on a phase IB peptide vaccineprotocol was performed with the authorization of theInstitutional Review Board of the University of Illinois atChicago. Twenty-eight HLA-A2+ patients with pathologi-cally confirmed CaP who had completed vaccination withPSA146-154 peptide between July 2002 and September 2004were included in the study [6]. The clinical characteristicsof patients are listed in Table 1. All patients had undergoneradiotherapy or surgical ablation of the prostate a minimumof 6 weeks prior to initiation of vaccine study. Patients eitherhad advanced local disease with high risk of recurrencebased on the presence of T3, T4 disease, a serum PSA level≥10 ng/ml, or a Gleason grade ≥7 (Group A), or theyhad confirmed metastatic disease which was associated withdeclining serum PSA on ADT or a stable or improving bonescan or CT scan in response to hormone therapy (GroupB). All patients were immunologically reactive to a panel ofmumps, measles, and candida.

The unique patient identifying number (UPIN) assignedin the original report was retained. Relevant informationpertinent to morbidity, disease-specific mortality, and OSwas collected from patients and/or family members followingappropriate informed consent.

2.2. Vaccine Protocol and Dendritic Cell Culture. Patientswere either treated by intradermal administration of nativePSA146-154 peptide and GM-CSF (protocol 1, n = 14) orby intravenous administration of peptide-pulsed, autologousDC (protocol 2, n = 14) as previously detailed [6]. Patientswere vaccinated on three occasions (weeks 1, 4, and 10) andmonitored. DC was derived from monocyte and culturedin serum-free AIM-V (Life Technologies, Grand Island, NY)medium with IL-4 and GM-CSF for a total of 8 days in T-150 flasks in clinical grade sterile laminar airflow hood perthe method of Lau et al. [7]. Release criteria for the final DCproduct included sterile bacterial, fungal and mycoplasmacultures, negative endotoxin per Limulus Amoebocyte lysateassay, viability of at least 90% and greater than 50% CD86,CD80, HLA-DR, or CD1a positive cells, and less than 10%CD 14 positive cells by flow cytometric analysis. The finalDC product was divided into 3 equal parts. The first infusionincluded fresh DC while the 2nd and 3rd infusions consistedof frozen DC product. At the time of infusion, DC wererapidly thawed at 37◦C, again checked for sterility and via-bility, and administered intravenously to patients.

2.3. Delayed-Type Hypersensitivity Skin Testing. Immuneresponses were monitored by delayed-type hypersensitivity(DTH) skin testing on weeks 4, 14, 26, and 52 by intradermalinjection of 0 (carrier only), 1, 10 and 20 microgram ofpeptide dissolved in 200 microliter of 33% DMSO aspreviously detailed [6]. DTH reactions were measured at 48–72 hours following injection. An induration of ≥15 mm wasconsidered as a positive reaction.

2.4. T-Cell Culture Induced from Peripheral Blood Mononu-clear Cells. Frozen PBMC obtained at various study timepoints, prevaccine (1 to 3 weeks prior to vaccination), week26, and week 52, were rapidly thawed, washed, checked forviability, and resuspended in RPMI-1640 medium (BioWhit-taker, Walkersville, MD) containing 10% human AB serum(complete medium). Viability was ≥90% (range 90 to 99%,mean 95± 1.26). PBMC (2× 106) were plated in 24 wellplates (Nunc, Naperville, IL) and cultured in completemedium containing PSA146-154 peptide (20 ug/mL) and IL-2 (20 U/mL) for 7± 1 days (1 cycle). PBMC were alternativelystimulated with HLA-A2 binding control peptide, Flu-M1, insome patients. Spent medium was aspirated and replenishedwith complete medium plus IL-2 and restimulated withirradiated autologous PBMC pulsed with peptide for 2additional cycles prior to tetramer and cytokine analysis.

2.5. Tetramer Analysis. PSA146-154 peptide stimulatedPBMC (1× 106 per tube) were doubly stained with PSA146-154 peptide-tetramer-PE (Immunomics, San Diego, CA)and CD8-FITC (BD Biosciences, San Diego, CA) at roomtemperature for 30 minutes in phosphate-buffered salinecontaining 0.5% para-formaldehyde (Sigma, St. Louis, MO).Cells were washed, resuspended in buffer, and analyzedby a Calibur flow cytometer (Becton Dickinson, MountainView, CA). Cells also were stained separately with a negativecontrol tetramer-PE, of unknown sequence that does notrecognize CD8+ T-cells of any HLA allele type, to assess thelevel of background PE fluorescence. As a positive control,tetramer-PE staining for Flu-M1 peptide also was performedin some patients. The percentage of CD8+ tetramer+ cellswas determined from the quadrant dot plots per CellQuest software (Becton Dickinson, Mountain View, CA).The results were represented as the number of tetramer+

cells per CD8+ cells and are calculated as the number oftetramer+CD8+ cells divided by total number of CD8+ cells.

2.6. Cytokine Bead Array Analysis. PSA146-154 peptidestimulated PBMC also were evaluated for specific release ofcytokines following recognition of peptide-pulsed targets.Cytokines released into the culture supernatant, including,IFN-γ, TNF-α, IL-4, IL-5, and IL-10, were measured concur-rently by cytokine bead array analysis (CBA, BD Biosciences,San Diego, CA) as described earlier [6]. Briefly, the antigenpresenting cell line, T2 (ATCC, Manassas VA), was usedas a stimulator and was pulsed with 20 μg/ml of PSA-peptide or control HLA-A2 binding peptide, HIV-RT476-484 or diluent alone (0.4% volume by volume). T2 cells(25,000/well) were cultured with T-cells (100,000/well) in


Table 1: Patient baseline characteristics.

Characteristic Group A Group B Protocol-1 Protocol-2 Total

n = 14 n = 14 n = 14 n = 14 n = 28Age

median (average) 61.5 (62.2) 62 (64) 64.5 (65.2) 60.5 (61) 62

range 51–73 51–80 51–80 51–75 51–80

Race

white 10 13 12 11 23

black 2 1 1 2 3

other 2 0 1 1 2

ECOG PS

0 or 1 14 14 14 14 28

2 or 3 0 0 0 0 0

Disease status

undetectable (PSA 0) 3 4 5 2 7

measurable 0 7 2 5 7

increased PSA only 11 3 7 7 14

Sites of disease

bone 0 4 2 2 4

soft tissue 0 4 0 4 4

Family history

positive 4 3 5 2 8

negative 9 9 6 12 17

unknown 1 1 2 0 3

Gleason score

median (average) 7 (7.14) 7 (7.3) 7 (7.2) 7.5 (7.25) 7

range 4–9 5–10 5–10 4–9 4–10

PSA at diagnosis

median (average) 5.8 (9.1) 15.25 (28.4) 8.4 (12.3) 13.4 (24.3) 10.5

range 4–23.4 3.4–139 <4–23.4 3.4–139 3.4–139

PSA at study entry

median (average) 0.32 (3.66) 0.4 (2.6) 0.4 (2.75) 0.4 (3.5) 0.4

range 0–12.5 0–13.8 0–12.1 0–13.8 0–13.8

Local therapy

RPE 2 4 2 4 6

RPE + EBRT 7 4 8 3 11

EBRT, primary 2 4 2 4 6

other 3 2 2 3 5

Hormone Rx

none 10 0 5 5 10

first line 4 11 8 7 15

second line 0 2 1 1 2

≥3 therapies 0 1 0 1 1

Basal biochemistry

Alkaline phosphatase 48–90 49–105 49–90 50–105 48–105

Haemoglobin 12.8–16.6 11.5–16.9 11.5–15.7 12.2–16.9 11.5–16.9

Creatinine 0.7–1.3 0.8–1.2 0.8–1.3 0.7–1.3 0.7–1.3

All patients had completed primary therapy a minimum of 6 weeks prior to enrollment in the vaccine study.

complete medium containing 30 U/ml of IL-2 in a totalvolume of 1 ml per well in 48-well plates. This particularstimulator to responder ratio was found to be optimal forculture in 48-well plates. Cells were incubated at 37◦C for 24

hours in 5% CO2 atmosphere. Supernatants were harvestedand stored in sterile vials at −80◦C. At the time of assay,samples were thawed and cytokines were measured using aCBA kit as per the manufacturer’s protocol with a Calibur


flow cytometer. Results are represented as net cytokine levels(pg/mL) which were obtained by subtracting nonspecificbackground responses (T2 cells pulsed with HIV-RT476-484or diluent).

2.7. Microarray and Bioinformatic Analysis. Total RNA wasextracted from unmanipulated prevaccine PBMC samplesof representative patients using RNeasy mini kit (Qiagen,Valencia, CA). The quantity and quality of RNA were esti-mated with a NanoDrop 3300 Fluorospectrometer (ThermoFisher Scientific, Waltham, MA) and an Agilent bioanalyzer(Agilent Technologies, Santa Clara, CA), respectively. AllRNA samples were stored at −80◦C. Microarray analysiswas performed at the functional Genomics Laboratory ofthe University of Illinois at Urbana, Champaign, using thehuman genome U133 plus 2.0 chip (Affymetrix, Santa Clara,CA). Data was extracted from the Affymetrix array andnormalized by the Robust Multichip Average (RMA) method[8]. All appropriate internal quality control was performedas per the guidelines for microarray gene expression studies.Class comparison analysis was conducted per the BiometricResearch Branch (BRB) array tool (National Cancer Insti-tute, Bethesda, MD). Gene expression data was comparedbetween strong immune responders (UPIN13, UPIN28,UPIN40, UPIN45, and UPIN71-positive DTH and tetramerresponses) and nonresponders (UPIN32, UPIN35, UPIN37,and UPIN70-negative DTH and tetramer responses).

2.8. Clinical Evaluation. The disease status of patients wasmonitored by clinical examination and serial serum PSAmeasurements on weeks 1, 4, 7, 14, 26, and 52. Biochemicalprogression (P) was defined as at least a 20% increase inserum PSA at week 52 over week 1 (study entry) with anabsolute PSA value ≥0.2 ng/mL. Stable biochemical diseaseor nonprogression (NP) was defined as less than a 20%increase in serum PSA over week 1 with an absolute PSAvalue less than 0.2 ng/mL.

Survival status was established for all 28 vaccinatedpatients by review of the Social Security Death RegistryIndex and by direct contact of patients or their relatives.Time (in months) from the onset of vaccine therapy (week1) till death or until May 1, 2010 for patients who weredeceased or surviving, respectively, was calculated followedby computation of OS per Kaplan-Meier analysis (SASsoftware version 9.2, Cary, NC). The median follow-upperiod was 6.30 years (mean 5.36 years; range 1.35 to 7.68years).

2.9. Statistical Analysis. A marginal longitudinal model wasused to compare tetramer or cytokine measurements overtime within similar groups of patients. The dependentvariable was the log of the tetramer values or cytokine mea-surements. The independent variables included intercept,group, time dummies, and interactions between group andtime dummies. Spearman analysis was used to evaluate thecorrelation of tetramer or cytokines values with serum PSAstatus. The two-sample t-test with unequal variance was usedto identify genes that were differentially expressed between

immune responders and nonresponders per BRB array tools.OS was evaluated per Kaplan-Meier analysis. Log-rank testswere used to evaluate differences in survival curves.

3. Results

3.1. Dendritic Cell Product. Two healthy donors and 14patients underwent 7–9 liter leukapheresis, and DC werecultured for 8 days under identical conditions and pheno-typed. The average HLA-DR% was 54.51 (median 52.92),the average CD86% was 58.77 (median 62.56), the averageCD1a% was 28.17 (median 30.95), and the average CD14%was 1.31 (median 0 or negative expression). DC productwas also phenotyped for CD80 and CD83. However, only2 of 14 patients’ DC showed CD80 expression, while CD83was negative in all the patients. The average percent HLA-DR, CD86, CD1a, and CD14 were 70.03%, 76.6%, 30.58%and 5.94%, respectively, in healthy individuals. The yieldof total DC from PBMC ranged from 0.94 to 2.02×108

cells (average 1.499, median 1.555) per vaccine in the14 patients. Functional activity of DC product also wastested in several patients. DC, cultured in IL-4/GM-CSFfor 8 days were able to stimulate significant (>20-fold)allogeneic T-cell proliferative responses compared to DC-pulsed autologous T-cells. Additionally, upon maturationwith TNF-tide induced release of IFN or LPS for 24 hours,the expression of CD83, a late DC marker, was up-regulated(negative expression to 25% expression).

3.2. Immunological Responses. Three distinct readouts wereused to detect specific immune responses. First via theinduction of DTH skin responses to PSA146-154 peptidein vivo, second via detection of CD8+ PSA146-154 peptide-tetramer+ T-cells, and third via PSA146-154 peptide inducedrelease of IFN-γ in pre- versus postvaccine PBMC samples.In vitro sensitization of PBMC with PSA146-154 peptide wasessential prior to tetramer and CBA analysis to detect specificT-cells in peripheral blood. This procedure was applieduniformly to all specimens and was necessary to overcomehigh background. Similar techniques have been employed inprevious cancer vaccine trials [7, 9]. Lau et al. have showninduction of peptide-specific CTL stimulated twice withmelanoma-associated peptides over 24 days in IFN-γ ELISA[7]. Meidenbauer et al. have shown PSA-reactive responsesper IFN-γ ELISPOT following two stimulations in patientswith prostate cancer [9].

Overall, fifty percent of patients demonstrated positiveDTH skin responses to PSA146-154 peptide (Table 2). Spe-cific DTH responses were negative in a majority of patients(13 of 14 patients) when they were tested initially at week4; however, measurable induration became evident overtime and increased with successive DTH testing. Responseswere dose-dependent with increasing doses of the PSA146-154 peptide eliciting increasing degrees of induration inresponding patients [6]. Injection of carrier only, that is, 33%DMSO, did not cause significant induration. Both CD4+and CD8+ T-cells were derived from the positive DTH skin


0

200

400

600

800

1000

SSC

-hei

ght

0 200 400 600 800 1000

FSC-height

Pre-vaccine

R1

(a)

0

200

400

600

800

1000

SSC

-hei

ght

0 200 400 600 800 1000

FSC-height

Post-vaccine (week 26)

R1

(b)

100

101

102

103

104

0.09%

FL2-

heig

ht

100 101 102 103 104

FL1-height

(c)

100

101

102

103

104

3.22%

FL2-

heig

ht

100 101 102 103 104

FL1-height

(d)

100

101

102

103

104

0.05%

FL2-

heig

ht

100 101 102 103 104

FL1-height

(e)

100

101

102

103

104

0.03%

FL2-

heig

ht

100 101 102 103 104

FL1-height

(f)

Figure 1: A representative flow cytometric data showing the detection of CD8+ PSA146-154 peptide-tetramer+ cells in patient UPIN28.PBMC were sensitized in vitro with PSA146-154 peptide for 3 cycles, and resulting T-cells were doubly stained with PSA146-154 peptide-tetramer-PE ((c) and (d)) or negative control tetramer-PE ((e) and (f)) and CD8-FITC ((e), (f), (c), and (d)). A greater number of CD8+PSA146-154 peptide-tetramer+ cells ((a) and (b)) were observed on postvaccine compared to prevaccine samples.


Table 2: Immunological outcomes based on specific DTH, tetramer, and IFN-γ responses.

PatientCode

DTHresponderspositive (+)negative (−)

Fold increase in tetramer

Tetramerresponderspositive (+)negative (−)

Absolute change in IFN-γ

IFN-γresponder

positive (+)negative (−)

week 26 week 52 week 26 week 52

UPIN13 + 22.25 34.49 + 141.4 0 +

UPIN16 + 22.77 121.29 + 44.4 241.6 +

UPIN28 + 29.25 12.5 + 525.4 847.9 +

UPIN40 + 15.85 2.47 + 313.5 488.9 +

UPIN45 + 11.39 6.97 + 30.6 88.8 −UPIN49 + 1.75 1.71 − 0 113.5 +

UPIN50 + 6.11 3.8 + −20.7 −20.7 −UPIN51 + 0.69 0.26 − 0 0 −UPIN53 + 1.58 0.69 − 25.3 2417.4 +

UPIN55 + 2.71 5.28 + 262.7 66 −UPIN69 + 2.09 0.23 − 1293.5 133.4 +

UPIN71 + 4.91 4.82 + 20.9 31 −UPIN81 + 1.72 0.09 − −2.9 −2.9 −UPIN88 + 1.42 0.06 − −34.5 230.4 +

UPIN2 − 3.09 5.83 + 63.9 −50.2 −UPIN21 − 0.51 5.43 + 1064 −9.3 +

UPIN26 − 0.35 0.81 − 255.5 1211.4 +

UPIN27 − 82.11 ND + 2236 ND +

UPIN32 − 1.13 1.05 − 0 0 −UPIN35 − 1.31 0.09 − −10.4 24.3 +

UPIN37 − 0.80 0.26 − 1.4 0 −UPIN38 − 10.79 1.37 + 112.1 1.3 +

UPIN43 − 1.11 5.06 + −3.5 −11.2 −UPIN67 − 1.72 0.07 − 40.3 0 −UPIN70 − 0.46 1.85 − 130.9 0 +

UPIN82 − 4.47 0.24 + −3.2 −3.2 −UPIN85 − 2.50 0.08 − 0 0 −UPIN89 − 2.93 0.02 − −46.9 −46.9 −

Fourteen of 28 (50%) patients developed positive tetramer, IFN-γ, and/or DTH responses to PSA146-154 peptide by week 52. A positive tetramer response isdefined as ≥4-fold increase in tetramer levels by week 52 over prevaccine levels, while positive IFN-γ response was defined as ≥100 ng/ml of absolute changein cytokine levels at week 26 or 52 minus prevaccine levels. A positive DTH reaction is defined as ≥15 mm of induration to PSA146-154 peptide. A stringentcutoff value was taken into consideration to measure true immune responses and to avoid false positives.

biopsy that demonstrated specific cytolytic and cytokineactivity as detailed in a previous publication [6].

Fourteen of 28 patients developed ≥4-fold increase inCD8+ PSA146-154-tetramer+ T-cells at week 26 and/orweek 52 over baseline levels (Table 2). On average, 3.5 CD8+PSA146-154-tetramer+ T-cells were observed for every 100CD8+ T-cells at week 26, while an average of 2.0 CD8+PSA146-154-tetramer+ T-cells were detected for every 100CD8+ T-cells at week 52. On an average, 1.0 CD8+ PSA146-154-tetramer+ T-cell could be detected per 100 CD8+ T-cells prior to the onset of immunotherapy. Figure 1 is arepresentative tetramer staining analysis showing increasedCD8+ PSA146-154-tetramer+ T-cells postvaccine (week 26)compared to prevaccine following in vitro sensitizationof PBMC with PSA146-154 peptide. Tetramer responses

were not detectable in unstimulated PBMC population.Comparable results were observed by Lau and coworkers ina peptide-DC based melanoma study [7].

Similarly, 14 of 28 patients demonstrated specific releaseof IFN-γ (defined as ≥100 ng/ml of absolute change) byweek 52 from the outset of immunotherapy. Specific releaseof other cytokines, namely, TNF-α, IL-4, and IL-5 also wasobserved (see Table 1 in supplementary material availableonline at doi:10.1155/2010/473453). The CBA analysis in thecurrent study was performed with unsorted T-cell popula-tions; therefore, it is not possible to determine whether IFN-γ was released by CD8+ and/or CD4+ T-cells.

Eight of 14 (57%) positive tetramer responders alsomounted specific DTH responses to PSA146-154 peptide,while only 4/14 (28%) tetramer nonresponders were positive


−0.04

−0.02

0

0.02

0.04

0.06

0.08

0.1

0.12

Tetr

amer

leve

ls

NP P

Serum PSA status

Figure 2: Correlation between augmented-specific tetramerresponses and serum PSA status. The average tetramer mea-surements at week 26 minus prevaccine levels (Δ26) inverselycorrelated with lower risk of serum PSA progression at six monthsfollowing the onset of immunotherapy (P = .02). “NP” denotesstable biochemical disease or nonprogression, while “P” denotesbiochemical progression.

for DTH responses to the peptide, indicating concordancebetween the development of peptide-specific DTH responsesin the skin and specific T-cell immune responses in periph-eral blood of patients.

3.3. Clinical Outcomes: Toxicity, Serum PSA, and

Survival Status

3.3.1. Toxicity. Both methods of vaccination were well toler-ated with no treatment-related grade 3/4 toxicities, gradedaccording to the NIH Common Terminology Criteria forAdverse Events, version 3.0. Mild pain, itching, and erythemawith or without transient induration were observed at the siteof injection for all patients treated under protocol-1. Therehave been no late safety concerns or deleterious sequelaeidentified after six to eight years of monitoring.

3.3.2. PSA Progression. Thirteen of 27 (48.1%) patientsmanifested stable or declining serum PSA, while 14 of 27(51.6%) patients evidenced PSA progression at one yearfollowing the initiation of PSA146-154 peptide vaccinetherapy. One patient, UPIN27, did not return for follow-up at week 52 and hence his biochemical status was notevaluable. However, the survival status was determinable inall 28 patients. As of May 1 2010, 15 of 28 (54%) patientswere alive while 13 (46%) patients had died. In most patients,death was CaP specific; however, one patient, UPIN16, diedof late occurring esophageal cancer.

3.3.3. Survival. OS is the most definitive standard to assessthe outcome of anticancer therapies and was determinedper Kaplan-Meier analysis eight years after the initiation ofthe protocol. The median follow-up period for individualpatients was 6.30 years (range 1.35 to 7.68 years) from theonset of immunotherapy. The mean OS was 60 months(95% CI 51 to 68 months) for all patients (Figure 3(a)).

0

0.25

0.5

0.75

1

All patients (n = 28)

Surv

ival

dist

ribu

tion

fun

ctio

n

0 20 40 60 80 100

Time (months)

(a)

0

0.25

0.5

0.75

1

Surv

ival

dist

ribu

tion

fun

ctio

n

High risk,locally advanced (n = 14)

0 20 40 60 80 100

Time (months)

(b)

0

0.25

0.5

0.75

1Metastatic, hormone-sensitive

patients (n = 14)

Surv

ival

dist

ribu

tion

fun

ctio

n

0 20 40 60 80 100

Time (months)

(c)

Figure 3: Overall Survival for high risk, locally advanced andmetastatic hormone-sensitive CaP. The mean OS was 60 months(95% CI 51 to 68 months) for all patients (a). The median OS wasgreater than 84 months for patients with high risk, locally advanceddisease (b), while the median OS was 75 months for patients withmetastatic, hormone-sensitive CaP (c) at a median follow-up of 6.30years since the onset of immunotherapy.

The median OS has not yet been reached for patients withhigh risk, locally advanced disease and exceeds 84 months,Figure 3(b). The median OS was 75 months for patients withmetastatic, hormone-sensitive CaP (Figure 3(c)).

3.4. Correlation of Clinical Outcome with the Induction ofSpecific Immune Responses. The development of specificT-cell immune responses was correlated with patients’ serumPSA and survival status. The results indicate that thedifference between average tetramer measurements at week26 and at baseline inversely correlated with changes in serumPSA levels (Figure 2, P = .02). Thus, a decreased risk of bio-chemical progression was observed in patients who devel-oped augmented tetramer responses at six months comparedto pre-vaccination levels. No significant correlation remained


0

0.25

0.5

0.75

1

DTH responders (n = 14)

DTH non responders (n = 14)

Surv

ival

dist

ribu

tion

fun

ctio

n

0 20 40 60 80 100

Time (months)

(a)

0

0.25

0.5

0.75

1

Tetramer responders (n = 14)

Tetramer non responders (n = 14)

Surv

ival

dist

ribu

tion

fun

ctio

n

0 20 40 60 80 100

Time (months)

(b)

0

0.25

0.5

0.75

1

IFN-γ non responders (n = 14)

IFN-γ responders (n = 14)

Surv

ival

dist

ribu

tion

fun

ctio

n

0 20 40 60 80 100

Time (months)

(c)

Figure 4: Comparison of overall survival between immune respon-ders versus nonresponders. There was a trend towards greater OSin men with high-risk, hormone-sensitive CaP who developedstrong specific DTH or tetramer response following vaccinationwith PSA146-154 peptide.

at one year, as specific immune responses became attenuatedover time.

OS of patients who developed positive DTH responses,tetramer or, IFN-γ responses to PSA146-154 peptide versusthat of patients who did not develop specific immuneresponses were correlated by log-rank testing. The mean OSwas 58 months (95% CI, 50 to 66 months) for strong DTHresponders versus 54 months (95% CI, 41 to 68 months)for nonresponders (P = .21). The mean OS was 61 months(95% CI, 50 to 71 months) in patients who showed strongtetramer responses versus 44 months (95% CI, 35 to 52months) for nonresponders (P = .46). The mean OS was 61months (95% CI, 50 to 73 months) in patients who showedstrong IFN-γ responses versus 55 months (95% CI, 43 to68 months) for nonresponders (P = .65). Although thesefindings did not reach statistical significance, the patients

who developed strong T-cell immunity in terms of specificDTH and tetramer responses to PSA146-154 peptide withinone year following vaccination demonstrated a trend towardsgreater OS (Figure 4).

3.5. Gene Expression Profiles of Immune Responders versusNonresponders. Affymetrix human genome U133 plus 2.0chips array analysis was performed on prevaccine PBMC,in order to identify genes and gene pathways that aredifferentially expressed between patients who developedstrong PSA146-154 peptide-specific immune responses ver-sus patients who did not. Immune responders includedpatients with strong tetramer (>4.9 fold) responses inconjunction with a positive DTH skin reaction to thePSA146-154 peptide, while nonresponders included patientswho were negative for both tetramer and DTH responses.

Class comparison analysis per BRB array tools revealedthat 166 of 54,675 genes were differentially expressed at asignificance level of P < .005 (Supplemental Table 2). Pre-dictably, the gene ontology class belonging to the biologicalprocess category of immune system development (GOID: 0002520) was affected with an observed to expectedratio of 2.1. Of the 166 differentially expressed genes, 12genes were members of the immune function associatedpathway (Table 3). A 4-fold increase in 2′–5′ oligoadenylatesynthetase 1 (OAS1) was noted in immune respondersversus nonresponders. Other genes that were overexpressedincluded mitogen-activated protein kinase 1, Sh2 domaincontaining 1B, vannin 1, CD58 molecule, and interferon-induced transmembrane protein-3. Tumor necrosis factorreceptor superfamily-member 25, chemokine C-C motifreceptor 7 and phosphoinositide-3-kinase, regulatory sub-unit 1 alpha genes, and epiregulin showed lower expressionin immune responders versus nonresponders.

4. Discussion

The field of cancer immunotherapy recently reached anexciting milestone with the approval of the first therapeu-tic cancer vaccine by the United States Food and DrugAdministration. Sipuleucel-T (Provenge), an autologouscellular immunotherapeutic product that is designed tostimulate T-cell immunity to prostatic acid phosphatase, wasfound to improve the median overall survival of patientswith metastatic, castration resistant CaP by 4.1 months(25.8 versus 21.7 months for placebo) [3, 10]. Similarly,immunotherapy with PROSTVAC-VF, a PSA-based viralvaccine construct, also improved survival in a phase 2trial for patients with metastatic, castration-resistant CaP[11]. These studies have highlighted the potential of specificimmunotherapy for the management of patients with CaP.

To date, the majority of tumor vaccines have beenevaluated in patients with the most advanced forms ofdisease. In the current study, we observed the developmentof specific T-cell immunity in terms of increased peptide-specific tetramer and IFN-γ responses (≥4-fold increaseor ≥100 pg/ml fold change, resp.) in 50% of patientsvaccinated at points in the spectrum of prostate cancer that


Table 3: Differentially expressed genes between immune responders and nonresponders∗.

Gene name (symbol) Probe set Fold changeAffected immune-function associated

pathway

2′–5′ oligoadenylate synthetase 1 (OAS1)202869 at

205552 s at4.052.48

Innate immune response

Vannin 1 (VNN1) 205844 at 2.74Innate immune response

Positive regulation of T-celldifferentiation in the thymus

Sh2 domain containing 1B (SH2D1B) 1553176 at 1.92 Natural killer cell mediated cytotoxicity

DEAD box polypeptide 58 (DDX58) 218943 s at 1.82 Innate immune response

Interferon-induced transmembraneprotein 3 1–8 U (IFITM3)

212203 x at 1.55 Immune response

Mitogen-activated protein kinase 1(MAPK1)

1552263 at1552264 a at

1.531.37

T-cell and B-cell receptor signalingVEGF signaling pathway TGF-beta

signalingnatural killer mediated cytotoxicity

CCR3 signaling in eosinophilsCXCR4 signaling pathway

CD58 molecule (CD58)216942 s at205173 x at

1.461.48

IL-17 signaling pathway

X-ray repair complementing defectiverepair in Chinese hamster cells 4(XRCC4)

210813 s at 1.38T-cell differentiation in the thymus

Immunoglobulin V(D)J recombination

Tumor necrosis factor receptorsuperfamily, member 25

211841 s at 0.65 Cytokine-cytokine receptor interaction

Chemokine C-C motif receptor 7 (CCR7) 206337 at 0.52 Cytokine-cytokine receptor interaction

Phosphoinositide-3-kinase, regulatorysubunit 1 alpha (PIK3R1)

212249 at 0.62

T-cell activationT-cell and B-cell receptor signaling

CXCR4 signaling pathwayVEGF signaling pathway

Toll-like receptor signaling pathway

Epiregulin (EREG) 205767 at 0.26positive regulation of innate immune

response

Gene expression analysis was performed on unmanipulated pre-vaccination PBMC. ∗Of the 166 genes differentially expressed, only genes affecting theimmune function associated pathway are shown.

precede the development of castrate-resistance. Importantly,patients who developed augmented tetramer responses at sixmonths compared to pre-vaccination levels had a decreasedrisk of biochemical progression at one year following theonset of immunotherapy. The inclusion of patients withhormone-sensitive disease who are immunologically robust,as reported here, may be key to harnessing the full potentialof novel vaccine regimens.

In the current study, 15 of 28 (54%) patients were alive ateight years from the initiation of the protocol while 13 (46%)patients had died. Of note, a trend towards greater survival inmen with high-risk, hormone-sensitive CaP who developedstrong specific DTH or tetramer responses following vacci-nation with PSA146-154 peptide was observed. Two previouscancer vaccine studies conducted in hormone-refractoryCaP patients showed that survival positively correlated withthe induction of specific immune responses [12, 13]. Thedemonstration of statistically significant survival advantagesby immunization of hormone-sensitive CaP patients withlonger life expectancies will require extended periods ofobservation and expanded patient cohorts.

The availability of quantitative metrics for monitoringthe induction of specific T-cell immunity to defined targetantigens as in the present study should provide an importantsurrogate for gauging vaccine efficacy, if a causal relationshipbetween the induction of specific T cell immunity andsurvival advantages can be definitively established. This inturn would speed vaccine optimization for early phasesof CaP. It also is critical to establish formal standardsfor reporting immunomonitoring in clinical trials to avoidvariations between laboratories and also among variousassays. However, this will entail storage of large amounts ofsamples to ensure the ability to conduct validating studiesand retrospectively apply emerging assays that becomerelevant over time.

DC are central to successful vaccination and can bedirectly targeted in vivo with antigen and adjuvants, such asGM-CSF, as demonstrated in early pioneering studies [14–16]. Alternatively, ex vivo generated monocytic or CD34-derived DC loaded with tumor antigen can be utilized forspecific active immunotherapy of cancer patients [17–20].However, DC-based vaccine formulations involve laborious


manipulations ex vivo and incur considerable cost. In thecurrent study, the efficacy of PSA146-154 peptide vaccine byboth techniques was compared in a randomized fashion. Thefinding that a simple method of intradermal vaccination isefficacious has important implications for the affordabilityand applicability of the technique to the general population.These results were corroborated by a similar study, wherein,intradermal injection of E75 HER2/neu peptide plus GM-CSF was found to be efficacious in high-risk node positivebreast cancer patients [21].

The present study shows for the first time, that a setof molecular determinants expressed within PBMC distin-guish immune responders and nonresponders undergoingvaccination with a peptide-based cancer vaccine. Genomicand bioinformatics analysis revealed 166 genes that weredifferentially expressed between strong immune respondersversus nonresponders. In particular, genes associated withinnate immune response were over-expressed, including,OAS1, which belongs to a family of IFN-stimulated proteins[22]. Interestingly, OAS1 also is postulated to be associatedwith radiation resistance in human breast cancer and CaPcell lines and with the regulation of cell growth in mammaryand prostate glands [23, 24]. Understanding the molecularintricacies of why some patients respond to a well definedpeptide target, while others do not, should lead to theapplication of optimal vaccine strategies for appropriatelyselected patients and shed light on novel strategies to maketargeted immunotherapy applicable to a wider array ofpatients.

Acknowledgments

Authors greatly appreciate the enthusiasm and active partic-ipation of patients and their families. Authors are indebtedto Dr. Nadim Mahmud, Director of Stem Cell laboratory,University of Illinois at Chicago, for overseeing the clinicalgrade dendritic cell production. The clinical trial was fundedby Grants from the National Cancer Institute (CA88062) andthe Department of Army (DAMD17-98-1-8489). Fundingfor conducting correlative studies was supported by Grantsfrom the Illinois Department of Public Health (IDPH4328301) and the Milheim Grant for Cancer Research,Denver, CO (award no. 2007-24). Statistical analysis wasmade possible by a grant from the National Center forResearch Resources (Grant no. UL1RR029879) awarded tothe Center for Clinical and Translational Science, Universityof Illinois at Chicago. The contents in the manuscript aresolely the responsibility of the authors and do not necessarilyreflect the official views of the funding agencies.

References

[1] A. Jemal, R. Siegel, E. Ward, Y. Hao, J. Xu, and M. J. Thun,“Cancer statistics, 2009,” CA-Cancer Journal for Clinicians, vol.59, no. 4, pp. 225–249, 2009.

[2] P. Pronzato and M. Rondini, “Hormonotherapy of advancedprostate cancer,” Annals of Oncology, vol. 16, no. 4, pp. iv80–iv84, 2005.

[3] P. Kantoff, C. S. Higano, E. R. Berger et al., “Updated survivalresults of the IMPACT trial of sipuleucel-T for metastaticcastration-resistant prostate cancer (CRPC),” in Proceedings ofthe Genitourinary Cancers Symposium of the American Societyof Clinical Oncology, 2010, Abstract #8.

[4] B. H. Xue, Y. Zhang, J. A. Sosman, and D. J. Peace, “Inductionof human cytotoxic T lymphocytes specific for prostate-specific antigen,” Prostate, vol. 30, no. 2, pp. 73–78, 1997.

[5] S. Perambakam, B. H. Xue, J. A. Sosman, and D. J. Peace,“Induction of Tc2 cells with specificity for prostate-specificantigen from patients with hormone-refractory prostate can-cer,” Cancer Immunology, Immunotherapy, vol. 51, no. 5, pp.263–270, 2002.

[6] S. Perambakam, S. Hallmeyer, S. Reddy et al., “Induction ofspecific T cell immunity in patients with prostate cancer byvaccination with PSA146-154 peptide,” Cancer Immunology,Immunotherapy, vol. 55, no. 9, pp. 1033–1042, 2006.

[7] R. Lau, F. Wang, G. Jeffery et al., “Phase I trial of intravenouspeptide-pulsed dendritic cells in patients with metastaticmelanoma,” Journal of Immunotherapy, vol. 24, no. 1, pp. 66–78, 2001.

[8] R. A. Irizarry, B. M. Bolstad, F. Collin, L. M. Cope, B. Hobbs,and T. P. Speed, “Summaries of Affymetrix GeneChip probelevel data,” Nucleic acids research, vol. 31, no. 4, p. e15, 2003.

[9] N. Meidenbauer, D. T. Harris, L. E. Spitler, and T. L. Whiteside,“Generation of PSA-reactive effector cells after vaccinationwith a PSA- based vaccine in patients with prostate cancer,”Prostate, vol. 43, no. 2, pp. 88–100, 2000.


[11] P. W. Kantoff, T. J. Schuetz, B. A. Blumenstein et al., “Overallsurvival analysis of a phase II randomized controlled trial ofa Poxviral-based PSA-targeted immunotherapy in metastaticcastration-resistant prostate cancer,” Journal of Clinical Oncol-ogy, vol. 28, no. 7, pp. 1099–1105, 2010.

[12] A. K. Thomas-Kaskel, R. Zeiser, R. Jochim et al., “Vaccinationof advanced prostate cancer patients with PSCA and PSApeptide-loaded dendritic cells induces DTH responses thatcorrelate with superior overall survival,” International Journalof Cancer, vol. 119, no. 10, pp. 2428–2434, 2006.

[13] J. L. Gulley, P. M. Arlen, R. A. Madan et al., “Immunologic andprognostic factors associated with overall survival employinga poxviral-based PSA vaccine in metastatic castrate-resistantprostate cancer,” Cancer Immunology, Immunotherapy, vol. 59,no. 5, pp. 663–674, 2010.

[14] A. K. Palucka, H. Ueno, J. W. Fay, and J. Banchereau,“Taming cancer by inducing immunity via dendritic cells,”Immunological Reviews, vol. 220, no. 1, pp. 129–150, 2007.

[15] M. L. Disis, H. Bernhard, F. M. Shiota et al., “Granulocyte-macrophage colony-stimulating factor: an effective adjuvantfor protein and peptide-based vaccines,” Blood, vol. 88, no. 1,pp. 202–210, 1996.

[16] M. L. Disis, T. A. Gooley, K. Rinn et al., “Generation of T-cell immunity to the HER-2/neu protein after active immu-nization with HER-2/neu peptide-based vaccines,” Journal ofClinical Oncology, vol. 20, no. 11, pp. 2624–2632, 2002.

[17] F. O. Nestle, S. Alijagic, M. Gilliet et al., “Vaccination ofmelanoma patients with peptide- or tumor lysate-pulseddendritic cells,” Nature Medicine, vol. 4, no. 3, pp. 328–332,1998.


[18] B. Thurner, I. Haendle, C. Roder et al., “Vaccinationwith Mage-3A1 peptide-pulsed nature, monocyte-deriveddendritic cells expands specific cytotoxic T cells andinduces regression of some metastases in advanced stage IVmelanoma,” Journal of Experimental Medicine, vol. 190, no. 11,pp. 1669–1678, 1999.

[19] J. Banchereau, A. K. Palucka, M. Dhodapkar et al., “Immuneand clinical responses in patients with metastatic melanomato CD34 progenitor-derived dendritic cell vaccine,” CancerResearch, vol. 61, no. 17, pp. 6451–6458, 2001.

[20] J. M. Timmerman, D. K. Czerwinski, T. A. Davis et al.,“Idiotype-pulsed dendritic cell vaccination for B-cell lym-phoma: clinical and immune responses in 35 patients,” Blood,vol. 99, no. 5, pp. 1517–1526, 2002.

[21] G. E. Peoples, J. M. Gurney, M. T. Hueman et al., “Clinical trialresults of a HER2/neu (E75) vaccine to prevent recurrence inhigh-risk breast cancer patients,” Journal of Clinical Oncology,vol. 23, no. 30, pp. 7536–7545, 2005.

[22] I. H. Haralambieva, N. Dhiman, I. G. Ovsyannikova et al.,“2′-5′-Oligoadenylate synthetase single-nucleotide polymor-phisms and haplotypes are associated with variations inimmune responses to rubella vaccine,” Human Immunology,vol. 71, no. 4, pp. 383–391, 2010.

[23] M. H. Tsai, J. A. Cook, G. V. R. Chandramouli et al.,“Gene expression profiling of breast, prostate, and glioma cellsfollowing single versus fractionated doses of radiation,” CancerResearch, vol. 67, no. 8, pp. 3845–3852, 2007.

[24] C. J. B. Maia, S. Socorro, F. Schmitt, and C. R. A. Santos,“Characterization of oligoadenylate synthetase-1 expressionin rat mammary gland and prostate: effects of 17β-estradiolon the regulation of OAS1g in both tissues,” Molecular andCellular Biochemistry, vol. 314, no. 1-2, pp. 113–121, 2008.


Research Article

Archaeosome Adjuvant Overcomes Tolerance toTumor-Associated Melanoma Antigens Inducing ProtectiveCD8+ T Cell Responses

Lakshmi Krishnan, Lise Deschatelets, Felicity C. Stark, Komal Gurnani, and G. Dennis Sprott

National Research Council of Canada, Institute for Biological Sciences, Ottawa, ON, Canada K1A 0R6

Correspondence should be addressed to Lakshmi Krishnan, [email protected]

Received 30 July 2010; Revised 15 December 2010; Accepted 23 December 2010

Academic Editor: Y. Yoshikai

Copyright © 2010 Lakshmi Krishnan et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Vesicles comprised of the ether glycerolipids of the archaeon Methanobrevibacter smithii (archaeosomes) are potent adjuvantsfor evoking CD8+ T cell responses. We therefore explored the ability of archaeosomes to overcome immunologic tolerance to self-antigens. Priming and boosting of mice with archaeosome-antigen evoked comparable CD8+ T cell response and tumor protectionto an alternate boosting strategy utilizing live bacterial vectors for antigen delivery. Vaccination with melanoma antigenic peptidesTRP181−189 and Gp10025−33 delivered in archaeosomes resulted in IFN-γ producing antigen-specific CD8+ T cells with strongcytolytic capability and protection against subcutaneous B16 melanoma. Targeting responses against multiple antigens affordedprolonged median survival against melanoma challenge. Entrapment of multiple peptides within the same vesicle or admixedformulations were both effective at evoking CD8+ T cells against each antigen. Melanoma-antigen archaeosome formulations alsoafforded therapeutic protection against established B16 tumors when combined with depletion of T-regulatory cells. Overall, wedemonstrate that archaeosome adjuvants constitute an effective choice for formulating cancer vaccines.

1. Introduction

Constant immunosurveillance against tumors both in exper-imental mouse models and cancer patients suggests thatimmunotherapy can be an effective way of controlling manyforms of cancer. For example, absence of the cytokine IFN-gamma or CD8+ T cells and perforin leads to aggressive oreven spontaneous tumor development [1–3]. The identifica-tion of many T cell-defined tumor antigens over the past twodecades has lead to the logical search for a reliable “widelyapplicable” cancer vaccine. Indeed the demonstration inboth mice and humans that T cells of defined antitumorspecificity can be generated and can eliminate cancers hasbeen very encouraging [4, 5]. Nevertheless, strategies havebeen limiting in their ability to sufficiently break toleranceand induce high avidity T cells against cancer self-anti-gens. Cancer vaccination as opposed to vaccination againstinfection poses a unique challenge in that the immuneactivation has to occur in the absence of “danger signal”recognition. Such a response is often weak, does not sustain

for long periods, and leads to tolerance and regulatory T cellinduction [6].

Vaccine adjuvants offer one potential solution to cir-cumventing the weak response to cancer antigens. Themolecular definition of danger-associated molecular pat-terns (DAMPs) has provided a rational choice of immunemodulators such as CpG and other TLR-ligand agonists[7, 8]. These have been explored as add-ons in vaccineformulations for boosting cancer vaccine efficacy. Anotherapproach includes heterologous prime-boost containingdifferent types of vaccine adjuvants that are often able toincrease the breadth and potency of the immune response[9, 10]. However, the short half-life of small moleculeimmunomodulators and potential toxicity of DAMPs oftenlimit their widespread use. Delivery systems includingliposomes, virus-like particles, microspheres, and ISCOMA-TRIX adjuvants provide alternative options for formulatingdiverse antigenic and adjuvant components and facilitat-ing specific immunity with minimal damaging side-effects[11, 12].


The polar membrane lipids of Archaea are characteristicof this Domain of life in having isoprenoid chains ofconstant length, with novel stereochemistry of ether bondlinkages to sn-2, 3 carbons of the glycerol backbone [13].We have developed a potent vaccine adjuvant delivery systemconstituted by these polar lipids, termed “Archaeosomes”[11]. Antigen may be entrapped within the hydrophilic coreof the vesicles or anchored in the membrane or linked tosurface-exposed groups akin to antigen-loading principlesfor liposomes. Our lead Archaeosome type is composedof the total natural mixture of polar lipids extracted fromthe methanogen, Methanobrevibacter smithii (Ms archaeo-somes). The superior features of Archaeosomes relative toconventional liposomes and other adjuvants are recruitmentand activation of dendritic cells in vivo, ability to directantigen cargo for MHC class I processing leading to potentinduction of CD8+ T cell response, and stability of archaeallipid cores facilitating profound immune memory [14]. Mostimportantly, these archaeosomes bypass TLR-2- and IL-12-mediated signaling for activating CD8+ T cells [15].

Thus, we have evaluated the ability of archaeosomes toeffectively break self-tolerance to native melanoma antigensand afford tumor protection. We show that archaeosomescan provide strong antitumor immunity to self-antigens andthat a prime-boost strategy with alternate adjuvants waswithout benefit.


2.1. Materials. Ovalbumin grade VI, 2-mercaptoethanol,RBC lysing solution, carboxyfluorescein, AEC chromogenkit, and PKH26 red fluorescent cell linker kit werefrom Sigma-Aldrich (Sigma-Aldrich Canada Ltd., Oakville,Ontario, Canada). Flow cytometry antibodies were pur-chased from BD Biosciences, RPMI 1640 medium andgentamicin from Invitrogen Life Technologies, FBS fromHyClone, G418 from Calbiochem, and murine recombinantIL-2 from ID Labs. Peptides H-2Kb restricted Ova257–264

(SIINFEKL) and HLA.A2/H-2Kb TRP-2180–188 (SVYDF-FVWL), CTL epitope from tyrosinase-related protein-2 [16],Gp10025–33 (KVPRNQDWL) from human melanoma anti-gen Gp100 [17] were synthesized in-house. Live recombinantMycobacterium bovis (BCG-OVA) and Listeria monocyto-genes (LM-OVA) expressing OVA257–264 were constructed asdescribed previously [18].

2.2. Preparation of Antigen-Loaded Archaeosomes. Peptide-loaded archaeosome formulations were prepared from thetotal polar lipids (TPLs) extract from Methanobrevibactersmithii (Ms) [19]. GP10025–33 or TRP-2181–189 was entrappedseparately, or a coentrapped preparation was preparedwhere both peptides were loaded simultaneously. For modelantigen studies the whole protein Ovalbumin (OVA) wasentrapped in Ms archaeosomes.

A dry lipid film containing 20 mg TPL was hydrated at35◦C in 0.5 mL of filter-sterilized Milli-Q water containing5 mg of OVA. In the case of peptides, the hydration wasdone in a high pH environment, and the proportion of

the peptide was diminished to avoid aggregate formationbetween the charged peptide and the negatively chargedpolar lipids. 100 mM Triethanolamine (TEA) pH 9 contain-ing 2 mg peptide was added to the dry lipids. Assessment ofarchaeosome integrity was done microscopically using phasecontrast at 2000X magnification. Archaeosomes were bathsonicated for 1–3 min to reduce the vesicles diameter to 100–200 nm and were annealed overnight at 4◦C for membranestabilization.

Nonentrapped antigen was removed by ultracentrifuga-tion at 207,000× g (raverage) for 2 h (Beckman centrifuge).The liposome pellet was washed three times with 8 mL ofpyrogen-free filter-sterilized Milli-Q water. The final pelletwas resuspended in 1.0 ml of water and filtered through0.45 μm filters.

2.3. Archaeosome Characterization. Gaussian, number-weighted size distributions were monitored with a Nicompparticle sizer Model 370, Santa Barbara, CA. All archaeosomevaccines used herein ranged between 80 and 118 nm averagediameter. Entrapment efficiency was determined based onthe dry weight of a known aliquot and quantification of theincorporated antigen. Ovalbumin content was evaluatedthrough SDS-PAGE gel electrophoresis and densitometryof bands revealed by Coomassie blue staining. Peptideamounts were assayed by RP-HPLC using a Zorbax C-18reverse-phase column (150× 4.6 mm) with a guard cartridgeinstalled in a DX-300 Dionex dual piston HPLC system(Sunnyvale, CA). The peptides were eluted at a flow rateof 1 mL/min using a gradient aqueous mobile phase from2% acetonitrile in 0.1% TFA to 70% acetonitrile in 0.085%TFA over 60 min and revealed by UV absorbance at a216 nm wavelength. Integration was done by a Dionex 4290integrator. Quantification was done using a calibrationcurve based on known amounts of each of the respectivepeptides.

2.4. Mice and Immunizations. Inbred, 6–8-week-old femaleC57BL/6J mice were obtained from the Jackson Laboratory(Bar Harbor, ME) and maintained in the Animal facilityof the Institute for Biological Sciences, NRC, in accordancewith guidelines from the Canadian Council on AnimalCare. Mice were injected subcutaneously (0.1 mL volume)at the base of the tail, with antigen-archaeosomes (Ms-antigen), antigen in PBS (no adjuvant), or live recombinantpreparations LM-OVA or BCG-OVA as per doses indicatedin figure legends. All final archaeosome preparations werein PBS prior to the immunizations. Antigen-archaeosomeimmunization scheme was based on peptide amounts of 1,10, 15, 20, or 30 μg in 0.1 mL. Each archaeosome formulationwas named according to the antigen entrapped as Ms-OVA,Ms-Gp100, and Ms-TRP. Archaeosome preparations withcoentrapped Gp100 and TRP peptides (both peptides withinthe same archaeosomes) were named Ms-Gp100-TRP co-entrap, whereas the ones named Admix were defined by thecombination prior to injection of two (Ms-Gp100 + Ms-TRPadmix) of the singly entrapped peptides. For comparisonpurposes, the admixed archaeosome immunization scenarioswere based on the same antigen amount as directed by


the coentrapped version. Refer to figure legends for loadingshown as μg antigen/mg dry weight of the different immu-nization cocktails.

2.5. Cell Lines. EL-4 (H-2b) was obtained from the Amer-ican Type Culture Collection (ATCC, Rockville, MD) andmaintained in RPMI 1640 medium (Life Technologies,Grand Island, NY) supplemented with 2-mercaptoethanol,8% FBS (HyClone, Logan, UT) and 10 μg/ml gentamicin(Life Technologies). B16 melanoma cells were cultured inRPMI plus 8% FBS. B16OVA cells, expressing the genefor OVA, were obtained from Dr. Edith Lord (Universityof Rochester, NY) and cultured in RPMI plus 8% FBS,additionally containing 400 μg/mL G418.

2.6. Assessment of Numbers of Antigen-Specific CD8+ T CellsIn Vivo. The activation of antigen-specific CD8+ T cells afterimmunization with Ms-OVA, LM-OVA, and BCG-OVA wastracked in vivo using the tetramer assay. Briefly, peripheralblood lymphocytes were incubated in 200 μL PBS plus 1%BSA (PBS-BSA) with anti-CD16/32 at 4◦C. After 10 min.,cells were stained with PE-tetramer and anti-CD8 for 30 min.at room temperature. Cells were washed with PBS, fixedin 0.5% formaldehyde, and acquired on BD BiosciencesFACS Canto analyzer. The tetramers used include H-2KbOVA257–264, H-2DbTRP-2180–188, and H-2KbGp10025–33,all purchased from Beckman Coulter.

2.7. CTL Assays. The antigen-specific cytolytic activity ofspleen cell effectors after recall stimulation with antigen wascarried out as described in detail previously [20]. Briefly,spleen cells were cultured with 0.01 μg/mL of the appropriateantigen (TRP2181–189 or Gp10025–33) for 5 days in vitro,and the ensuing effectors were used in a standard 51Cr-release CTL assay against nonspecific and antigen-specifictargets. EL-4 cells served as the nonspecific target, whereasthey were preincubated with 10 μg/ml of the CTL peptidefor 1 h to generate specific targets. The percent specificlysis at various effector:target ratios was calculated usingthe formula: [(cpm experimental-cpm spontaneous)/(cpmtotal-cpm spontaneous)]× 100.

In vivo cytolytic activity of antigen-specific CD8+ T cellswas enumerated according to the protocol of Barber et al.[21]. Donor spleen-cell suspensions from syngeneic micewere prepared and red blood cells lysed using trisbufferedammonium chloride (RBC lysing solution). Cells werestained with the dye PKH26 (4 μM) and split into twoaliquots. One aliquot was stained with low concentrationof CFSE (0.5 μM) and incubated in R8 medium. Thesecond aliquot was stained with 10X CFSE (5 μM) andincubated with the appropriate CTL peptide (10 μg/mL) inR8 medium. After 30 min. of incubation, the two aliquotswere mixed 1 : 1 and injected (20× 106/mouse) into pre-viously immunized recipient mice. PBS-injected recipientmice served as controls. At 24 h after the donor cell transfer,spleens were removed from recipients, single cell suspensionsprepared, and cells analyzed by flow cytometry. The in vivolysis percentage of peptide pulsed targets was enumeratedaccording to previously published equation [21].

2.8. Enumeration of IFN-γ Secreting Cells. Enumeration ofIFN-γ secreting cells was done by ELISPOT assay. Briefly,ELISPOT plates were coated with anti-IFN-γ antibody,blocked, and incubated with spleen cells in various numbers(in a final cell density of 5× 105/well using feeder cells)in the presence of IL-2 (0.1–1 ng/mL) and R8 media orthe appropriate CTL epitope peptide (5–25 μg/mL) for 48 hat 37◦C, 8% CO2. The plates were then incubated withthe biotinylated secondary antibody (37◦C, 2 h) followedby avidin-peroxidase conjugate (room temperature for 1 h).Spots were revealed using 3-amino-9-ethylcarbazole (AECchromogen kit).

2.9. Tumor Model. Mice were injected with 106 B16-OVA orB16 tumor cells (in PBS plus 0.5% normal mouse serum) inthe shaved lower dorsal region. From day 5 onwards, palpablesolid tumors were measured using digital calipers. Tumorsize, expressed in mm2, was obtained by multiplicationof diametrically perpendicular measurements. Mice wereeuthanized when the tumor sizes reached a maximum of 300mm2.

2.10. Statistical Analysis. Unpaired, Student’s t-test was usedto determine the statistical difference between two groupsof data, whereas one-way ANOVA was used to comparemultiple sets of data. Tumor survival curves were analyzedby log-rank test. The statistical package of the Graph Prismsoftware was used for statistical analyses of all data.

3. Results

3.1. Priming and Boosting with Archaeosomes Confers SuperiorQuantity of Antigen-Specific CD8+ T Cell Response. Wehave previously shown that Ms archaeosomes prime CD8+

T cell response to entrapped antigen. However, in manyvaccination regimens, particularly for breaking tolerance,priming and boosting with alternate adjuvants, has beensuggested to increase the magnitude and longevity ofantigen-specific T cell response. We therefore evaluatedwhether boosting with a live vector may improve responseto a primary Ms-OVA injection. Mice that received a singleinjection of BCG-OVA or LM-OVA evoked OVA-specificCD8+ T cells (based on OVA-tetramer binding) similar innumbers to those injected with particulate Ms-OVA onday 7 after immunization (Figure 1(a)). When Ms-OVAvaccination was followed by LM-OVA injection 30 days later,a clear increase in the number of antigen-specific CD8+ Tcells was seen on day 37. However, the magnitude of theresponse was similar to priming and boosting with Ms-OVA(Figures 1(a) and 1(b)). In contrast, a BCG-OVA boosterprovided very little enhancement of the antigen-specificresponse primed by Ms-OVA. The in vivo CTL responseto vaccination showed a similar trend (Figure 1(c)), withsingle dose vaccines yielding low level of specific killing,whereas boosting with Ms-OVA or LM-OVA yielded strongenhancement of CD8+ T cell cytolytic ability. Therefore,there was no benefit to using an alternate delivery vectorfor boosting.


Ms-OVA

(a)

(b) (c)

LM-OVABCG-OVA

CD

8C

D8

OV

A25

7−26

4te

tram

er+

cells

(%)

Sin

gle

inj.

Pri

me-

boos

t

Invi

voki

llin

g(%

)

1.4%

BC

G-O

VA

LM-O

VA

Ms-

OV

A

Ms-

OV

A/B

CG

-OV

A

Ms-

OV

A/L

M-O

VA

Ms-

OV

A/M

s-O

VA

1.3%1%

23%22%2.3%

Ms-OVA/Ms-OVAMs-OVA/LM-OVAMs-OVA/BCG-OVA

60

0

5040

(days)Single inj. Prime-boostBCG-OVALM-OVAMs-OVA

Ms-OVA/BCG-OVA

Ms-OVA/LM-OVAMs-OVA/Ms-OVA

10

20

30

OVA257−264 tetramer OVA257−264 tetramer OVA257−264 tetramer

OVA257−264 tetramer OVA257−264 tetramer OVA257−264 tetramer

In vivo CTL

0

25

50

75

100∗∗

Figure 1: CD8+ T cell response after heterologous prime-boost with archaeosomes and live vectors. Mice were vaccinated (subcutaneously)with Ms-OVA (20 μg OVA entrapped in Ms archaeosomes, 94.1 nm average size, 38 μg/mg loading), or 104 CFU of live LM-OVA or BCG-OVA. Prime-boost regimens involved the same Ms-OVA vaccine given on days 0 and 30, or heterologous boost on day 30 with the BCG-OVAor LM-OVA vector. The CD8+ T cell response was evaluated based on the percentage of OVA257–264 tetramer positive cells in the blood (a, b)and in vivo CTL response (c). (a) Representative scatter plots showing tetramer positive cells on day 7 after a single injection (top panel) orafter prime-boost on day 37 (bottom panel). The square gate indicates percentage of tetramer positive cells in the blood of immunized mice.(b, c), Mean± SD of 5 mice per group. ∗Response was statistically significant from single dose group by student’s t-test (P < .05).


3.2. Prime-Boost Vaccination Affords Long-Term Protectionagainst Melanoma Challenge. We next correlated the CD8+

T cell response evoked by various vaccination regimensto protection against subcutaneous melanoma challenge.Mice vaccinated with a single dose of Ms-OVA, LM-OVA,or BCG-OVA exhibited a median survival of 60, 60, and31 days, respectively, following subcutaneous melanomachallenge (Figure 2(a)). In contrast, prime-boost regimensthat involved priming with Ms-OVA followed by boostingwith Ms-OVA or LM-OVA afforded superior protection, with>90% of vaccinated mice being tumor-free for indefiniteperiods. Boosting with BCG-OVA resulted in a mediansurvival of 45 days, which was greater than a single injec-tion regimen (Figure 2(b)). Thus, prime-boost vaccinationafforded superior tumor protective responses, but alternatingvector delivery systems had no further benefit.

3.3. Archaeosome Vaccines Break Tolerance to Self-AntigenCargo. As priming and boosting with model antigensentrapped in archaeosomes yielded a strong CD8+ T cellresponse and tumor protection, we next evaluated theresponse to self-antigen cargo delivered in archaeosomes.Immunization with 10 to 30 μg of TRP-2181–189 peptide(H-2Kb/HLA A.2 CD8+ T cell epitope) entrapped inarchaeosomes evoked strong peptide-specific IFN-gammaproduction by CD8+ T cells as evaluated in an ELISPOT assay(Figure 3(a)). This correlated to a strong antigen-specificCTL response evoked in immunized mice both in vitroand in vivo (Figures 3(b) and 3(c)). Above all, immunizedanimals showed protection against a melanoma challenge(Figure 3(d)). Entrapment of Gp10025–33 (H-2Kb/HLA A.2CD8+ T cell epitope) peptide in archaeosomes also resultedin IFN-gamma production by CD8+ T cells (Figure 4(a))and peptide-specific in vitro and in vivo CTL responses(Figures 4(b) and 4(c)). Additionally, the CD8+ T cellsfrom vaccinated mice exhibited killing of B16 targets (notpulsed with peptide) in vitro in a standard chromium releasekilling assay (Figure 4(b)). It is likely that the weaker killingresponse against B16 targets relative to EL-4 peptide targetmay be attributable to lower endogenous expression ofMHC-peptides complexes in B16 cells. Importantly, Gp100-archaeosome vaccination protected mice against subcu-taneous melanoma challenge (Figure 4(d)). Although theresponses to Gp100 were less dramatic than with TRP-2,archaeosomes were able in general to evoke functional CD8+

T cell responses to native melanoma antigens.

3.4. Archaeosome-Dual Antigen Formulations Evoke CTLResponses to Each Individual Antigenic Component. Whilearchaeosomes could evoke CTL responses to native melan-oma antigens, the response to the individual antigenic com-ponent provided vaccinated animals with only a relativelyshort-lived protection to tumor challenge. We thereforecoentrapped two melanoma CTL epitopes, TRP-2181–189 andGp10025–33 within the same archaeosome vesicles. Follow-ing vaccination with a coentrapped antigen-archaeosomeformulation, spleen cells of vaccinated mice exhibited CTLresponses specific to both TRP-2 and Gp100 epitopes(Figure 5(a)). Nevertheless, the TRP-2-specific CTL response

was weaker than that of Gp100, and this may be attributedto differential loading of the two respective peptides inthe coentrapped formulation. A 25 μg injection of peptidescorresponded to 20 μg Gp100 and 5 μg TRP in 1.56 mgarchaeosomes. However, vaccination with Ms-TRP and Ms-Gp100 archaeosomes admixed to achieve the equivalent20 μg Gp100 and 5 μg TRP as in the coentrapped formula-tion, but this time entrapped in only 0.5 mg archaeosomes,resulted in a strong and comparable CTL activity beingevoked against both antigens (Figure 5(b)). Consistent withthe intensity of CTL responses, the coentrapped formulationafforded a median survival of only 29.5 days to tumorchallenge, whereas immunization with the admixed Ms-TRP and Ms-Gp100 archaeosomes proved more effective atproviding tumor protection (median survival of 49 days)(Figure 5(c)). Nevertheless, there was no statistical differencebetween the coentrapped and admixed vaccination group,and the median survival may be attributable to small samplesize of the study. However, both groups demonstrated statis-tically significant difference from the control, nonvaccinatedgroup. Therefore, admixed formulations simply provide aconvenient means of controlling a more optimized dose andloading.

3.5. Targeting Dual Melanoma Antigens with an AdmixedArchaeosome Formulation Affords Long-Term Tumor Protec-tion. In the above experiments, the admixed formulation ofarchaeosomes contained suboptimal amounts of melanomaantigens in order to serve as appropriate comparison to thecoentrapped formulation. In order to test the full potentialof targeting multiple melanoma antigens with archaeosomeadjuvants, vaccination was carried out with a mixture ofarchaeosome melanoma peptide formulations such that eachantigen was provided in an equivalent dose of 30 μg, on day0 and 21. Firstly, IFN-gamma production by CD8+ T cells, asevaluated in an ELISPOT assay on day 28 after vaccination,indicated a response against both TRP-2 and Gp100 peptidestimulations using the admixed formulation (Figure 6(a)and 6(b)). However, the antigen-specific response to Gp100peptide was weaker when administered in an admixed for-mulation relative to Gp100-archaeosome by itself. Antigen-specific tetramer was also detectable in the blood on day 28after priming against both peptides with the admixed for-mulation (Figure 6(c)). Here again, the endogenous responseto Ms-Gp100 vaccine was stronger when administeredby itself than in an admixed formulation. The CD8+ Tcells evoked after admixed immunization effectively killedtargets expressing TRP or Gp100 (Figure 6(d)), althoughthe killing against TRP-expressing targets was stronger thanthe killing against Gp100 targets. Overall it appears thatwhen coadministered, the TRP epitope may dominate inresponse over Gp100. Finally, we addressed the ability of theadmixed vaccine to evoke tumor protection. In all previousexperiments, tumor challenge was conducted at 6 weeks pos-timmunization, possibly during the declining phase of theT cell response. Therefore, we compared protection againsttumors given at either 4 or 6 weeks postvaccination. Ms-Admixed formulation afforded significant protection (P <.01) when tumor challenge was carried out at either 4 or 6


0

30

60

Surv

ival

(%)

90

0 25 50

∗∗

∗∗∗

(days)

NaiveMs-OVA

LM-OVABCG-OVA

(a) Single Inj.

0

30

60

Surv

ival

(%)

90

0 25 50

∗∗∗

∗∗∗

(days)

Naive

Ms-OVA/BCG-OVA

Ms-OVA/LM-OVAMs-OVA/Ms-OVA

(b) Prime-boost.

Figure 2: Tumor protection following heterologous prime-boost vaccination. C57BL/6J mice were vaccinated subcutaneously with a singledose of Ms-OVA (20 μg OVA, 38 μg/mg lipid loading, 94.1 nm average archaeosome size), LM-OVA (104 CFU), or BCG-OVA (104 CFU) orwith a prime-boost regimen as indicated. Mice were challenged with subcutaneous B16-OVA tumors 4 weeks post vaccination. Survival plotsare based on euthanizing animals upon reaching a maximum tumor size of 300 mm2 (n = 5/group). Survival curves for vaccinated groupswere significantly different from naıve group by log-rank test (∗∗P < .01, ∗∗∗P < .001).

weeks postvaccination (Figure 6(e)). Furthermore, there wasno statistically significant difference in the protection seenat early (4 week) versus late (6 weeks) postvaccination (P =.4) suggesting ability of archaeosomes to afford memoryresponses to self-antigens. Comparing the aggregate datafrom several experiments, the median survival for TRP-2-or Gp100-archaeosomes single-peptide vaccine administeredpreventatively was 22 days relative to <17 days for naıvenonvaccinated mice. In contrast, targeting responses toboth peptides using an admixed vaccine markedly improvedmedian survival to 35 days.

3.6. Therapeutic Protection against Established B16 Tumorsby Peptide-Archaeosome Vaccine. Induction of prophylacticprotection against cancer vaccines is often relatively easy toachieve, whereas breaking tolerance against an establishedtumor in a therapeutic setting can be challenging. Indeed,the admixed formulation of TRP- and Gp100-archaeosomesafforded only marginal protection (median survival of 24days relative to median survival of 20 days for nonvaccinatedgroup) when administered therapeutically following tumorchallenge (Figure 7). However, when combined with priordepletion of T-regulatory cells using anti-CD25 antibody asignificant decrease (P < .01) in tumor size (Figure 7(a))and increased median survival (P < .01) were observed(Figure 7(b)). Thus, archaeosome vaccines appear to holdpromise for tumor vaccination when combined with othervaccination strategies that target tumor evasion.

4. Discussion

Archaeosomes are effective self-adjuvanting delivery systemsthat target processing of antigenic cargo for MHC class

I presentation leading to potent long-term CD8+ T cellresponses. In previous studies using model antigens wedemonstrated that following immunization of mice withOvalbumin (OVA)-Archaeosomes,∼3.5% of all CD8+ T cellsin the spleen were OVA-specific by day 7, and boostingon day 21 resulted in expansion to ∼20% on day 28 [14].Furthermore, a prolonged memory response ensued thatwas complemented with a strong functional cytolytic abilityof CD8+ T cells and protection against OVA-expressingtumors in prophylactic and therapeutic settings [15]. Thus,we pursued the ability of archaeosomes to evoke adaptiveimmune responses to entrapped tumor-associated antigens.The generation of high frequencies of functional antigen-specific CD8+ T cells against two tumor-associated antigens,as assessed by IFN-γ secretion, enumeration of tetramer-binding antigen-specific CD8+ T cells, and in vitro andin vivo cytotoxic assays prove archaeosomes as a highlyeffective adjuvant for breaking tolerance to tumor self-antigens.

Microparticulate carrier systems such as archaeosomespreferentially accumulate in APCs and thus can effectivelytarget antigens for presentation onto MHC class I. However,secondary costimulatory signals are required for breakingtolerance to self-antigens which are often lacking in thecontext of tumors due to lack of danger-associated molecularpatterns. Advantageously, archaeosomes exhibit the uniqueability to induce dendritic cell maturation and increasecostimulation [22]. Furthermore, archaeosomes evoke CD8+

T cell response in an IL-12-independent manner [15].Thus, archaeosome vesicles uniquely bring together manybeneficial features for tumor antigen presentation: antigentargeting to APC and costimulation in the absence of overtinflammation leading to effective adaptive immunity in a safemanner.


0 07

(days)

28

10

40Ms-TRP

30

(a) (c)

(b) (d)

20

1 μg 10 μg

Ms-TRP (in vivo)

100

30 μg

IFN

-γse

cret

ing

cell

nu

mbe

r

Invi

voki

llin

g(%

)

200

Surv

ival

(%)

Spec

ific

lysi

s(%

)

300

400

101

NaivePBS-TRPMs-TRP

NaivePBS-TRPMs-TRP

E : T ratio

102101102

0

30

60

90

0

0

∗∗∗

20 30

(days)

40

30

60

EL-4 TRP EL-490

In vitro stimulation0 μg/mL-TRP peptide

5 μg/mL-TRP peptide

Figure 3: CD8+ T cell response and tumor protection induced by Ms-TRP vaccine. C57BL/6J mice were immunized in (a) with 1, 10, or 30 μgTRP peptide entrapped in Ms archaeosomes on days 0 and 21. At 5 weeks, representative mice (n = 2 per group) were euthanized, the spleencells were stimulated with IL-2 (0.1 ng/mL) and peptide (5 μg/mL) for 48 h, and the frequency of IFN-gamma secreting cells was enumeratedby ELISPOT. Mean± SD (n = 3) of IFN-gamma secreting cells per 106 spleen cells is indicated. Spleen cells from mice immunized with10 μg TRP-peptide per injection were also cultured for 5 days with antigen to generate CTL effectors. The ability of effectors to kill peptidespecific (EL-4-TRP) versus nonspecific (EL-4) targets was evaluated in an in-vitro CTL assay (b). Mean killing± SD of 2 mice per group atdifferent effector : target ratio is indicated for Naıve, PBS-TRP, and Ms-TRP vaccinated mice (b). In another group of representative micevaccinated with 20 μg Ms-TRP, in vivo CTL response was evaluated on day 7 and day 28. Mean± SD of n = 4 mice per group is indicated(c). Finally, groups of naıve (n = 12), Ms-TRP vaccinated (n = 12), and TRP-PBS (n = 4) vaccinated mice were challenged with B16 tumorsat 6 weeks. Survival was monitored based on a maximum tumor size of 300 mm2. Tumor survival data are presented as an aggregate from 3different experiments conducted, and TRP dose was 15–30 μg/injection. Loading was 19 μg peptide/mg lipid, and average archaeosome sizewas 99 nm. Survival with Ms-TRP is significantly different (P < .001) by log-rank test relative to naıve group.

Priming and boosting of an immune response has beentraditionally recognized as an efficient way of maintaininglong-term immunity. Indeed most vaccines against infec-tious diseases are often given multiple times and since

priming and boosting occurs with the same vaccine, theyare considered homologous prime-boost regimens. Het-erologous prime-boosting with unmatched vaccine deliverysystems or adjuvants but the same antigen was first reported


0 07

(days)

28

10

40Ms-Gp100

30

20

1 μg 10 μg

Ms-Gp100 (in vivo)

25

30 μg

IFN

-γse

cret

ing

cell

nu

mbe

r

Invi

voki

llin

g(%

)

50

Surv

ival

(%)

Spec

ific

lysi

s(%

)

75

100

101

NaiveMs-Gp100

Naive

Ms-Gp100PBS-Gp100

E : T ratio

102101 101 102102

0

60

30

90

0

0

∗∗

25 50

(days)

25

50

EL-4 B16EL-4-Gp100

75

In vitro stimulation5 μg-Gp100 peptide

25 μg-Gp100 peptide

(a) (c)

(b) (d)

Figure 4: CD8+ T cell response and tumor protection induced by Ms-Gp100 vaccine. Mice were immunized with 1, 10, or 30 μg Gp100peptide entrapped in Ms archaeosomes at days 0 and 21. At 5 weeks, representative mice (n = 2 per group) were euthanized, the spleen cellswere stimulated with IL-2 (0.1 ng/mL) and peptide (25 μg/ml) for 48 h and the frequency of IFN-gamma secreting cells was enumerated byELISPOT. Mean± SD (n = 3) of IFN-gamma secreting cells per 106 spleen cells is indicated. (a) At 5 weeks, in vitro CTL assay was alsocarried out on spleen cells of representative mice (n = 2 per group) immunized with 10 μg Ms-Gp100 (b). Mean Killing± SD of 2 mice pergroup at different effector : target ratio is indicated (b). In another group of representative mice vaccinated with 20 μg Ms-Gp100, in vivoCTL response was evaluated on day 7 and day 28. Mean± SD of n = 4 mice per group is indicated (c). Finally, response to subcutaneous B16tumor challenge was evaluated at 6 weeks in groups of naıve (n = 12), Ms-Gp100 vaccinated (n = 12), and PBS-Gp100 (n = 4) vaccinatedmice. Survival was monitored based on a maximum tumor size of 300 mm2. Tumor survival data are presented as an aggregate from 3different experiments conducted, and Gp100 peptide vaccination dose ranged from 25 to 30 μg/injection. Loading was 112 μg peptide/mgarchaeosomes and average size 94.3 nm. Survival for Ms-Gp100 group is significantly different (P < .05) from naıve mice by log-rank test.

in the early 1990s to be an effective strategy for evoking abalanced humoral and cell-mediated immunity [23]. Earlystudies focused on live viral vectors expressing antigensfollowed by a protein or peptide boost. With the adventof DNA vaccines that were weakly immunogenic, onceagain priming with DNA vaccine followed by boostingwith a protein and peptide antigen proved to be effec-tive strategy [24]. Since then, prime-boost regimens have

been evaluated for several new generation vaccines againstHIV, malaria, and tuberculosis. Heterologous prime-boostapproaches can improve vaccine immunity by influencingthe immunogenicity of antigens, allowing dose sparing,diversifying the quality of immunity, and circumventingthe negative effects of neutralizing antibodies against thepriming vector [9, 10, 24–26]. A heterologous prime-boost protocol involving different viral vectors expressing


0

25

Spec

ific

lysi

s(%

)

50

75TRP

101 102 101 102

Gp100

E : T ratioTargets

EL-4EL-4 peptide

(a) Ms-TRP-Gp100 (co-entrap)

0

25

Spec

ific

lysi

s(%

)

50

75TRP

101 102 101 102

Gp100

E : T ratioTargets

EL-4EL-4 peptide

(b) Ms-TRP + Ms-Gp100 (admix)

0

30

60

surv

ival

(%)

90

0 25 50

∗∗

∗

(days)

NaiveMs-TRP-Gp100 (co-entrap)Ms-TRP + Ms-Gp100 (admix)

(c)

Figure 5: CD8+ T cell response and tumor protection induced by coentrapped melanoma peptide-archaeosome vaccine. Mice werevaccinated (days 0 and 21) with 25 μg of peptides (coentrapped 5 μg TRP and 20 μg Gp100) or an equivalent admixed formulation ofMs-TRP and Ms-Gp100. In vitro CTL response of spleen effectors from representative mice (n = 2) was evaluated at 5 weeks (a, b) onTRP and Gp100 specific targets. Mean± SD of triplicate cultures of effectors: targets at various ratios are indicated. At 6 weeks mice werechallenged subcutaneously with B16 melanoma, and survival (n = 4/group) was evaluated based on a maximum tumor size of 300 mm2 (c).Loadings were 13 μg Gp100 and 3 μg TRP/mg archaeosomes for the coentrapped vaccine used in (a, c), and 60 μg Gp100/mg lipid and 30 μgTRP/mg archaeosomes for the admixed used in (b, c). Archaeosome size ranged from 104 to 110 nm. Survival for the vaccinated groups wassignificantly different compared to naıve animals by log-rank test (∗P < .05; ∗∗P < .01).

the same melanoma-polypeptide induced 100 times greaterfrequencies of vaccine-induced CD8+ T cells [27]. We thusrationalized that if a heterologous prime-boost regimes usingmodel antigen-archaeosome vaccines were beneficial, thisapproach may be beneficial for cancer antigen delivery usingarchaeosomes. Our results indicate that while priming andboosting (2 injections) afforded superior tumor protectionrelative to a single injection regimen, there was no benefit toboosting with an alternative delivery system. The inefficiencyof BCG-OVA to boost a good immune response may berelated to its slow growth and low antigen expression. In

contrast, 2 injections of Ms-OVA archaeosomes affordedsuperior tumor immunity comparable in efficacy to boostingwith LM-OVA. Archaeosomes comprised of polar lipids ofM. smithii do not usually evoke antilipid responses andhence demonstrate an advantage to be repeatedly utilized invaccination regimens.

Use of tumor antigenic peptides provides a quick, simple,and inexpensive manner of targeting induction of antitumorresponses. T cell immunity against multiple antigenic deter-minants may be evoked. Peptides can be easily formulatedto be tested in various settings, including altered peptide


(a) (c)

(b) (d) (e)

Ms-admixedMs-Gp100Ms-TRPNaive

TR

P2-

tetr

amer

Gp1

00-t

etra

mer

IFN

-γse

cret

ing

cell

nu

mbe

rIF

N-γ

secr

etin

gce

lln

um

ber

Spec

ific

killi

ng

(%)

1.4%

EL-

4

EL-

4-T

RP

EL-

4-G

p100

1.7%±0.63%

0.32%±0.28%

1.29%±1.47%

0.18%±0.08%

0.72%±0.09%

0% 0.03%±0.04%

8.76%±4.73%

CD8

0

200

400

600

0

50

100

150

Naive

In vitro/TRP peptide

0 μg/mL

5 μg/mL

Naive Ms-Gp100 Ms-admixed

In vitro/Gp100 peptide

0 μg/mL25 μg/mL

Ms-TRP Ms-admixed

0

25

50

75

0 15 30

(days)

Naive

Surv

ival

(%)

Ms-admixed (4 weeks)

Ms-admixed (6 weeks)

0

30

50

90∗

∗

Figure 6: Optimized induction of CD8+ T cell response to dual melanoma antigen vaccine. Mice were vaccinated (days 0 and 21) with 30 μgof individual peptide-archaeosome vaccines or in an admixed formulation containing 30 μg of each peptide. At 4 weeks, representative mice(n = 3 per group) were euthanized, the spleen cells were stimulated with IL-2 (0.1 ng/mL) and each peptide (5–25 μg/mL) for 48 h, and thefrequency of IFN-gamma secreting cells was enumerated by ELISPOT. Mean± SD (n = 3) of IFN-gamma secreting cells per 106 spleen cellsis indicated for TRP-peptide (a) and Gp100 peptide (b) stimulation. The percentage of tetramer-specific CD8+ T cells was enumerated in theblood on day 28. Representative plots for the various groups are shown (c), and the Mean± SD (n = 3) of the response is indicated withineach panel. In vitro CTL response in representative mice (n = 3/group) vaccinated with the admixed formulation was evaluated on day 28against EL-4, EL-4-TRP, and EL-4-Gp100 targets (d). Mean± SD at 100 : 1 effector : target ratio is indicated. At 4 or 6 weeks postvaccination,mice (n = 5 per group) were challenged with B16 melanoma (e). Survival curves for the vaccinated groups were significantly different fromnaıve by log-rank test (P < .05). Archaeosome loadings were 40 μg Gp100/mg and 20 μg TRP/mg lipid. Archaeosome size ranged from 110to 117 nm.

ligands, lipidated peptides, combination with heterologoushelper peptides, and comparison of various adjuvants. Overthe past decade, CTL epitopes of several melanoma antigenshave been identified, and epitopes of melanoma antigensGp100 and TRP-2 are widely targeted in clinical trials [4, 28–32]. Thus, we chose the use of these peptides for proof-of-concept studies to demonstrate the ability of archaeosomesto overcome self-tolerance. The overall negative chargeof the total polar lipids mixture from Methanobrevibacter

smithii with its high phospholipids content together withthe charge/solubility of the peptides being used is majorfactors to consider during vaccine formulation. Each peptidetype was solubilized and made into an appropriate overallcharge and which had to be compatible with the hydrationof the dry lipid film required for archaeosome formation.GP100 and OVA peptides were all more or less hydrophilicin character, whereas TRP-2 was more hydrophobic andrequired the addition of some propanol. TRP-2 peptide


0

100

Tum

our

size

(mm

2)

200

300

0 10 20 30

∗∗

(days)

NaivePBS-admixed

Ms-admixedMS-admixed + αCD25-Ab

(a)

0

30

Surv

ival

(%)

60

90

0 10 20 30

∗∗

Tumour

NaivePBS-admixed

Ms-admixedMS-admixed + αCD25-Ab

(b)

Figure 7: Therapeutic tumor protection by the dual melanoma antigen-archaeosome vaccine. Mice were injected with 105 B16 melanomacells subcutaneously on day 0. Vaccination with 30 μg of each peptide in PBS (PBS-admixed) or as mixture of Ms-Gp100 and Ms-TRP (Ms-admixed) was carried out on days 1 and 21 posttumor challenge. One group of mice received the anti-CD25 antibody injection (100 μg),intraperitoneally on day 1 posttumor challenge. Mean tumor size± SD (n = 5/group) over time is indicated for all groups (a). Animalsthat received the Ms-admixed vaccine and anti-CD25 antibody showed significantly slower tumor progression over time based on one-wayANOVA Bonferronis Post-test compared to the naıve group (P < .01). Tumor survival (b) is based on animals reaching a maximum tumorsize of 300 mm2. Survival for the Ms-Admixed plus anti-CD25 antibody group was also significantly longer (P < .01) relative to the naıvegroup (n = 5 mice/group) based on Log rank test. The admixed group contained 2.2 mg of lipid and 60 μg of peptide (30 μg of each peptide).

is acidic whereas GP100 peptide is basic. To avoid theaggregation of archaeosomes, a basic pH environment wasfavoured, and the lipid-peptide ratio was kept around 10 : 1w/w. In this study, results are showing equivalent immuneresponse using the admixed approach. The formulation ofmultiple peptide-archaeosome by admixing is appealing,since it requires optimization for only a single peptide ata time and may facilitate consistency of formulation andreproducibility.

Vaccination with two melanoma antigen-archaeosomesafforded superior tumor protection in comparison to sin-gle antigen-archaeosome vaccines indicating that targetingresponses to multiple antigens is an efficient way of break-ing tolerance. Moreover coentrapment and/or admixtureof individual peptide encapsulated archaeosomes affordedtumor protection. As archaeosomes are efficiently phago-cytosed by antigen presenting cells (APC), coentrappedantigens within the same vesicle may be targeted forrelease within the same APC. Once released in the intra-cellular milieu, peptides may compete for binding to thetransporter associated with antigen processing (TAP), anddifferential binding of melanoma peptides to TAP has beenreported [33, 34]. The differential CTL response to thetwo peptides following coentrapment may be reflectiveof their differential binding to TAP or just difference

in the amount of each peptide that was coentrapped.Moreover, peptides may also compete for binding to thesame MHC molecules, and epitopes with higher avidity forthe MHC-I complex may outwit weakly binding peptides.In contrast, when peptides are individually entrapped andused in an admixed formulation, each antigen is targetedindividually for presentation by the APC. Furthermore,optimized amounts of each antigen may be included inthe vaccine. Our results demonstrate that admixed peptide-archaeosome formulations induced CTL responses againstboth epitopes. It has been previously shown that recognitionof melanoma CTL epitopes with different affinities canbe achieved following multiple peptide formulation [35].Nevertheless, even with the admixed vaccine, responses tothe TRP-2 peptide was stronger based on frequency ofIFN-gamma secreting cells, tetramer+ CD8+ T cell number,and CTL response. Indeed, it has been suggested thatwith multiple peptide vaccinations, competition betweenCTLs can narrow the repertoire of the immune responseevoked [27]. Despite such a possibility, multiple peptideimmunization often provides a cumulatively stronger over-all CD8+ CTL response [36]. Indeed, the dual-peptidearchaeosome vaccine afforded longer-term tumor protec-tion relative to the single-peptide archaeosome formula-tion.


A number of particulate delivery systems have been eval-uated for cancer vaccine delivery. Conventional liposomescarrying cytokines such as IL-2 and IFN-gamma were testedfor their efficacy to deliver cancer antigens but showedlimited promise in breaking tolerance [37, 38]. Cationicliposomes posed the ease for efficient association with diversenegatively charged antigens including DNA [39] but theirhigh positive charge and large size lead to quick clearancefrom the blood and to tissue toxicity. More recently, lipidparticle-based nanovesicles with neutral charge were utilizedto deliver negatively charged CpG oligonucleotides in asafe and consistent manner [40]. A novel proteoliposomalvaccine prepared from the cell membrane proteins oflymphoma cells was also shown to be more efficient ininducing antitumor responses than conventional liposomes[41]. ISCOM vaccines that combine TLR9 agonists havebeen reported to break tolerance in an orthotopic model ofpancreatic carcinoma [42].

In many cases combination immunotherapy includesnovel vaccination regimens to overcome immunologic tol-erance. A synthetic TRP2180–188 peptide vaccine was recentlycombined with several toll-like receptor agonists and an anti-CD40 antibody to induce potent CD8+ T cells with therapeu-tic tumor efficacy against established B16 melanoma [43].Other strategies to counteract tumor evasion and promotelong-term tumor recession have included blocking T regu-latory cells and programmed death-ligand interactions, orco-use of adoptive T cell transfer [44, 45]. We also observedthat while both Gp100- and TRP-2-archaeosome vaccinesevoked antigen-specific CD8+ T cell response indicatingtheir ability to break tolerance, tumor protection againstestablished tumors required depletion of T regulatory cells.Indeed, it is now well established that T regulatory cellscan hinder the protective efficacy of tumor-specific CD8+

T cells. An anti-CD25 monoclonal antibody, daclizumab,has been effectively trialed for the depletion of T-regulatorycells and improved efficacy of a coadministered peptide-vaccine against human breast cancer [46]. Our studiesreiterate that induction of functional CD8+ T cells aloneis an insufficient marker for predicting tumor protectionand immunotherapy may require additional strategies tocounteract tumor evasion. Thus, archaeosomes should befurther evaluated as a promising adjuvant delivery system forimmunotherapeutic vaccination regimens.

Acknowledgments

The authors acknowledge Mr. J.-R. Barbier and Dr. GordonWillick for peptide synthesis. This study was conducted inpart with the support of the Ontario Institute for CancerResearch through funding provided by the Government ofOntario, Canada.

References

[1] H. Winter, H. M. Hu, K. McClain, W. J. Urba, and B. A.Fox, “Immunotherapy of melanoma: a dichotomy in therequirement for IFN-γ in vaccine-induced antitumor immu-nity versus adoptive immunotherapy,” Journal of Immunology,vol. 166, no. 12, pp. 7370–7380, 2001.

[2] M. F. Van Den Broek, D. Kagi, F. Ossendorp et al., “Decreasedtumor surveillance in perforin-deficient mice,” Journal ofExperimental Medicine, vol. 184, no. 5, pp. 1781–1790, 1996.

[3] M. J. Smyth, D. I. Godfrey, and J. A. Trapani, “A fresh lookat tumor immunosurveillance and immunotherapy,” NatureImmunology, vol. 2, no. 4, pp. 293–299, 2001.

[4] S. E. Finkelstein, D. M. Heimann, C. A. Klebanoff et al.,“Bedside to bench and back again: how animal models areguiding the development of new immunotherapies for cancer,”Journal of Leukocyte Biology, vol. 76, no. 2, pp. 333–337, 2004.

[5] C. S. Hinrichs, L. Gattinoni, and N. P. Restifo, “ProgrammingCD8+ T cells for effective immunotherapy,” Current Opinionin Immunology, vol. 18, no. 3, pp. 363–370, 2006.

[6] G. Lizee, L. G. Radvanyi, W. W. Overwijk, and P. Hwu,“Improving antitumor immune responses by circumventingimmunoregulatory cells and mechanisms,” Clinical CancerResearch, vol. 12, no. 16, pp. 4794–4803, 2006.

[7] A. Pashine, N. M. Valiante, and J. B. Ulmer, “Targeting theinnate immune response with improved vaccine adjuvants,”Nature Medicine, vol. 11, no. 4, pp. S63–S68, 2005.

[8] E. De Gregorio, U. D’Oro, and A. Wack, “Immunologyof TLR-independent vaccine adjuvants,” Current Opinion inImmunology, vol. 21, no. 3, pp. 339–345, 2009.

[9] V. P. Badovinac, K. A. N. Messingham, A. Jabbari, J. S.Haring, and J. T. Harty, “Accelerated CD8+ T-cell memory andprime-boost response after dendritic-cell vaccination,” NatureMedicine, vol. 11, no. 7, pp. 748–756, 2005.

[10] I. A. Ramshaw and A. J. Ramsay, “The prime-boost strategy:exciting prospects for improved vaccination,” ImmunologyToday, vol. 21, no. 4, pp. 163–165, 2000.

[11] L. Krishnan and G. D. Sprott, “Archaeosome adjuvants:immunological capabilities and mechanism(s) of action,”Vaccine, vol. 26, no. 17, pp. 2043–2055, 2008.

[12] G. F. A. Kersten and D. J. A. Crommelin, “Liposomes andISCOMs,” Vaccine, vol. 21, no. 9-10, pp. 915–920, 2003.

[13] M. Kates, “Archaebacterial lipids: structure, biosynthesis andfunction,” Biochemical Society Symposia, vol. 58, pp. 51–72,1992.

[14] L. Krishnan, K. Gurnani, C. J. Dicaire et al., “Rapid clonalexpansion and prolonged maintenance of memory CD8+

T cells of the effector (CD44highCD62Llow) and central(CD44highCD62Lhigh) phenotype by an archaeosome adjuvantindependent of TLR2,” Journal of Immunology, vol. 178, no. 4,pp. 2396–2406, 2007.

[15] L. Krishnan, S. Sad, G. B. Patel, and G. D. Sprott, “Archaeo-somes induce enhanced cytotoxic T lymphocyte responses toentrapped soluble protein in the absence of interleukin 12 andprotect against tumor challenge,” Cancer Research, vol. 63, no.10, pp. 2526–2534, 2003.

[16] M. W. J. Schreurs, A. A. O. Eggert, C. J. A. Punt, C. G. Figdor,and G. J. Adema, “Dendritic cell-based vaccines: from mousemodels to clinical cancer immunotherapy,” Critical Reviews inOncogenesis, vol. 11, no. 1, pp. 1–17, 2000.

[17] W. W. Overwijk, A. Tsung, K. R. Irvine et al., “gp100/pmel17 is a murine tumor rejection antigen: induction of “self”-reactive, tumoricidal T cells using high-affinity, altered peptideligand,” Journal of Experimental Medicine, vol. 188, no. 2, pp.277–286, 1998.

[18] R. Dudani, Y. Chapdelaine, H. Van Faassen et al., “Preexistinginflammation due to Mycobacterium bovis BCG infectiondifferentially modulates T-cell priming against a replicating ornonreplicating immunogen,” Infection and Immunity, vol. 70,no. 4, pp. 1957–1964, 2002.


[19] G. D. Sprott, J. R. Brisson, C. J. Dicaire et al., “A structuralcomparison of the total polar lipids from the human archaeaMethanobrevibacter smithii and Methanosphaera stadtmanaeand its relevance to the adjuvant activities of their liposomes,”Biochimica et Biophysica Acta, vol. 1440, no. 2-3, pp. 275–288,1999.

[20] L. Krishnan, S. Sad, G. B. Patel, and G. D. Sprott, “Archaeo-somes induce long-term CD8+ cytotoxic T cell response toentrapped soluble protein by the exogenous cytosolic pathway,in the absence of CD4+ T cell help,” Journal of Immunology,vol. 165, no. 9, pp. 5177–5185, 2000.

[21] D. L. Barber, E. J. Wherry, and R. Ahmed, “Cutting edge:rapid in vivo killing by memory CD8+ T cells,” Journal ofImmunology, vol. 171, no. 1, pp. 27–31, 2003.

[22] L. Krishnan, S. Sad, G. B. Patel, and G. D. Sprott, “Thepotent adjuvant activity of archaeosomes correlates to therecruitment and activation of macrophages and dendritic cellsin vivo,” Journal of Immunology, vol. 166, no. 3, pp. 1885–1893,2001.

[23] S. L. Hu, K. Abrams, G. N. Barber et al., “Protection ofmacaques against SIV infection by subunit vaccines of SIVenvelope glycoprotein gp160,” Science, vol. 255, no. 5043, pp.456–459, 1992.

[24] S. Lu, “Heterologous prime-boost vaccination,” Current Opin-ion in Immunology, vol. 21, pp. 346–351, 2009.

[25] M. Vaine, S. Wang, A. Hackett, J. Arthos, and S. Lu,“Antibody responses elicited through homologous or heterol-ogous prime-boost DNA and protein vaccinations differ infunctional activity and avidity,” Vaccine, vol. 28, no. 17, pp.2999–3007, 2010.

[26] T. I. Naslund, C. Uyttenhove, E. K. L. Nordstrom et al.,“Comparative prime-boost vaccinations using semliki forestvirus, adenovirus, and ALVAC vectors demonstrate differencesin the generation of a protective central memory CTL responseagainst the P815 tumor,” Journal of Immunology, vol. 178, no.11, pp. 6761–6769, 2007.

[27] M. J. Palmowski, E. M. L. Choi, I. F. Hermans et al.,“Competition between CTL narrows the immune responseinduced by prime-boost vaccination protocols,” Journal ofImmunology, vol. 168, no. 9, pp. 4391–4398, 2002.

[28] G. Parmiani, C. Castelli, P. Dalerba et al., “Cancerimmunotherapy with peptide-based vaccines: what havewe achieved? Where are we going?” Journal of the NationalCancer Institute, vol. 94, no. 11, pp. 805–818, 2002.

[29] G. Q. Phan, C. E. Touloukian, J. C. Yang et al., “Immunizationof patients with metastatic melanoma using both class I-and class II-restricted peptides from melanoma-associatedantigens,” Journal of Immunotherapy, vol. 26, no. 4, pp. 349–356, 2003.

[30] A. G. Sikora, N. Jaffarzad, Y. Hailemichael et al., “IFN-α enhances peptide vaccine-induced CD8+ T cell numbers,effector function, and antitumor activity,” Journal of Immunol-ogy, vol. 182, no. 12, pp. 7398–7407, 2009.

[31] Y. Kawakami, S. Eliyahu, K. Sakaguchi et al., “Identificationof the immunodominant peptides of the MART-1 humanmelanoma antigen recognized by the majority of HLA-A2-restricted tumor infiltrating lymphocytes,” Journal of Experi-mental Medicine, vol. 180, no. 1, pp. 347–352, 1994.

[32] M. R. Parkhurst, E. B. Fitzgerald, S. Southwood, A. Sette, S. A.Rosenberg, and Y. Kawakami, “Identification of a shared HLA-A∗0201-restricted T-cell epitope from the melanoma antigentyrosinase-related protein 2 (TRP2),” Cancer Research, vol. 58,no. 21, pp. 4895–4901, 1998.

[33] S. A. Chang, V. G. Lacaille, D. S. Guttoh, and M. J.Androlewicz, “Binding and transport of melanoma-specificantigenic peptides by the transporter associated with antigenprocessing,” Molecular Immunology, vol. 33, no. 15, pp. 1165–1169, 1996.

[34] P. A. Wearsch and P. Cresswell, “The quality control of MHCclass I peptide loading,” Current Opinion in Cell Biology, vol.20, no. 6, pp. 624–631, 2008.

[35] L. W. Thompson, C. F. Garbee, S. Hibbitts et al., “Competitionamong peptides in melanoma vaccines for binding to MHCmolecules,” Journal of Immunotherapy, vol. 27, no. 6, pp. 425–431, 2004.

[36] D. E. Speiser and P. Romero, “Molecularly defined vaccinesfor cancer immunotherapy, and protective T cell immunity,”Seminars in Immunology, vol. 22, no. 3, pp. 144–154, 2010.

[37] O. C. Krup, I. Kroll, G. Bose, and F. W. Falkenberg, “Cytokinedepot formulations as adjuvants for tumor vaccines. I.Liposome-encapsulated IL-2 as a depot formulation,” Journalof Immunotherapy, vol. 22, no. 6, pp. 525–538, 1999.

[38] M. L. Van Slooten, G. Storm, A. Zoephel et al., “Liposomescontaining interferon-gamma as adjuvant in tumor cell vac-cines,” Pharmaceutical Research, vol. 17, no. 1, pp. 42–48, 2000.

[39] I. Gursel, M. Gursel, K. J. Ishii, and D. M. Klinman,“Sterically stabilized cationic liposomes improve the uptakeand immunostimulatory activity of CpG oligonucleotides,”Journal of Immunology, vol. 167, no. 6, pp. 3324–3328, 2001.

[40] B. B. Gowen, J. Fairman, S. Dow et al., “Prophylaxis withcationic liposome-DNA complexes protects hamsters fromphleboviral disease: importance of liposomal delivery andCpG motifs,” Antiviral Research, vol. 81, no. 1, pp. 37–46, 2009.

[41] T. Towata, Y. Komizu, S. Suzu, R. Ueoka, and S. Okada,“Highly selective fusion and accumulation of hybrid lipo-somes into primary effusion lymphoma cells along withinduction of apoptosis,” Biochemical and Biophysical ResearchCommunications, vol. 393, no. 3, pp. 445–448, 2010.

[42] C. Jacobs, P. Duewell, K. Heckelsmiller et al., “An ISCOMvaccine combined with a TLR9 agonist breaks immuneevasion mediated by regulatory T cells in an orthotopic modelof pancreatic carcinoma,” International Journal of Cancer, vol.128, no. 4, pp. 897–907, 2011.

[43] H. I. Cho and E. Celis, “Optimized peptide vaccines elicitingextensive CD8+ T-cell responses with therapeutic antitumoreffects,” Cancer Research, vol. 69, no. 23, pp. 9012–9019, 2009.

[44] J. L. Perez-Gracia, P. Berraondo, I. Martinez-Forero etal., “Clinical development of combination strategies inimmunotherapy: are we ready for more than one investiga-tional product in an early clinical trial?” Immunotherapy, vol.1, no. 5, pp. 845–853, 2009.

[45] G. Dotti, “Blocking PD-1 in cancer immunotherapy,” Blood,vol. 114, no. 8, pp. 1457–1458, 2009.

[46] A. J. Rech and R. H. Vonderheide, “Clinical use of anti-CD25antibody daclizumab to enhance immune responses to tumorantigen vaccination by targeting regulatory T cells,” Annals ofthe New York Academy of Sciences, vol. 1174, pp. 99–106, 2009.


Review Article

Immune Response in Ovarian Cancer:How Is the Immune System Involved in Prognosis and Therapy:Potential for Treatment Utilization

Nikos G. Gavalas, Alexandra Karadimou, Meletios A. Dimopoulos, and Aristotelis Bamias

Department of Clinical Therapeutics, Medical School, University of Athens, Alexandra Hospital, 80 Vasilissis Sofias Avenue,115 28 Athens, Greece

Correspondence should be addressed to Aristotelis Bamias, [email protected]

Received 1 July 2010; Accepted 17 December 2010

Academic Editor: Stuart Berzins

Copyright © 2010 Nikos G. Gavalas et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Ovarian cancer is one of the leading causes of cancer-related death among women. Resistance to the disease occurs in morethan 70% of the cases even after treated with chemotherapy agents such as paclitaxel- and platinum-based agents. The immunesystem is increasingly becoming a target for intense research in order to study the host’s immune response against ovarian cancer.T cell populations, including NK T cells and Tregs, and cytokines have been associated with disease outcome, indicating theirincreasing clinical significance, having been associated with prognosis and as markers of disease progress, respectively. Harnessingthe immune system capacity in order to induce antitumor response remains a major challenge. This paper examines the recentdevelopments in our understanding of the mechanisms of development of the immune response in ovarian cancer as well as itsprognostic significance and the existing experience in clinical studies.

1. Introduction

Cancer is one of the leading causes of death in the developedworld outnumbering even heart disease in the United States[1]. In turn, ovarian cancer remains the leading cause ofdeath among gynaecological malignancies and is the fourthmost common cause of cancer-related death among women.Epithelial ovarian cancer is the main type of the diseaseaccounting for more than 90% of all malignant ovariantumors. According to the initial FIGO stage, the prognosisof ovarian cancer varies; a 5-year survival reaches 90% whenthe disease is confined within the ovary, but it drops to below50% for the cases that cancer has spread outside the pelvis.Ovarian cancer is usually diagnosed in advanced stages(FIGO stages III and IV), and prognosis is generally ratherpoor. Major established prognostic factors, apart from FIGOstage of the disease, include tumor grade, histologic subtype,and the volume of disease remaining after cytoreductivesurgery [2]. Nevertheless, the value of these factors in

a population with advanced stage and usually high-gradetumors is limited.

Current treatment of advanced ovarian carcinomaincludes debulking and chemotherapy, mainly the combi-nation of the use of paclitaxel and platinum agents and atleast 70% of the patients treated with the above combinationinitially respond to treatment. Intraperitoneal drug admin-istration has substantially improved the survival of patientswho have minimal gross disease remaining after surgery andcan also tolerate the side effects of aggressive treatment [3].

Despite the significant advances in surgery and chemo-therapy, the disease is more likely to relapse in about 70% ofthe cases [4] with resistance being prevalent in most cases.As a result, new ways of treating the disease are currentlybeing explored focusing on the biology of cancer and morespecifically within the ovarian tumor microenvironment.Therefore, clinical research has focused on molecular mark-ers, which are related either to the behaviour of the disease orthe response to chemotherapy in order to define the outcome


VEGFR ID

VEGFR TD

Signalling

VEGFmolecules

VEGFR ED

Binding

Plasmamembrane

Tumor environment

vasculogenesis, proliferation, etc.

Angiogenesis,

Figure 1: VEGF exerts its signalling effect via its receptor VEGFR.VEGF, mainly the VGEFA isoform exerts its effects via binding itsreceptor VEGFR (mainly VEGFR2). It is a powerful angiogenicfactor that holds a pivotal role in tumor progress and metastasis.It comprises an attractive target for possible agents that will block itsfunction and therefore enhance patients’ survival. ID: Intracellulardomain, ED: extracellular domain, TD: Transmembrane domain.

in these patients and establish furthermore potential targetsfor therapy.

Oncogenesis in all types of cancer, including ovarian can-cer, is a process that involves multiple molecular pathways,which regulate important functions of cancer cells. In 2004,the Baltimore group proposed a model for the division ofepithelial ovarian tumors into two rather broad categoriestermed type I and type II, that correspond to two mainpathways of tumorigenesis [5].

The major groups are genes involved in apoptosis and cellcycle regulation, genes encoding for growth factors and genesinvolved in angiogenesis. The prognostic and predictive valueof several factors implicated, in these pathways, has beenrecently studied. Genetic alterations in associated genes, suchas mutations of p53, malfunctioning genes of the BRCAfamily (BRCA1 and BRCA2) in about 15% of inherited typesof ovarian cancer [6], malfunction of tumor suppressor genessuch as ARHI [7], the cyclinE/CDK2 and cyclinD/CDK4complexes and the cell cycle regulators p27, p15, and p16have all been studied in this context [8–11]. Although somestudies have reported relevant associations, the prognosticrole of these factors remains to be elucidated in full.

Angiogenesis is a critical function for the expansion ofa tumor and also for its metastatic potential, and it is influ-enced by the tumor microenvironment [12]. Its significancein ovarian cancer has been well established, and a numberof angiogenic factors have been identified. The vascularendothelial growth factor (VEGF) holds a pivotal role in theangiogenic process [13]. It is produced by cancer cells andassists tumor progression and metastasis (Figure 1) exertinga central role in the formation of ascitic fluid and metastasisin the peritoneum. It is also related to the invasive andmetastatic potential of ovarian cancer [14–16].

Immune surveillance has long been recognized as animportant element of host anticancer response. Agents whichaugment immune response as well as antibodies againstcertain tumor antigens have been approved for the treatmentof different types of neoplasms. In the recent years, we havewitnessed important developments in our understanding ofcancer immunology. Many of these developments involveovarian cancer, and this paper will focus on them.

2. Cancer and the Immune System

The immune system responds to the presence of cancer anti-gens. A key advance in recent advances in immunology hasbeen the elucidation of antigen-specific cell recognition anddestruction of target cells. Mutations can occur in commonantigens that are found in otherwise normal functioninggenes in the cell; these were initially termed the tumor-specific antigens [17], and on those that can be found inboth normal and cancer cells called the tumor-associatedantigens (TAA) [18]. This terminology is still extensivelyused but it has been termed as imperfect by researchers andalthough still present in the literature, other modern antigenclassifications have emerged based on the antigens’ molecularstructure and source. More modern terminology dividesantigens into categories such as differentiation antigensand overexpression antigens [19] and also viral antigens.A distinct example of the latter category is Epstein-BarrVirus Nuclear Antigen (EBNA-1), which is associated withBurkitt’s lymphoma and nasopharyngeal carcinoma [20].Identifying tumor antigens has been an ongoing process witha number of techniques, having been employed, based onseveral components of the immune system [21–24].

In contrast to early theories that a tumor could notelicit an immune reaction, later experiments showed thatit actually does provoke the onset of an immune response[25–27]. More specific studies have shown that both theinnate and adaptive “arms” of the immune system are impli-cated in antitumor response [28, 29]. There is a number ofcomponents of the immune system that have been implicatedwith cancer cell elimination, equilibrium, and also escapefrom immune surveillance; all three comprising what iscalled “immunoediting” [30], a process that emphasizes inthe dynamic interaction of the immune system with cancer,and it is present in almost all types of tumor includingovarian cancer. It is a process that has been reinforced in thelast few years for its usage in cancer progress. Immunoeditingis divided in elimination, equilibrium, and escape. At first,cancer is eliminated, rendered nondetectable, followed by


a period of being kept in check by the immune system,and finally cancer becomes clinically detectable when ithas escaped antitumor immunity. Thus the immune systemprotects the host from cancer and it also plays a rolein “sculpting” immunogenicity, and this has actually beenexperimentally shown [31].

Elimination and equilibrium are achieved via lympho-cytes, mainly the T cell subpopulation [32]. In cancerpatients the “healthy” response against the tumor is counter-acted by a suppressive, tumor-driven effect. This hypothesisis strengthened by recent studies showing that the absence orpresence of T cells in colorectal cancer specimens more accu-rately predicted the outcome than using standard prognosticfactors [33]. Other studies in different types of tumor, mainlycervical and breast cancer, have also shown similar results[34, 35]. These studies further confirmed the importanceof the immune response in prognosis alongside other moreestablished factors. Recent studies also support the case ofimmunoediting by observing that tumor infiltration by lym-phocytes is linked to tumor-associated immune response,mainly showing that the presence of tumor infiltratinglymphocytes may be associated with improved prognosisand clinical outcome in cancer patients [36–38] includingovarian carcinoma [39, 40]. These observations as well aspreclinical data also suggest that by enhancing the hostimmune system, it may achieve tumor destruction and actsynergistically with other anticancer therapies.

Although the development of antitumor immuneresponse has been well established, there is also evidence thattumors can escape destruction by suppressing the immunesystem both within the cancer microenvironment and alsoon a systemic basis. T regulatory cells (Tregs), for example,that can suppress effector T cells action have been found inthe microenvironment of several types of tumor [41–43].Similar effects on regulation of Tregs can also be broughtabout in systemic modes of immunosuppression by tumors.For example, an increase in blood Tregs content has beenobserved in melanoma [44]. In colorectal cancer, increasednumbers of activated granulocytes [45] have also beenreported. Such cell types have shown to suppress tumor-specific T cells in mouse models [46]. Other types of im-munosuppression consist of the downregulation of MajorHistocompatibily Complex (MHC) and tumor antigen loss[47]. They also include disruption of specific NaturalKiller (NK) cells employment that inhibit immune system-mediated tumor destruction [41, 48].

3. Ovarian Cancer and Lymphocyte Response

Epithelial ovarian cancer is characterized by periods ofremission and relapse of sequentially shortening durationuntil chemoresistance occurs [40]. Such patients are the bestcandidates for immunological studies, since T cells’ presencecan be utilized as markers for disease progress and can beevaluated at different stages of the disease. The progressionof cancer in the peritoneal cavity and the frequent formationof ascites, which characterize advanced stages of ovariancancer, mainly stage IV, make this tumor a model for thestudy of different lymphocytic populations. Ascitic fluid as

well as peritoneal metastases can be easily obtained throughparacentesis, laparoscopy, or open surgery, and cells can bescreened by various techniques such as flow cytometry orimmunohistochemistry.

It is believed that the presence or absence of specific pop-ulations of T cells, which hold a central role in immunoedit-ing within epithelial ovarian cancer tumors, is associatedwith important differences in prognosis. Studies in paraffin-embedded tissues have reinforced this notion and haveshown that the presence of tumor infiltrating lymphocytes(TIL) such as CD3+ cells and increased number of cytotoxicCD8 lymphocytes were associated with prolongation ofsurvival [49–51]. For example, in the case of CD3 TILs,Tomsova et al. have shown that patients exhibiting higherCD3 cell numbers had an improved overall survival of 60months over 29 months for patients that had lower CD3 cellnumbers.

Elimination is also conferred by CD3+ CD56+ cells,containing the NK-like T cytotoxic cells which have cytotoxicproperties against tumor cells and contain the highestsuch property among effector killer cells in vitro [52, 53].Experiments, using blood cells from lymphoma patients,showed significant expansion of this cell population in exvivo conditions, which accounted for the 20% of a cytokine-induced population that resulted in significant cytotoxicityagainst cancer cells in vitro [54]. Frozen tissue has alsobeen used in immunohistochemical studies showing similarresults [55], where the presence of CD3+ cells was shown inmost cancer specimens. In this paper, immunohistochemicalstudies also showed the presence of CD4+ and CD8+ TILswith numbers that were closely related. Moreover, both typesof cells, CD4+ and CD8+, were both present or absentin specimens examined. The 5-year progress-free survivalpercentage for patients with the presence of TILs accordingto Zhang et al. was 38%. Nesbeth et al. have recently shownthe positive effect of CD4+ T cells in ovarian cancer via theuse of a novel mechanism that recruits dendritic cells to thetumor site that in turn activate tumor-specific CD8+ cellswhich then mediate long-term protection [56].

The presence of CD3+ CD56+ cells in ascitic fluid takenfrom advanced ovarian cancer patients has been shown to beinversely correlated with the presence of vascular endothelialgrowth factor (VEGF) [57]. In addition, low CD3+ CD56+content was correlated with poor prognosis and platinumresistance. NK cells’ rapid activation, and cytotoxic activitywithout need for prior sensitization and the release ofcytokines such as IFN-γ, TNF-α, and IL-10, indicates theirimportance [58]. Early studies have shown the efficacy ofNK cells against tumors when activated by cytokines [59, 60]or when ex vivo stimulated lymphokine-activated killer cellswere adoptively transferred into patients [61, 62]. Recentstudies though have shown that the expression of mucin(MUC) molecules on the ovarian cancer cell surface, namely,MUC16 which is a carrier for the CA125 tumor marker, assistin the avoidance of the tumor cells’ recognition by NK cells[63]. Human Leukocyte Antigen (HLA) class I antigens thatcan play a negative role in antitumor functionality of NK cellsare downregulated in ovarian cancer, hence making the use ofNK cells possibly quite important in ovarian carcinoma [64].


This is enhanced by findings that the formation of ascites inlate stage ovarian cancer may be inhibited by Cd-1-mediatedactivation of NK cells [65].

Another factor, termed programmed cell death 1 (PD-L1) which is expressed on tumor cells, has been shown to actas a prognostic factor. Its expression level has been shown tobe inversely correlated with CD8+ cell count rendering thisprotein a factor of poor prognosis, since it has been suggestedto directly inhibit CD8+ cells [66].

Dendritic cells migrate in a transendothelial manner viathe use of L1 IgCaM molecule as has been recently shown byMaddaluno et al. [67], an observation that may play a rolein tumor metastasis. L1 is a glycosylated protein that hasbeen recently reported to be expressed in 40%–70% of casesof epithelial ovarian cancer and is associated with poorprognosis [68].

In contrast to the augmentation of antitumor response bythe aforementioned populations, another specific subset ofT cells has been shown to play a key role in tumor immunity.T regulatory cells (Tregs) play a key role in peripheraltolerance. Since tumor-associated antigens (TAA) are selfantigens, they are subjected to control by peripheral tol-erance. Tregs within the CD4+ CD25+ T cell populationare characterized by the expression of the FoxP3+ protein[69, 70]. Humans bearing tumors show an elevated amountof Tregs in their blood as well as malignant effusions [71, 72].Sato et al. [69] identified cells in ovarian tumors expressingboth CD25 and FoxP3. Recently, the presence of Tregs inovarian cancer ascites in comparison to normal asciteshas been shown [72]. The presence of Tregs in ovariantumors has been associated with reduced overall survival[73, 74]. More specifically, Curiel et al. showed for thefirst time that CD4+ CD25+ FoxP3+ Treg cells correspondto poor clinical outcome in epithelial ovarian cancer. Thesame study also showed that CD4+ CD25+ CD3+ cellpopulations were much more concentrated in malignantascites rather than nonmalignant ones and in blood. It wasalso shown that CD4+ CD25+ cells were preferentiallyconcentrated in tumor mass rather than in tumor draininglymph nodes. Furthermore, the presence of FoxP3 alonewas an independent prognostic factor for progress-free andoverall survival.

Therefore, Tregs depletion can be expected to lead tomore efficient treatment and better prognosis. Current ther-apeutic agents may be useful in this respect. Classical cyto-toxics, such as cyclophosphamide [75] as well as antibody-based immunotherapy with Trastuzumab have been shownto result in a substantial decrease in the number of Tregsin cancer patients [76]. A recent study has shown selec-tive accumulation of NK-T cells, activated CD4 and CD8lymphocytes and also Tregs in ascites formed in ovariancancer [72], which complements previous evidence thattumor-associated lymphocytes are indeed present in ascites[70, 73, 77] and may be important for the immune responseagainst the tumor. These results indicate that the presence ofcancer cells can activate lymphocytes and could also resultin a parallel accumulation of Tregs that may inhibit CD8-mediated immune response against the tumor as has beensuggested before [71, 78]. Recent studies also indicate that

in the case of epithelial ovarian cancer, local treatment withinterleukin 2 may play a role in converting Tregs into Th17cells, a new player in the field of cancer immunotherapy,with a concomitant relief of Treg-mediated immune sup-pression and enhancement of antitumor immunity [79, 80].Plasmacytoid dendritic cells (PDc) have also been shownto contribute to immunosuppression in ovarian cancer byinducing tumor microenvironment Tregs [81].

Another type of cells of the immune system, namelymacrophages, are also found in ovarian cancer [82, 83].The presence of macrophages in tumors has been associatedwith tumor growth and metastasis in rodents [84, 85].Kryczek et al. [83] have shown that the B7-H4+ receptorexpression, which is a negative T cell regulator on tumor-associated macrophages, in ovarian cancer, induces sup-pression of T cells encompassing tumor-associated antigensimmunity.

Finally, since the increased concentration of autoanti-bodies can induce the production of Tregs and clinical studieshave reported autoimmune paraneoplastic syndromes (dif-ferent from autoimmune diseases) [86, 87], there may belinks between cancer and autoimmune disease that remainto be elucidated in full. These studies may provide us witha greater insight into Tregs activity and association withovarian cancer.

Lately, different populations such as vascular lympho-cytes have shown the ability to form functional blood vessels,and they may be proven to be an important target forblocking cancer progression [88].

The identification of important subsets of lymphocytesin tumors and ascites from ovarian cancer has led to the studyof possible immunomodulatory effects of current therapies.Chemotherapy, in particular paclitaxel, may have a positiveeffect on the immune response by directly downregulatingTregs [89]. Tregs can also be suppressed by cyclophos-phamide as has been exhibited in mouse models [75, 90],and NK cells can be activated at the same time. The useof gemcitabine, which is a nucleoside analog, reduced thenumber of myeloid suppressor T cells, without reducingcytotoxic cells such as NK cells [91]. Gemcitabine, inassociation with oxaliplatin and interleukins such as IL-2 and GM-CSF, can have a suppressive effect on Tregs[92, 93]: therefore, it could possibly have a positive effect onreducing drug resistance and influence prognosis and diseaseoutcome.

4. Cytokines, Growth Factors and Associationwith Lymphocytes’ Mobility and Response

The composition of lymphocytic populations in blood,ascites and tumors is regulated by various cytokines andchemokines produced by the tumors or the components ofthe immune system. A simple schematic representation ofthese interactions is depicted in Figure 2.

A number of cytokines have been associated with a directeffect on tumor cells, via surface receptors such as Toll-likereceptors [94], but mainly they have been attributed roles inassisting the immune response of the body against tumors.Host antitumor response results from the balance between


TNF-α↓, H2O2↓,NO↓

Macrophages

Tregs Cell number ↑

Proliferation ↓,apoptosis↓,

CD8

Activation

CD4 Th1

CD4 Th2

IL-10, VEGF,TGF-β

IL4↑, IL10↑,IL13↑

Cell promotion↑

Dendritic cells

Cell numbers ↓,dysfunctional,

apoptosisActivation

Activation

NK cells

TNF-α ↓, IL2↓,IL12↓

Figure 2: Schematic representation of characteristic immune cells, growth factors, and cytokines interactions in cancer. Interactions betweengrowth factors such as VEGF, cytokines (e.g., TNFα) and T cells (e.g., NK, Tregs) are shown in this diagram. Tumor cells bring about theproduction of cytokines that assist in the mobilization of T cells and induce the production of further cytokines, and they also utilize growthfactors such as VEGF to promote neovasculirisation implicated in metastasis. ↑means increase where ↓means decrease.

the T helper 1 (Th1) response, which potentiates immuneresponse and the T helper 2 (Th2) response with a shiftin favor of the latter characterising oncogenesis and diseaseprogression. Both Th1 and Th2 immune responses have beenassociated with the production of cytokines such as Inter-leukin 12 (IL-12), Interleukin 4 (IL-4), Interferon gamma(IFN-γ), Tumor Necrosis Factor (TNF-α) (Th1 response),and IL-10 (Th2 response) [16, 95–97]. These cytokinescan also be produced by cancer cells; they are present inascites and have been associated with prognosis in ovariancancer [71, 98–100]. Gradients between blood and ascitesmay play a role in migration of leukocytes [101] andfactors that facilitate such movements may include L1 [67].As a consequence, different lymphocytic populations areinvolved in the two types of response: for example, CD3+CD56+ cells are associated with Th1 whereas CD4+ CD25+cells are associated with Th2 response.

The prognostic role of various cytokines has beenstudied, but no absolutely firm conclusions can be drawn sofar. It is conceivable that cytokines involved in Th1 responseare expected to predict for better prognosis, while theopposite is expected in those associated with Th2 response.Interleukins in that respect have received much attention. IL-2 initiates the activation of T and NK cells and is also essential

for the maintenance of self-tolerance through generation andmaintenance of Tregs [102] or by activation-induced celldeath [103] to eliminate self reactive T cells. Cytokines suchas IL-12 [104] and IL-21 [105] are currently considered fortheir therapeutic potential in other types of cancer and mayhave the same effect in ovarian cancer. In glioma, in thecase of IL-12, the cytokine is fused with normal glioma cellsand dendritic cells and administered to malignant gliomapatients [104]. IL-12 is associated with favorable prognosis,and in this study, four patients exhibited a glioma reductionof 50%. For IL-21, Dou et al. have shown that when the geneexpressing IL-21 is administered in rodents, it has a positiveantitumor effect in squamous cell carcinoma, and thereforeIL-21 may be associated with favorable prognosis. This hasfurther been enhanced by a recent study showing that theantitumor effect is increased by human ovarian cancer cellssecreting IL21 alone or in combination with GM-CSF [105].TNFα may also be associated with prognosis [72, 106, 107],but reports on whether it is a signature of poor or betterprognosis vary. IL-6 levels have been shown to be increasedin ovarian cancer patients’ serum [108, 109], and it wascorrelated with poor overall survival. Another cytokine thatwas shown to be associated with the growth of cancer cellsand tumor proliferation is IL-1 [110, 111]. IL-15 has also


been recently shown to activate CD8+ and NKT cells thatmay inhibit tumor growth [112]. Further functional studiesare necessary to confirm the above results.

A cytokine that seems to be heavily involved in tumor im-munosuppression is transforming growth factor beta (TGF-β), a protein that affects proliferation, activation, and differ-entiation of immune cells and inhibits antitumor immuneresponse [113]. In cancer cells, the production of TGF-β isincreased, which in turn increases the proteolytic activityof cells and the binding to cell adhesion molecules in theextracellular matrix. TGF-β can also convert effector T cellsinto Tregs [114]. It has been reported that it can also promoteangiogenesis and that process can be blocked by anti-TGF-βantibodies [115].

TNFα is produced by tumor cells and can induceautocrine proliferation and disease progression in ovariancancer [107, 116, 117]. The autocrine action of TNFα mayhave direct effects on tumor cell spread via acting on thechemokine receptor CXCR4 and also stimulation of bloodvessel formation in the peritoneal tumor by inducing expres-sion of VEGF and CXCL12 [118]. In contrast, TNFα levelshave also been inversely correlated with the presence ofCD4+ CD25+ cells, and have been shown to directly down-regulate Tregs [119]. This might indicate a favorable effect ofthis cytokine on prognosis and underlines the complexity ofthe functions that each of these factors may possess.

A family of proteins called chemokines (CC) may also beinfluencing cellular composition in biological fluids. Recentstudies have exhibited the detection of mRNA for CCL2,CCL3, CCL4, and CCL5 in solid ovarian tumors by in situhybridization [120]. Moreover, CCL5 has been shown to besecreted by CD4+ T cells, recruits CCR5+ dendritic cellsto the tumor location, and activates them through CD40-CD40L interactions [56]. The newly matured dendritic cellsprime tumor-specific CD8+ cells thus providing with longterm protection.

In the protein-rich ascitic fluid, different chemokinemolecules have been shown to be expressed, with CCL2 beingthe predominant one [121]. In addition, chemokine stromal-derived factor-1 (CXCL-1) induced the migration of plas-macytoid dendritic cells into the tumor microenvironmentin cases of ovarian cancer and induced delivery of survivalsignals to PDC. In turn, the tumor microenvironmental PDCinduced IL-10 expressing Tregs [122], which is correlated topoor prognosis and shorter progress-free survival. Tregs, andIL-10 are associated with poor prognosis in many types ofcancer. In the case of Tregs it has been exhibited that CCL22plays a central role in inducing influx of these cells into tumorsites, and it binds CCR4 that is expressed on Treg surface[123].

Interferon gamma (IFN-γ) plays a stimulatory role formacrophages turning them from immunosuppressive to im-munostimulatory cells [124]. It also skewed monocyte dif-ferentiation from tumor-associated macrophages- (TAM-)like cells to M1-polarized immunostimulatory macrophages.Taken together these data show that IFN-γ overcomes TAM-induced immunosuppression by preventing TAM generationand functions.

Furthermore, cytokines such as interleukin 18 (IL-18)[125] and stroma derived factor 1 (SDF-1) [126] have beenshown to be correlated with poor prognosis in ovarian cancerpatients, but further studies are required to fully evaluatethem in the tumor microenvironment and the periphery.

VEGF holds a very important role in the oncogenesis aswell as progression and prognosis in ovarian cancer [55, 127].It is selectively accumulated in ascites and occurs in advancedstages of the disease but not in ascites from cirrhosis [55, 57].Up to now, this has been attributed solely to its angiogenicproperties. Recently, it has been suggested that VEGF alsoexerts an immunosuppressive effect in cancer, as it wascorrelated with low levels of IL12, inhibition of dendritic cellmaturation, low numbers of NK-T cells, and upregulation ofTregs [58, 59, 128–130]. It can also induce expression of theT cell cosignaling molecule B7-H1 on myeloid dendritic cells(MDC). Barnett et al. [15] have reported that the blockageof B7-H1 improved T cell-mediated immune response andtumor clearance in an ovarian cancer mouse model. VEGFexerts its effects via its receptor, VEGFR, mainly VEGFR2[13, 131]. This type of receptor has the ability of activatingthe mTOR protein through the Akt/mTOR pathway [131].Inactivation of mTOR may lead to downregulation of IL-2,thus conferring a direct negative effect in T cell proliferationas well as cancer cell proliferation [132, 133]. Except cancercells, the VEGFR2 protein has been recently shown to beexpressed selectively on a subset of T cells, namely, the CD4+FoxP3+ Tregs [134]. Since FoxP3high Tregs are associatedwith poor prognosis, the expression of VEGFR2 on theirsurface may be attributed with a more prominent role inangiogenesis in the future.

The prognostic significance of VEGF in ovarian cancerhas received much attention recently. Several studies haveassociated serum or plasma levels of VEGF with prog-nosis [127, 135, 136]. Ascites VEGF levels may be moreinformative, since it reflects the site of the most intensedisease activity. It has been shown that VEGF levels above1900 pg/ml were associated with inferior survival in a seriesof 41 patients with advanced ovarian cancer [57, 72]. Theseresults have been confirmed by a more recent analysis ofa larger series and longer followup (Figures 3(a) and 3(b)show the updated results). Finally, in recent studies serumFas protein (sFas) levels and serum VEGF levels have beenfound to be increased in ovarian cancer patients correlatedwith a short duration of the relapse-free period [137].

5. Harnessing the Immune System for CancerTherapy: A Driven Response

In general, there are three approaches to harnessing theimmune system response in order to fight cancer: (1) useexogenously administered antibodies, (2) elicit a humoraland a cellular response, and (3) explore the activation and/orgeneration of antigen-specific CD4+ and CD8+ cells. Thestrategies which are in the more advanced stages of drugdevelopment are the use of monoclonal antibodies andcytokines. The other strategies will be discussed more briefly.

Antibodies with the potential to be used in cancertreatment are often targeting either the tumor directly,


0 10 20 30 40 500

0.2

0.4

0.6

0.8

1

Pro

babi

lity

ofP

FS

Months after start of chemotherapy

VEGF > 1900Censored

VEGF ≤ 1900Censored

(a)

0 10 20 30 40 500

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Pro

babi

lity

ofsu

rviv

al

Months after start of chemotherapy

VEGF > 1900Censored

VEGF ≤ 1900Censored

(b)

Figure 3: Clinical data concerning patients undertaking chemotherapy. Progression-free survival (a) and overall survival (b) of 54 patientswith advanced ovarian cancer receiving first-line, platinum-based chemotherapy, according to VEGF levels in ascites. The lower levels wereassociated with significantly longer progression-free (P = .0297) and overall (P = .0164). Median followup: 33 months.

the tumor microenvironment, or function as modulators ofimmune response [138]. Another way is to target intra-cellular pathway molecules by the use of cell penetratingagents [139]. Antibody immunotherapy does not seem tointerfere with suppressor mechanisms that could limit itstreatment capacity. Antibodies usually act by the inductionof death pathways by engaging with receptors on cell surface,antibody-dependent cellular cytotoxicity (ADCC), and theblockade of tumor growth factors such as vascular endothe-lial growth factor (VEGF). There is a growing number ofpotential agents, mainly antibodies, currently undergoingevaluation in clinical trials [140]. Such antibodies includeTrastuzumab [141], Oregovomab [142], Bevacizumab, andCetuximab [143, 144]. Published data are shown in Table 1.

VEGF, mainly the VEGF-A isoform, may be the morepromising therapeutic target. It is a powerful angiogenicmolecule that has been associated with tumor progression,poor prognosis, and drug resistance in ovarian cancer.In addition, it has immunosuppressive properties, as pre-viously discussed. Recent data have suggested that an anti-VEGF monoclonal antibody (Bevacizumab) is efficient inplatinum-resistant disease [145–147, 154]. By combiningpaclitaxel and/or carboplatin agents with VEGF inhibitors,such as bevacizumab, we may overcome resistance to chem-otherapy. This hypothesis is currently tested in two ran-domised studies [148, 149]. Both studies showed a significantPFS prolongation by the administration of Bevacizumab.Another monoclonal antibody already tested in a phaseIII randomized study is oregovomab, which recognizes

an epitope on CA125. The formation of the oregovomab-CA125 complex results in the development of CA125-specificimmune response [150]. The development of such responsehas been shown to predict improved survival in a small phaseII study [151]. In the phase III study, although no survivaladvantage was found when it was given as maintenanceafter remission following first-line chemotherapy, subgroupanalysis showed that patients with low-volume residualdisease (<2 cm), Ca125 ≤ 65 IU/mL after the 3rd cycle ofchemotherapy, and CA125 ≤ 35 IU/mL at entry experienceda 2-fold increase in median time to progress (TTP) [152].The IMPACT study is currently evaluating the role oforegovomab in this subset of patients.

The use of cytokines in cancer therapy has also beenevaluated. Certain cytokines, such as IFNs, can augmentantitumor response and were considered as promising agentsin cancer therapy. IFN-α, is approved for the treatmentof malignant melanoma and kidney cancer. It has beenshown that GM-CSF-secreting tumor cell immunotherapywith VEFG-blocking agents prolonged survival of cancerbearing mice [155, 156], while IL-2 and GM-CSF can have asuppressive effect on Tregs [92, 93]. GM-CSF in combinationwith recombinant IFN-γ1 and carboplatin in a phase IItrial has been recently shown to have a reasonable responseagainst recurrent platinum sensitive ovarian cancer [157]. Allthese preclinical data suggest that the use of cytokines may beefficacious in ovarian cancer. IFN is the most well-studiedagent. Several randomized studies, based on promisingphase II results, have been published during the last decade


Table 1: Selected clinical studies of monoclonal antibodies used for the treatment of ovarian cancer.

AntibodyMechanism ofaction

Representative Phase II studies Phase III studies

Population Treatment Results Population Treatment Results

Bevacizumab(Genentech/R-oche)

Binds to VEGFAntiangiogenicImmunosuppre-ssive

Refractory(n = 32) [145]

Monotherapy(n = 23) Withchemotherapy(n = 9)

RR 16%PFS 5.5 mOS 6.9 m

First-line ICON7 [146]

Carboplatin/PaclitaxelversusCarboplatin/Paclitax-el/Bevacizumab

MedianPFS17.3 mversus19 m,P = .004110.3 mversus11.2 mversus14.1,P < .00001

Refractory(n = 44) [147]

MonotherapyRR 16%PFS 4.4 mOS 10.7

GOG 218 [148]

Carboplatin/PaclitaxelversusCarboplatin/Paclitax-el/Bevacizum abversusCarboplatin/Paclitax-el/Bevacizum ab+ Bevacizumabmaintenance

Oregovomab(AltaRex Corp)

Binds to CA125Development ofa humoral andcellular antitumorresponse

2nd linetreatment(n = 20) [149]

Withchemotherapy

Developmentof T cellresponse wasassociatedwithimprovedsurvival

Maintenanceafter first-line(n = 147) [150]

Oregovomab versusplacebo

MedianPFS13.3 mversus10.3 m,P = .71

Maintenanceafter first-lineResidual<2 cm,CA125 < 65after 3rd cycle,CA125 < 35 atentry (n = 354)

Oregovomab versusplacebo

Awaited

Trastuzumab(Genentech)

Binds to HER2extracellulardomain

Recurrent(n = 41) [141]

MonotherapyRR 7.3% PFS2 m

Pertuzumab(Genentech)

Inhibitor of HERdimerization

87% platinum-resistant(n = 123) [151]

monotherapyRR 4.3% PFS6.6 w

Cetuximab(Bristol-MyersSquibb)

EGFR inhibitorFirst-line(n = 41) [152]

Combinationwithpaclitaxel/carb-oplatin

PFS 14.4 m

Matuzumab(Merck/Sero-no/Takeda)

EGFR inhibitorPlatinum-resistant(n = 37) [153]

MonotherapyRR 16.2 mTTP 54d OS13.3 m

evaluating the role of interferon in addition to first-linetherapy or as maintenance strategy. The results of thesestudies are summarized in Table 2. The first study showeda PFS but not OS benefit [158]. Nevertheless, the standardof Cisplatin/Cyclophosphamide, used in that study has beensubstituted by Paclitaxel/Carboplatin, and thus these resultsare difficult to be viewed in the context of current practicein ovarian cancer. Two randomized studies using the currentstandard showed no benefit from the addition of IFNs in thetreatment of ovarian cancer [159, 160].

Methods to augment an immune response against tumorantigens have also been explored [161, 162]. The moststudied have been vaccines or macrophage-activated killer

(MAK) cells. Within this context, IFN-γ has recently beenshown to reverse the immunosuppressive properties ofmacrophages so its local administration could potentiallyincrease the efficacy of antitumor immunotherapies based onthe generation of effector T cells [163], an observation thatcontradicts previous studies mentioned above where IFN-γ showed no positive effect within the tumor microenvi-ronment. Tumor antigens, synthetic tumor peptides, wholetumor cells, tumor cell lysates, or anti-idiotypic antibodiesare among the list of initiators of an immune response[161]. In some protocols, injection of synthetic peptides incombination with GM-CSF is performed. In different pro-tocols, dendritic cells (antigen presenting cells) loaded with


Table 2: Selected clinical studies of cytokines for the treatment of ovarian cancer.

Cytokine Phase III studies

Population Treatment Results

IFN-γ First line (n = 148) [156]Cisplatin/Cyclophosphamide versusCisplatin/Cyclophosphamide/IFNγ

3-year OS58% versus 74% (P = .23)3-year PFS38% versus 51% (P = .031)

IFNa-2a Maintenance after first-line (n = 300) [157]IFNa-2a versusObservation

No benefit

IFN-γ First line (n = 847) [158]Carboplatin/Paclitaxel versusCarboplatin/Paclitaxel/IFN-γ

Median OSNot estimated versus1138dHR: 1.45, P = .001

synthetic peptides, immunocomplexes of tumor-associatedantigens with antibodies [162] through activating Fcγ-R[163], or fusion of dendritic cells with tumor cells areutilized. Dendritic cells present antigens to CD4+ CD8+ cellswhile delivering stimulatory signals necessary for effectiveT cell activation. They can also directly downregulate animmune response or induce immune tolerance [164]. Vac-cines using either gene-modified dendritic cells or wholetumor cells have also been explored [163, 165]. Peptidevaccines have been used so far in a lesser extent sincethey have some important limitations [166]. Although thedevelopment of a specific immune response could be shownin patients undergoing such approaches [167, 168], their roleremains investigational.

The ex vivo expansion of immunologically relevantautologous populations have also been studied. MAK cellshave been used as a form of adoptive immunotherapy aloneor in combination with monoclonal antibodies [58, 169,170]. MAK can reach tumor sites by intraperitoneal infusion,but most studies are small and the role of this approachremains undetermined. Using specific CD4+ and CD8+cells against tumor antigens may provide another way offighting cancer. These cells need to be activated againsttumor antigens before being administered to the patient.Activation can be achieved by either stimulating peripheralblood mononuclear cells (PBMC) in vitro, or by ex vivoexpansion of TILs [162, 167]. Recently, the adoptive transferof T cells expressing chimeric NKG2D receptors can lead tolong-term, tumor-free survival in a murine model of ovariancancer [171]. Genetic modification of T cells is anotheremerging approach but its application in ovarian cancer hasnot been successful so far [165].

Agents such as oligodeoxynucleotides containing dinu-cleotides with unmethylated CpG motifs (CpG-ODN) thatrecruit and activate innate effector cells throughout theabdominal cavity to the tumor site might control tumor cellgrowth and ascites formation [172].

Reports for the implication of Tegs in suppression ofantitumor response in cancer development and prognosishave already been discussed. There are currently clinical trialsusing ONTAK in ovarian cancer patients, with encouragingresults [15, 173]. ONTAK is a fusion toxin that consistsof IL-2 genetically fused to the enzymatically active and

translocating domains of diphtheria toxin. It can depletefunctional Tregs, as shown by Curiel et al. [173] in ovariancancer patients (including one patient at stage IV), by 50%in serum and it is considered to lead to better prognosis.ONTAK is approved by the FDA to be used in the treatmentof CD4+ CD25+ Treg-mediated tumors.

6. Conclusion and Future Considerations

Both the innate and adaptive immune response can be ofgreat importance in the battle against ovarian cancer.Throughout this paper, mechanisms of reaction of theimmune system against tumors were highlighted, stressingthe importance of such anti tumor response.

The prognosis of advanced ovarian cancer has beenimproved in the recent years. Nevertheless, after the intro-duction of paclitaxel in first-line treatment, no dramaticadvance in progress-free survival of the patients usingcytotoxic chemotherapy can be foreseen in the immediatefuture. On the contrary, targeted therapies may hold a sig-nificant promise, as shown in other neoplasms. The immuneresponse against the tumor may be a promising target,especially after much recent data has associated variouselements with prognosis.

The previous decade was characterized by many attemptsto establish interferon as a standard in the treatment ofovarian cancer. The failure of those attempts stresses thedisease’s complexity. At the moment, monoclonal antibodiesseem to be the most promising agents, currently tested inphase III trials.

There is still much to clarify regarding the mechanismsgoverning the development of host antitumor response inorder to find strategies to augment it. The interaction withother important functions, such as angiogenesis, may implythat more than one function needs to be blocked for achiev-ing an efficient therapy. Further progress in basic researchin combination of the awaited results of large randomizedclinical trials will hopefully enrich our armamentariumagainst ovarian cancer.


No conflict of interests is to be reported.


References

[1] A. Jemal, R. Siegel, E. Ward, Y. Hao, J. Xu, and M. J. Thun,“Cancer statistics, 2009,” CA Cancer Journal for Clinicians,vol. 59, no. 4, pp. 225–249, 2009.

[2] R. E. Bristow, I. Puri, and D. S. Chi, “Cytoreductive surgeryfor recurrent ovarian cancer: a meta-analysis,” GynecologicOncology, vol. 112, no. 1, pp. 265–274, 2009.

[3] D. K. Armstrong, B. Bundy, L. Wenzel et al., “Intraperitonealcisplatin and paclitaxel in ovarian cancer,” The New EnglandJournal of Medicine, vol. 354, no. 1, pp. 34–43, 2006.

[4] R. F. Ozols, B. N. Bundy, B. E. Greer et al., “Phase III trialof carboplatin and paclitaxel compared with cisplatin andpaclitaxel in patients with optimally resected stage III ovariancancer: a Gynecologic Oncology Group study,” Journal ofClinical Oncology, vol. 21, no. 17, pp. 3194–3200, 2003.

[5] I. M. Shih and R. J. Kurman, “Ovarian tumorigenesis:a proposed model based on morphological and moleculargenetic analysis,” American Journal of Pathology, vol. 164, no.5, pp. 1511–1518, 2004.

[6] H. A. Risch, J. R. McLaughlin, D. E. C. Cole et al.,“Population BRCA1 and BRCA2 mutation frequencies andcancer penetrances: a kin-cohort study in Ontario, Canada,”Journal of the National Cancer Institute, vol. 98, no. 23, pp.1694–1706, 2006.

[7] Y. Yu, R. Luo, Z. Lu et al., “Biochemistry and biology ofARHI (DIRAS3, an impronted tumor suppressor gene whoseexpression is lost in ovarian and breast cancers,” in Methodsin Enzymology: Regulators and Effectors of Small GTPases: RasFamily., W. E. Balch, C. Der, and A. Hall, Eds., vol. 407, pp.455–467, 2006.

[8] A. Bali, P. M. O’Brien, L. S. Edwards, R. L. Sutherland, N.F. Hacker, and S. M. Henshall, “Cyclin D1, p53, and p21expression is predictive of poor clinical outcome in serousepithelial ovarian cancer,” Clinical Cancer Research, vol. 10,no. 15, pp. 5168–5177, 2004.

[9] L. G. Buchynska, I. P. Nesina, N. P. Yurchenko, O. O. Bilyk,V. N. Grinkevych, and V. S. Svintitsky, “Expression of p53,p21, p16 and Ki-67 proteins in serous ovarian tumors,”Experimental Oncology, vol. 29, no. 1, pp. 49–53, 2007.

[10] D. G. Rosen, G. Yang, K. Q. Cai et al., “Subcellularlocalization of p27 expression predicts poor prognosis inhuman ovarian cancer,” Clinical Cancer Research, vol. 11, no.2, pp. 632–637, 2005.

[11] J. E. Quinn, J. E. Carser, C. R. James, R. D. Kennedy, andD. P. Harkin, “BRCA1 and implications for response tochemotherapy in ovarian cancer,” Gynecologic Oncology, vol.113, no. 1, pp. 134–142, 2009.

[12] P. Carmeliet and R. K. Jain, “Angiogenesis in cancer and otherdiseases,” Nature, vol. 407, no. 6801, pp. 249–257, 2000.

[13] M. Shibuya and L. Claesson-Welsh, “Signal transduction byVEGF receptors in regulation of angiogenesis and lymphan-giogenesis,” Experimental Cell Research, vol. 312, no. 5, pp.549–560, 2006.

[14] C. Evans, A. G. Dalgleish, and D. Kumar, “Review article:immune suppression and colorectal cancer,” AlimentaryPharmacology and Therapeutics, vol. 24, no. 8, pp. 1163–1177,2006.

[15] B. Barnett, I. Kryczek, P. Cheng, W. Zou, and T. J. Curiel,“Regulatory T cells in ovarian cancer: biology and therapeu-tic potential,” American Journal of Reproductive Immunology,vol. 54, no. 6, pp. 369–377, 2005.

[16] R. Kim, M. Emi, and K. Tanabe, “Cancer immunosuppres-sion and autoimmune disease: beyond immunosuppressive

networks for tumour immunity,” Immunology, vol. 119, no.2, pp. 254–264, 2006.

[17] P. K. Srivastava and L. J. Old, “Individually distinct transplan-tation antigens of chemically induced mouse tumors,” Trendsin Immunology, vol. 9, no. 1–12, pp. 84–88, 1988.

[18] M. G. Mathieu, A. K. Miles, G. Li, S. E. B. McArdle, and R.C. Rees, “Cancer/testis antigens for therapeutic use,” Journalof the Balkan Union of Oncology, vol. 14, no. 1, pp. S97–S102,2009.

[19] C. S. Chu, S. H. Kim, C. H. Jun, and G. Coukos,“Immunotherapy opportunities in ovarian cancer,” ExpertReview of Anticancer Therapy, vol. 8, no. 2, pp. 243–257, 2008.

[20] K. E. Hellstrom, I. Hellstrom, and J. P. Brown, “Humantumor-associated antigens identified by monoclonal anti-bodies,” Springer Seminars in Immunopathology, vol. 5, no.2, pp. 127–146, 1982.

[21] N. Renkvist, C. Castelli, P. F. Robbins, and G. Parmiani,“A listing of human tumor antigens recognized by T cells,”Cancer Immunology, Immunotherapy, vol. 50, no. 1, pp. 3–15,2001.

[22] S. K. Nair, D. Bockzowski, M. Morse, I. R. Cumming, H.K. Lyerly, and E. Gilboa, “Design and testing of peptide-based cytotoxic T cell mediated immunotherapeutics to treatinfectious diseases and cancer,” Nature Biotechnology, vol. 16,pp. 364–369, 1998.

[23] J. Banchereau and R. M. Steinman, “Dendritic cells and thecontrol of immunity,” Nature, vol. 392, no. 6673, pp. 245–252, 1998.

[24] A. Berchuck, A. Kamel, R. Whitaker et al., “Overexpressionof HER-2/neu is associated with poor survival in advancedepithelial ovarian cancer,” Cancer Research, vol. 50, no. 13,pp. 4087–4091, 1990.

[25] D. J. Slamon, W. Godolphin, L. A. Jones et al., “Studies ofthe HER-2/neu proto-oncogene in human breast and ovariancancer,” Science, vol. 244, no. 4905, pp. 707–712, 1989.

[26] Y. Wang, G. B. Kristensen, A. Helland, J. M. Nesland,A. L. Borrensen-Dale, and R. Holm, “Protein expression andprognostic value of genes in the erb-b signaling pathway inadvanced ovarian carcinomas,” American Journal of ClinicalPathology, vol. 124, no. 3, pp. 392–401, 2005.

[27] S. Camilleri-Broet, A. C. Hardy-Bessard, A. Le Tourneauet al., “HER-2 overexpression is an independent marker ofpoor prognosis of advanced primary ovarian carcinoma:a multicenter study of the GINECO group,” Annals ofOncology, vol. 15, no. 1, pp. 104–112, 2004.

[28] M. Pieretti, C. Hopenhayn-Rich, N. H. Khattar, Y. Cao, B.Huang, and T. C. Tucker, “Heterogeneity of ovarian cancer:relationships among histological group, stage of disease,tumor markers, patient characteristics, and survival,” CancerInvestigation, vol. 20, no. 1, pp. 11–23, 2002.

[29] M. S. Shahin, J. H. Hughes, A. K. Sood, and R. E. Buller,“The prognostic significance of p53 tumour suppressor genealterations in ovarian carcinoma,” Cancer, vol. 89, pp. 2006–2017, 2000.

[30] S. J. Piersma, E. S. Jordanova, M. I. E. van Poelgeest et al.,“High number of intraepithelial CD8+ tumor-infiltratinglymphocytes is associated with the absence of lymph nodemetastases in patients with large early-stage cervical cancer,”Cancer Research, vol. 67, no. 1, pp. 354–361, 2007.

[31] G. P. Dunn, L. J. Old, and R. D. Schreiber, “The three Es ofcancer immunoediting,” Annual Review of Immunology, vol.22, pp. 329–360, 2004.

[32] H. E. Kohrt, N. Nouri, K. Nowels, D. Johnson, S. Holmes,and P. P. Lee, “Profile of immune cells in axillary lymph


nodes predicts disease-free survival in breast cancer,” PLoSMedicine, vol. 2, no. 9, pp. 0904–0919, 2005.

[33] P. J. Klemi, L. Pylkkanen, P. Kiilholma, K. Kurvinen, andH. Joensuu, “p53 Protein detected by immunohistochemistryas a prognostic factor in patients with epithelial ovariancarcinoma,” Cancer, vol. 76, no. 7, pp. 1201–1208, 1995.

[34] Y. Hashiguchi, H. Tsuda, T. Inoue, S. Nishimura, T. Suzuki,and N. Kawamura, “Alteration of cell cycle regulators cor-relates with survival in epithelian ovarian cancer patients,”Human Pathology, vol. 35, no. 2, pp. 165–175, 2004.

[35] L. Wang, J. Ma, F. Liu et al., “Expression of MUC1 inprimary and metastatic human epithelial ovarian cancer andits therapeutic significance,” Gynecologic Oncology, vol. 105,no. 3, pp. 695–702, 2007.

[36] M. J. Anderson, K. Shafer-Weaver, N. M. Greenberg, andM. A. Hurwitz, “Tolerization of tumor-specific T cells despiteefficient initial priming in a primary murine model ofprostate cancer,” Journal of Immunology, vol. 178, pp. 1268–1276, 2007.

[37] J. Schmielau and O. J. Finn, “Activated granulocytes andgranulocyte-derived hydrogen peroxide are the underlyingmechanism of suppression of T-cell function in advancedcancer patients,” Cancer Research, vol. 61, no. 12, pp. 4756–4760, 2001.

[38] M. Dougan and G. Dranoff, “The immune response totumors,” Current Protocols in Immunology, no. 85, pp.20.11.1–20.11.4, 2009.

[39] M. Terabe, S. Matsui, N. Noben-Trauth et al., “NKT cell-mediated repression of tumor immunosurveillance by IL-13and the IL-4R-STAT6 pathway,” Nature Immunology, vol. 1,no. 6, pp. 515–520, 2000.

[40] M. Markman, J. Markman, K. Webster et al., “Duration ofresponse to second-line, platinum-based chemotherapy forovarian cancer: implications for patient management andclinical trial design,” Journal of Clinical Oncology, vol. 22, no.15, pp. 3120–3125, 2004.

[41] T. L. Whiteside, “Immune responses to malignancies,” Jour-nal of Allergy and Clinical Immunology, vol. 125, no. 2,supplement, pp. S272–S283, 2010.

[42] E. Yakirevich, E. Sabo, O. Lavie, S. Mazareb, G. C. Spagnoli,and M. B. Resnick, “Expression of the MAGE-A4 and NY-ESO-1 cancer-testis antigens in serous ovarian neoplasms,”Clinical Cancer Research, vol. 9, no. 17, pp. 6453–6460, 2003.

[43] A. D. Hislop, G. S. Taylor, D. Sauce, and A. B. Rickinson,“Cellular responses to viral infection in humans: lessons fromEpstein-Barr virus,” Annual Review of Immunology, vol. 25,pp. 587–617, 2007.

[44] H. Ji, A. M. Houghton, T. J. Mariani et al., “K-ras activationgenerates an inflammatory response in lung tumors,” Onco-gene, vol. 25, no. 14, pp. 2105–2112, 2006.

[45] E. Saeland, S. J. Van Vliet, M. Backstrom et al., “The C-type lectin MGL expressed by dendritic cells detects glycanchanges on MUC1 in colon carcinoma,” Cancer Immunology,Immunotherapy, vol. 56, no. 8, pp. 1225–1236, 2007.

[46] G. Zeng, M. E. Aldridge, X. Tian et al., “Dendritic cell surfacecalreticulin is a receptor for NY-ESO-1: direct interactionsbetween tumor-associated antigen and the innate immunesystem,” Journal of Immunology, vol. 177, no. 6, pp. 3582–3589, 2006.

[47] K. Imai, S. Matsuyama, S. Miyake, K. Suga, and K. Nakachi,“Natural cytotoxic activity of peripheral-blood lymphocytesand cancer incidence: an 11-year follow-up study of a generalpopulation,” The Lancet, vol. 356, no. 9244, pp. 1795–1799,2000.

[48] S. Roithmaier, A. M. Haydon, S. Loi et al., “Incidence ofmalignancies in heart and/or lung transplant recipients:a single-institution experience,” Journal of Heart and LungTransplantation, vol. 26, no. 8, pp. 845–849, 2007.

[49] C. Badoual, S. Hans, J. Rodriguez et al., “Prognostic value oftumor-infiltrating CD4+ T-cell subpopulations in head andneck cancers,” Clinical Cancer Research, vol. 12, no. 2, pp.465–472, 2006.

[50] K. Hiraoka, M. Miyamoto, Y. Cho et al., “Concurrentinfiltration by CD8+ T cells and CD4+ T cells is a favourableprognostic factor in non-small-cell lung carcinoma,” BritishJournal of Cancer, vol. 94, no. 2, pp. 275–280, 2006.

[51] M. Tomsova, B. Melichar, I. Sedlakova, and I. Steiner, “Prog-nostic significance of CD3+ tumor-infiltrating lymphocytesin ovarian carcinoma,” Gynecologic Oncology, vol. 108, no. 2,pp. 415–420, 2008.

[52] Y. C. Linn and K. M. Hui, “Cytokine-induced killer cells:NK-like T cells with cytotolytic specificity against leukemia,”Leukemia and Lymphoma, vol. 44, no. 9, pp. 1457–1462, 2003.

[53] P. H. Lu and R. S. Negrin, “A novel population of expandedhuman CD3CD56 cells derived from T cells with potentin vivo antitumor activity in mice with severe combinedimmunodeficiency,” Journal of Immunology, vol. 153, no. 4,pp. 1687–1696, 1994.

[54] T. Leemhuis, S. Wells, C. Scheffold, M. Edinger, and R.S. Negrin, “A phase I trial of autologous cytokine-inducedkiller cells for the treatment of relapsed Hodgkin diseaseand non-Hodgkin lymphoma,” Biology of Blood and MarrowTransplantation, vol. 11, no. 3, pp. 181–187, 2005.

[55] L. Zhang, J. R. Conejo-Garcia, D. Katsaros et al., “Intratu-moral T cells, recurrence, and survival in epithelial ovariancancer,” The New England Journal of Medicine, vol. 348, no. 3,pp. 203–213, 2003.

[56] Y. C. Nesbeth, D. G. Martinez, S. Toraya et al., “CD4+ T cellselicit host immune responses to MHC class II—ovarian can-cer through CCL5 secretion and CD4+0-mediated licensingof dendritic cells,” Journal of Immunology, vol. 184, no. 10,pp. 5654–5662, 2010.

[57] A. Bamias, V. Koutsoukou, E. Terpos et al., “Correlation ofNK T-like CD3+CD56+ cells and CD4+CD25+(hi) regulatoryT cells with VEGF and TNFα in ascites from advancedovarian cancer: association with platinum resistance andprognosis in patients receiving first-line, platinum-basedchemotherapy,” Gynecologic Oncology, vol. 108, no. 2, pp.421–427, 2008.

[58] M. Papamichail, S. A. Perez, A. D. Gritzapis, and C. N. Bax-evanis, “Natural killer lymphocytes: biology, development,and function,” Cancer Immunology, Immunotherapy, vol. 53,no. 3, pp. 176–186, 2004.

[59] S. H. Lim, A. C. Newland, S. Kelsey et al., “Continuousintravenous infusion of high-dose recombinant interleukin-2 for acute myeloid leukaemia—a phase II study,” CancerImmunology Immunotherapy, vol. 34, no. 5, pp. 337–342,1992.

[60] N. J. Meropol, G. M. Barresi, T. A. Fehniger, J. Hitt, M.Franklin, and M. A. Caligiuri, “Evaluation of natural killercell expansion and activation in vivo with daily subcutaneouslow-dose interleukin-2 plus periodic intermediate-dose puls-ing,” Cancer Immunology Immunotherapy, vol. 46, no. 6, pp.318–326, 1998.

[61] R. J. Soiffer, C. Murray, R. Gonin, and J. Ritz, “Effect of low-dose interleukin-2 on disease relapse after T-cell-depletedallogeneic bone marrow transplantation,” Blood, vol. 84, no.3, pp. 964–971, 1994.


[62] S. A. Rosenberg, M. T. Lotze, and L. M. Muul, “Observationson the systemic administration of autologous lymphokine-activated killer cells and recombinant interleukin-2 topatients with metastatic cancer,” The New England Journal ofMedicine, vol. 313, no. 23, pp. 1485–1492, 1985.

[63] J. A. A. Gubbels, M. Felder, S. Horibata et al., “MUC16provides immune protection by inhibiting synapse formationbetween NK and ovarian tumor cells,” Molecular Cancer, vol.9, article 11, 2010.

[64] M. Carlsten, N. K. Bjorkstrom, H. Norell et al., “DNAXaccessory molecule-1 mediated recognition of freshly isolatedovarian carcinoma by resting natural killer cells,” CancerResearch, vol. 67, no. 3, pp. 1317–1325, 2007.

[65] T. J. Webb, R. L. Giuntoli, O. Rogers, J. Schneck, andM. Oelke, “Ascites specific inhibition of CDId-mediatedactivation of natural Killer T Cells,” Clinical Cancer Research,vol. 14, no. 23, pp. 7652–7658, 2008.

[66] J. Hamanishi, M. Mandai, M. Iwasaki et al., “Programmedcell death 1 ligand 1 and tumor-infiltrating CD8+ T lym-phocytes are prognostic factors of human ovarian cancer,”Proceedings of the National Academy of Sciences of the UnitedStates of America, vol. 104, no. 9, pp. 3360–3365, 2007.

[67] L. Maddaluno, S. E. Verbrugge, C. Martinoli et al., “Theadhesion molecule L1 regulates transendothelial migrationand trafficking of dendritic cells,” Journal of ExperimentalMedicine, vol. 206, no. 3, pp. 623–635, 2009.

[68] S. Zecchini, M. Bianchi, N. Colombo et al., “The differentialrole of L1 in ovarian carcinoma and normal ovarian surfaceepithelium,” Cancer Research, vol. 68, no. 4, pp. 1110–1118,2008.

[69] E. Sato, S. H. Olson, J. Ahn et al., “Intraepithelial CD8+

tumor-infiltrating lymphocytes and a high CD8+/regulatoryT cell ratio are associated with favorable prognosis in ovariancancer,” Proceedings of the National Academy of Sciences of theUnited States of America, vol. 102, no. 51, pp. 18538–18543,2005.

[70] I. Kryczek, R. Liu, G. Wang et al., “FOXP3 defines regulatoryt cells in human tumor and autoimmune disease,” CancerResearch, vol. 69, no. 9, pp. 3995–4000, 2009.

[71] E. Y. Woo, C. S. Chu, T. J. Goletz et al., “RegulatoryCD4+CD25+ T cells in tumors from patients with early-stagenon-small cell lung cancer and late-stage ovarian cancer,”Cancer Research, vol. 61, no. 12, pp. 4766–4772, 2001.

[72] A. Bamias, M. L. Tsiatas, E. Kafantari et al., “Significant dif-ferences of lymphocytes isolated from ascites of patients withovarian cancer compared to blood and tumor lymphocytes.Association of CD3+CD56+ cells with platinum resistance,”Gynecologic Oncology, vol. 106, no. 1, pp. 75–81, 2007.

[73] T. J. Curiel, G. Coukos, L. Zou et al., “Specific recruitmentof regulatory T cells in ovarian carcinoma fosters immuneprivilege and predicts reduced survival,” Nature Medicine,vol. 10, no. 9, pp. 942–949, 2004.

[74] D. Wolf, A. M. Wolf, H. Rumpold et al., “The expressionof the regulatory T cell-specific forkhead box transcriptionfactor FoxP3 is associated with poor prognosis in ovariancancer,” Clinical Cancer Research, vol. 11, no. 23, pp. 8326–8331, 2005.

[75] M. E. C. Lutsiak, R. T. Semnani, R. De Pascalis, S. V.S. Kashmiri, J. Schlom, and H. Sabzevari, “Inhibition ofCD4+25+ T regulatory cell function implicated in enhancedimmune response by low-dose cyclophosphamide,” Blood,vol. 105, no. 7, pp. 2862–2868, 2005.

[76] S. A. Perez, M. V. Karamouzis, D. V. Skarlos et al.,“CD4+CD25+ regulatory T-cell frequency in HER-2/neu

(HER)-positive and HER-negative advanced-stage breastcancer patients,” Clinical Cancer Research, vol. 13, no. 9, pp.2714–2721, 2007.

[77] E. Y. Woo, C. S. Chu, T. J. Goletz et al., “RegulatoryCD4+CD25+ T cells in tumors from patients with early-stagenon-small cell lung cancer and late-stage ovarian cancer,”Cancer Research, vol. 61, no. 12, pp. 4766–4772, 2001.

[78] A. M. Ercolini, B. H. Ladle, E. A. Manning et al., “Recruit-ment of latent pools of high-avidity CD8+ T cells tothe antitumor immune response,” Journal of ExperimentalMedicine, vol. 201, no. 10, pp. 1591–1602, 2005.

[79] L. Leveque, F. Deknuydt, G. Bioley et al., “Interleukin 2-mediated conversion of ovarian cancer-associated CD4+

regulatory T cells into proinflammatory interleukin 17-producing helper T cells,” Journal of Immunotherapy, vol. 32,no. 2, pp. 101–108, 2009.

[80] I. Kryczek, M. Banerjee, P. Cheng et al., “Phenotype, distri-bution, generation, and functional and clinical relevance ofTh17 cells in the human tumor environments,” Blood, vol.114, no. 6, pp. 1141–1149, 2009.

[81] S. Wei, I. Kryczek, L. Zou et al., “Plasmacytoid dendriticcells induce CD8+ regulatory T cells in human ovariancarcinoma,” Cancer Research, vol. 65, no. 12, pp. 5020–5026,2005.

[82] A. H. Klimp, H. Hollema, C. Kempinga, A. G. J. vander Zee, E. G. E. De Vries, and T. Daemen, “Expressionof cyclooxygenase-2 and inducible nitric oxide synthase inhuman ovarian tumors and tumor-associated macrophages,”Cancer Research, vol. 61, no. 19, pp. 7305–7309, 2001.

[83] I. Kryczek, L. Zou, P. Rodriguez et al., “B7-H4 expres-sion identifies a novel suppressive macrophage populationin human ovarian carcinoma,” Journal of ExperimentalMedicine, vol. 203, no. 4, pp. 871–881, 2006.

[84] A. Mantovani, S. Sozzani, M. Locati, P. Allavena, and A. Sica,“Macrophage polarization: tumor-associated macrophagesas a paradigm for polarized M2 mononuclear phagocytes,”Trends in Immunology, vol. 23, no. 11, pp. 549–555, 2002.

[85] J. Wyckoff, W. Wang, E. Y. Lin et al., “A paracrine loopbetween tumor cells and macrophages is required for tumorcell migration in mammary tumors,” Cancer Research, vol.64, no. 19, pp. 7022–7029, 2004.

[86] R. Solans-Laque, C. Perez-Bocanegra, A. Salud-Salvia et al.,“Clinical significance of antinuclear antibodies in malignantdiseases: association with rheumatic and connective tissueparaneoplastic syndromes,” Lupus, vol. 13, no. 3, pp. 159–164, 2004.

[87] C. Sommer, A. Weishaupt, J. Brinkhoff et al., “Paraneoplasticstiff-person syndrome: passive transfer to rats by means ofIgG antibodies to amphiphysin,” The Lancet, vol. 365, no.9468, pp. 1406–1411, 2005.

[88] K. Balint, J. R. Conejo-Garcia, R. Buckanovich, and G.Coukos, “Role of vascular leukocytes in ovarian cancerneovascularization,” Advances in Experimental Medicine andBiology, vol. 622, pp. 273–280, 2008.

[89] L. A. Emens and E. M. Jaffee, “Leveraging the activityof tumor vaccines with cytotoxic chemotherapy,” CancerResearch, vol. 65, no. 18, pp. 8059–8064, 2005.

[90] F. Ghiringhelli, N. Larmonier, E. Schmitt et al., “CD4+CD25+

regulatory T cells suppress tumor immunity but are sen-sitive to cyclophosphamide which allows immunotherapyof established tumors to be curative,” European Journal ofImmunology, vol. 34, no. 2, pp. 336–344, 2004.

[91] E. Suzuki, V. Kapoor, A. S. Jassar, L. R. Kaiser, andS. M. Albelda, “Gemcitabine selectively eliminates splenic


Gr-1+/CD11b+ myeloid suppressor cells in tumor-bearinganimals and enhances antitumor immune activity,” ClinicalCancer Research, vol. 11, no. 18, pp. 6713–6721, 2005.

[92] P. Correale, M. G. Cusi, M. T. Del Vecchio et al., “Dendriticcell-mediated cross-presentation of antigens derived fromcolon carcinoma cells exposed to a highly cytotoxic mul-tidrug regimen with gemcitabine, oxaliplatin, 5-fluorouracil,and leucovorin, elicits a powerful human antigen-specificCTL response with antitumor activity in vitro,” Journal ofImmunology, vol. 175, no. 2, pp. 820–828, 2005.

[93] P. Correale, M. G. Cusi, K. Y. Tsang et al., “Chemoimmuno-therapy of metastatic colorectal carcinoma with gemcitabineplus FOLFOX 4 followed by subcutaneous granulocytemacrophage colony-stimulating factor and interleukin-2induces strong immunologic and antitumor activity inmetastatic colon cancer patients,” Journal of Clinical Oncol-ogy, vol. 23, no. 35, pp. 8950–8958, 2005.

[94] R. Chen, A. B. Alvero, D. A. Silasi, K. D. Steffensen, and G.Mor, “Cancers take their Toll - The function and regulationof Toll-like receptors in cancer cells,” Oncogene, vol. 27, no. 2,pp. 225–233, 2008.

[95] S. Romagnani, “The Th1/Th2 paradigm,” ImmunologyToday, vol. 18, no. 6, pp. 263–266, 1997.

[96] P. Pisa, E. Halapi, E. K. Pisa et al., “Selective expression ofinterleukin 10, interferon γ, and granulocyte- macrophagecolony-stimulating factor in ovarian cancer biopsies,” Pro-ceedings of the National Academy of Sciences of the UnitedStates of America, vol. 89, no. 16, pp. 7708–7712, 1992.

[97] L. L. Chen, F. Ye, W. G. Lu, Y. Yu, H. Z. Chen, andX. Xie, “Evaluation of immune inhibitory cytokine profilesin epithelial ovarian carcinoma,” Journal of Obstetrics andGynaecology Research, vol. 35, no. 2, pp. 212–218, 2009.

[98] A. G. Zeimet, M. Widschwendter, C. Knabbe et al., “Asciticinterleukin-12 is an independent prognostic factor in ovariancancer,” Journal of Clinical Oncology, vol. 16, no. 5, pp. 1861–1868, 1998.

[99] J. J.W. Chen, P. -L. Yao, A. Yuan et al., “Up-regulation oftumor interleukin-8 expression by infiltrating macrophages:its correlation with tumor angiogenesis and patient survivalin non-small cell lung cancer,” Clinical Cancer Research, vol.9, no. 2, pp. 729–737, 2003.

[100] J. G.W. Asschert, E. Vellenga, H. Hollema, A. G. J. vander Zee, and E. G.E. de Vries, “Expression of macrophagecolony-stimulating factor (M-CSF), interleukin-6 (IL-6),interleukin-1β (IL-1β), interleukin-11 (IL-11) and tumournecrosis factor-α (TNF-α) in p53-characterised human ovar-ian carcinomas,” European Journal of Cancer, vol. 33, no. 13,pp. 2246–2251, 1997.

[101] D. Milliken, C. Scotton, S. Raju, F. Balkwill, and J. Wilson,“Analysis of chemokines and chemokine receptor expressionin ovarian cancer ascites,” Clinical Cancer Research, vol. 8, no.4, pp. 1108–1114, 2002.

[102] R. Setoguchi, S. Hori, T. Takahashi, and S. Sakaguchi,“Homeostatic maintenance of natural Foxp3+ CD25+ CD4+

regulatory T cells by interleukin (IL)-2 and induction ofautoimmune disease by IL-2 neutralization,” Journal ofExperimental Medicine, vol. 201, no. 5, pp. 723–735, 2005.

[103] M. J. Lenardo, S. Boehme, L. Chen et al., “Automne feedbackdeath and the regulation of mature T lymphocyte antigenresponses,” International Reviews of Immunology, vol. 13, no.2, pp. 115–134, 1996.

[104] T. Kikuchi, Y. Akasaki, T. Abe et al., “Vaccination of gliomapatients with fusions of dendritic and glioma cells and

recombinant human interleukin 12,” Journal of Immunother-apy, vol. 27, no. 6, pp. 452–459, 2004.

[105] J. Dou, Y. Wang, J. Wang et al., “Antitumor efficacy inducedby human ovarian cancer cells secreting IL-21 alone orcombination with GM-CSF cytokines in nude mice model,”Immunobiology, vol. 214, no. 6, pp. 483–492, 2009.

[106] H. Nakano, T. Kishida, H. Asada et al., “Interleukin-21 triggers both cellular and humoral immune responsesleading to therapeutic antitumor effects against head andneck squamous cell carcinoma,” Journal of Gene Medicine,vol. 8, no. 1, pp. 90–99, 2006.

[107] T. Kusuda, K. Shigemasa, K. Arihiro, T. Fujii, N. Nagai, and K.Ohama, “Relative expression levels of Th1 and Th2 cytokinemRNA are independent prognostic factors in patients withovarian cancer,” Oncology Reports, vol. 13, no. 6, pp. 1153–1158, 2005.

[108] C. Tempfer, H. Zeisler, G. Sliutz, G. Haeusler, E. Hanzal,and C. Kainz, “Serum evaluation of interleukin 6 in ovariancancer patients,” Gynecologic Oncology, vol. 66, no. 1, pp. 27–30, 1997.

[109] G. Scambia, U. Testa, P. B. Panici et al., “Interleukin-6 serumlevels in patients with gynecological tumors,” InternationalJournal of Cancer, vol. 57, no. 3, pp. 318–323, 1994.

[110] C. Marth, A. G. Zeimet, M. Herold et al., “Different effectsof interferons, interleukin-1β and tumor necrosis factor-α innormal (OSE) and malignant human ovarian epithelial cells,”International Journal of Cancer, vol. 67, no. 6, pp. 826–830,1996.

[111] Z. Chen, A. Fadiel, Y. Feng, K. Ohtani, T. Rutherford,and F. Naftolin, “Ovarian epithelial carcinoma tyrosinephosphorylation, cell proliferation, and ezrin translocationare stimulated by interleukin 1α and epidermal growthfactor,” Cancer, vol. 92, no. 12, pp. 3068–3075, 2001.

[112] J. Rowley, A. Monie, C. -F. Hung, and T. -C. Wu, “Inhibitionof tumor growth by NK1.1+ cells and CD8+ T cells activatedby IL-15 through receptor β/common γ signaling in trans,”Journal of Immunology, vol. 181, no. 12, pp. 8237–8247, 2008.

[113] M. A. Kriegel, M. O. Li, S. Sanjabi, Y. Y. Wan, and R. A. Flavell,“Transforming growth factor-β: recent advances on its role inimmune tolerance,” Current Rheumatology Reports, vol. 8, no.2, pp. 138–144, 2006.

[114] G. C. Rodriguez, C. Haisley, J. Hurteau et al., “Regulation ofinvasion of epithelial ovarian cancer by transforming growthfactor-β,” Gynecologic Oncology, vol. 80, no. 2, pp. 245–253,2001.

[115] M. S. Pepper, “Transforming growth factor-beta: vasculoge-nesis, angiogenesis, and vessel wall integrity,” Cytokine andGrowth Factor Reviews, vol. 8, no. 1, pp. 21–43, 1997.

[116] S. Wu, C. M. Boyer, R. S. Whitaker et al., “Tumor necrosis fac-tor α as an autocrine and paracrine growth factor for ovariancancer: monokine induction of tumor cell proliferation andtumor necrosis factor α expression,” Cancer Research, vol. 53,no. 8, pp. 1939–1944, 1993.

[117] M. S. Naylor, G. W. H. Stamp, W. D. Foulkes, D. Eccles, andF. R. Balkwill, “Tumor necrosis factor and its receptors inhuman ovarian cancer: potential role in disease progression,”Journal of Clinical Investigation, vol. 91, no. 5, pp. 2194–2206,1993.

[118] H. Kulbe, R. Thompson, J. L. Wilson et al., “The inflam-matory cytokine tumour necrosis factor-alpha generates anautocrine tumor promoting network in epithelial ovariancancer cells,” Cancer Research, vol. 67, pp. 585–592, 2007.

[119] I. Goldstein, S. Ben-Horin, A. Koltakov et al., “α1β1 integrin+and regulatory Foxp3+ T cells constitute two functionally


distinct human CD4 + T cell subsets oppositely modulatedby TNFα blockade,” Journal of Immunology, vol. 178, no. 1,pp. 201–210, 2007.

[120] D. Milliken, C. Scotton, S. Raju, F. Balkwill, and J. Wilson,“Analysis of chemokines and chemokine receptor expressionin ovarian cancer ascites,” Clinical Cancer Research, vol. 8, no.4, pp. 1108–1114, 2002.

[121] R. P. M. Negus, G. W. H. Stamp, J. Hadley, and F. R.Balkwill, “Quantitative assessment of the leukocyte infiltratein ovarian cancer and its relationship to the expression of C-C chemokines,” American Journal of Pathology, vol. 150, no.5, pp. 1723–1734, 1997.

[122] S. Wei, I. Kryczek, L. Zou et al., “Plasmacytoid dendriticcells induce CD8+ regulatory T cells in human ovariancarcinoma,” Cancer Research, vol. 65, no. 12, pp. 5020–5026,2005.

[123] H. Nishikawa and S. Sakaguchi, “Regulatory T cells in tumorimmunity,” International Journal of Cancer, vol. 127, no. 4,pp. 759–767, 2010.

[124] D. Duluc, M. Corvaisier, S. Blanchard et al., “Interferon-γreverses the immunosuppressive and protumoral propertiesand prevents the generation of human tumor-associatedmacrophages,” International Journal of Cancer, vol. 125, no.2, pp. 367–373, 2009.

[125] A. Samsami Dehaghani, K. Shahriary, M. A. Kashef et al.,“Interleukin-18 gene promoter and serum level in womenwith ovarian cancer,” Molecular Biology Reports, vol. 36, no.8, pp. 2393–2397, 2009.

[126] K. Jaszczynska-Nowinka and A. Markowska, “New cytokine:stromal derived factor-1,” European Journal of GynaecologicalOncology, vol. 30, no. 2, pp. 124–127, 2009.

[127] C. Rudlowski, A.-K. Pickart, C. Fuhljahn et al., “Prognosticsignificance of vascular endothelial growth factor expressionin ovarian cancer patients: a long-term follow-up,” Interna-tional Journal of Gynecological Cancer, vol. 16, supplement 1,pp. 183–189, 2006.

[128] J. E. Ohm, D. I. Gabrilovich, G. D. Sempowski et al.,“Vascular endothelial growth factor C gene expression isclosely related to invasion phenotype in gynaecologicaltumor cells,” Blood, vol. 101, pp. 4878–4886, 2003.

[129] M. M. Dikov, J. E. Ohm, N. Ray et al., “Differential rolesof vascular endothelial growth factor receptors 1 and 2 indendritic cell differentiation,” Journal of Immunology, vol.174, no. 1, pp. 215–222, 2005.

[130] P. Lissoni, F. Malugani, A. Bonfanti et al., “Abnormallyenhanced blood concentrations of vascular endothelialgrowth factor (VEGF) in metastatic cancer patients and theirrelation to circulating dendritic cells, IL-12 and endothelin-1,” Journal of Biological Regulators and Homeostatic Agents,vol. 15, no. 2, pp. 140–144, 2001.

[131] X. B. Trinh, W. A. A. Tjalma, P. B. Vermeulen et al., “TheVEGF pathway and the AKT/mTOR/p70S6K1 signallingpathway in human epithelial ovarian cancer,” British Journalof Cancer, vol. 100, no. 6, pp. 971–978, 2009.

[132] L. J. Costa and H. A. Drabkin, “Renal cell carcinoma: newdevelopments in molecular biology and potential for targetedtherapies,” Oncologist, vol. 12, no. 12, pp. 1404–1415, 2007.

[133] A. Gaumann, H. J. Schlitt, and E. K. Geissler, “Immunosup-pression and tumor development in organ transplant recip-ients: the emerging dualistic role of rapamycin,” TransplantInternational, vol. 21, no. 3, pp. 207–217, 2008.

[134] H. Suzuki, H. Onishi, J. Wada et al., “VEGFR2 is selectivelyexpressed by FOXP3high CD4+ Treg,” European Journal ofImmunology, vol. 40, no. 1, pp. 197–203, 2010.

[135] C. A. Chen, W. F. Cheng, C. N. Lee et al., “Serum vascularendothelial growth factor in epithelial ovarian neoplasms:correlation with patient survival,” Gynecologic Oncology, vol.74, no. 2, pp. 235–240, 1999.

[136] A. A. Secord, R. Sayer, S. A. Snyder et al., “The relationshipbetween serum vascular endothelial growth factor, persistentdisease, and survival at second-look laparotomy in ovariancancer,” Gynecologic Oncology, vol. 94, no. 1, pp. 74–79, 2004.

[137] M. M. Vysotskii, M. A. Digaeva, N. E. Kushlinskii et al.,“Serum sFas, Leptin, and VEGF in patients with ovariancancer and benign tumors,” Bulletin of Experimental Biologyand Medicine, vol. 148, no. 5, pp. 810–814, 2009.

[138] B. C. Schultes and C. F. Nicodemus, “Using antibodiesin tumour immunotherapy,” Expert Opinion on BiologicalTherapy, vol. 4, no. 8, pp. 1265–1284, 2004.

[139] D. K. Niarchos, S. A. Perez, and M. Papamichail, “Character-ization of a novel cell penetrating peptide derived from Bag-1protein,” Peptides, vol. 27, no. 11, pp. 2661–2669, 2006.

[140] R. M. Sharkey and D. M. Goldenberg, “Targeted therapyof cancer: new prospects for antibodies and immunoconju-gates,” Ca-A Cancer Journal for Clinicians, vol. 56, no. 4, pp.226–243, 2006.

[141] J. A. Menendez, I. Mehmi, and R. Lupu, “Trastuzumabin combination with heregulin-activated Her-2 (erbB-2)triggers a receptor-enhanced chemosensitivity effect in theabsence of Her-2 overexpression,” Journal of Clinical Oncol-ogy, vol. 24, no. 23, pp. 3735–3746, 2006.

[142] J. S. Berek, O. Dorigo, B. Schultes, and C. Nicodemus, “Spe-cific keynote: immunological therapy for ovarian cancer,”Gynecologic Oncology, vol. 88, no. 1, pp. S105–S109, 2003.

[143] M. G. del Carmen, I. Rizvi, Y. Chang et al., “Synergismof epidermal growth factor-targeted immunotherapy withphotodynamic treatment of ovarian cancer in vivo,” Journalof the National Cancer Institute, vol. 97, pp. 1516–1524, 2005.

[144] C. Aghajanian, “The role of bevacizumab in ovarian cancer—an evolving story,” Gynecologic Oncology, vol. 102, no. 2, pp.131–133, 2006.

[145] B. J. Monk, E. Han, C. A. Josephs-Cowan, G. Pugmire, andR. A. Burger, “Salvage bevacizumab (rhuMAB VEGF)-basedtherapy after multiple prior cytotoxic regimens in advancedrefractory epithelial ovarian cancer,” Gynecologic Oncology,vol. 102, no. 2, pp. 140–144, 2006.

[146] R. A. Burger, M. F. Brady, M. A. Bookman et al., “PhaseIII trial of bevacizumab (BEV) in the primary treatmentof advanced Epithelial Ovarian Cancer (EOC), PrimaryPeritoneal Cancer (PPC), or Fallopian Tube Cancer (FTC):a Gynecologic Oncology Group study,” Journal of ClinicalOncology, vol. 28, Supplement 15, p. 5s, 2010.

[147] S. A. Cannistra, U. A. Matulonis, R. T. Penson et al.,“Phase II study of bevacizumab in patients with platinum-resistant ovarian cancer or peritoneal serous cancer,” Journalof Clinical Oncology, vol. 25, no. 33, pp. 5180–5186, 2007.

[148] T. Perren, A. M. Swart, J. Pfisterer et al., “ICON7: a phaseIII randomized Gynaecologic Cancer Intergroup trial ofconcurrent bevacizumab and chemotherapy followed bymaintenance bevacizumab, verus chemotherapy alone inwomen with newly diagnosed epithelial ovarian (EOC),primary peritoneal (PPC) or fallopian tube cancer (FTC),” inProceedings of the 35th ESMO Congress, October 2010, LBA4.

[149] V. J. Mobus, R. P. Baum, M. Bolle et al., “Immune responsesto murine monoclonal antibody-B43.13 correlate with pro-longed survival of women with recurrent ovarian cancer,”American Journal of Obstetrics and Gynecology, vol. 189, no.1, pp. 28–36, 2003.


[150] J. S. Berek, P. T. Taylor, A. Gordon et al., “Randomized,placebo-controlled study of oregovomab for consolidation ofclinical remission in patients with advanced ovarian cancer,”Journal of Clinical Oncology, vol. 22, no. 17, pp. 3507–3516,2004.

[151] M. S. Gordon, D. Matei, C. Aghajanian et al., “Clinicalactivity of pertuzumab (rhuMAb 2C4), a HER dimerizationinhibitor, in advanced ovarian cancer: potential predictiverelationship with tumor HER2 activation status,” Journal ofClinical Oncology, vol. 24, no. 26, pp. 4324–4332, 2006.

[152] J. Konner, R. J. Schilder, F. A. DeRosa et al., “A phase II studyof cetuximab/paclitaxel/carboplatin for the initial treatmentof advanced-stage ovarian, primary peritoneal, or fallopiantube cancer,” Gynecologic Oncology, vol. 110, no. 2, pp. 140–145, 2008.

[153] M. V. Seiden, H. A. Burris, U. Matulonis et al., “A phase IItrial of EMD72000 (matuzumab), a humanized anti-EGFRmonoclonal antibody, in patients with platinum-resistantovarian and primary peritoneal malignancies,” GynecologicOncology, vol. 104, no. 3, pp. 727–731, 2007.

[154] R. A. Burger, “Experience with bevacizumab in the man-agement of epithelial ovarian cancer,” Journal of ClinicalOncology, vol. 25, no. 20, pp. 2902–2908, 2007.

[155] L. Hu, J. Hofmann, J. Holash, G. D. Yancopoulos, A. K.Sood, and R. B. Jaffe, “Vascular endothelial growth factortrap combined with paclitaxel strikingly inhibits tumor andascites, prolonging survival in a human ovarian cancermodel,” Clinical Cancer Research, vol. 11, no. 19, pp. 6966–6971, 2005.

[156] B. Li, A. S. Lalani, T. C. Harding et al., “Vascular endothelialgrowth factor blockade reduces intratumoral regulatoryT cells and enhances the efficacy of a GM-CSF-secretingcancer immunotherapy,” Clinical Cancer Research, vol. 12,no. 22, pp. 6808–6816, 2006.

[157] K. M. Schmeler, S. Vadhan-Raj, P. T. Ramirez et al., “A phaseII study of GM-CSF and rIFN-γ1b plus carboplatin for thetreatment of recurrent, platinum-sensitive ovarian, fallopiantube and primary peritoneal cancer,” Gynecologic Oncology,vol. 113, no. 2, pp. 210–215, 2009.

[158] G. H. Windbichler, H. Hausmaninger, W. Stummvoll etal., “Interferon-gamma in the first-line therapy of ovariancancer: a randomized phase III trial,” British Journal ofCancer, vol. 82, no. 6, pp. 1138–1144, 2000.

[159] G. D. Hall, J. M. Brown, R. E. Coleman et al., “Maintenancetreatment with interferon for advanced ovarian cancer:results of the Northern and Yorkshire gynaecology grouprandomised phase III study,” British Journal of Cancer, vol.91, no. 4, pp. 621–626, 2004.

[160] D. S. Alberts, C. Marth, R. D. Alvarez et al., “Randomizedphase 3 trial of interferon gamma-1b plus standard carbo-platin/paclitaxel versus carboplatin/paclitaxel alone for first-line treatment of advanced ovarian and primary peritonealcarcinomas: results from a prospectively designed analysisof progression-free survival,” Gynecologic Oncology, vol. 109,no. 2, pp. 174–181, 2008.

[161] A. Ribas, L. H. Butterfield, J. A. Glaspy, and J. S. Economou,“Current developments in cancer vaccines and cellularimmunotherapy,” Journal of Clinical Oncology, vol. 21, no. 12,pp. 2415–2432, 2003.

[162] U. Wagner, S. Kohler, S. Reinartz et al., “Immuno-logical consolidation of ovarian carcinoma recurrenceswith monoclonal anti-idiotype antibody ACA125: immuneresponses and survival in palliative treatment,” ClinicalCancer Research, vol. 7, no. 5, pp. 1154–1162, 2001.

[163] D. H. Schuurhuis, N. Van Montfoort, A. Ioan-Facsinay et al.,“Immune complex-loaded dendritic cells are superior tosoluble immune complexes as antitumor vaccine,” Journal ofImmunology, vol. 176, no. 8, pp. 4573–4580, 2006.

[164] I. Fricke and D. I. Gabrilovich, “Dendritic cells and tumormicroenvironment: a dangerous liaison,” ImmunologicalInvestigations, vol. 35, no. 3-4, pp. 459–483, 2006.

[165] J. W. Hodge, K. Y. Tsang, D. J. Poole, and J. Schlom, “Generalkeynote: vaccine strategies for the therapy of ovarian cancer,”Gynecologic Oncology, vol. 88, no. 1, pp. S97–S104, 2003,Discussion pp. S103–S110.

[166] C. L. Slingluff and D. E. Speiser, “Progress and controversiesin developing cancer vaccines,” Journal of TranslationalMedicine, vol. 3, article no. 18, 2005.

[167] M. E. Dudley and S. A. Rosenberg, “Adoptive-cell-transfertherapy for the treatment of patients with cancer,” NatureReviews Cancer, vol. 3, no. 9, pp. 666–675, 2003.

[168] A. de Gramont, D. Gangji, C. Louvet, M. L. Garcia, D. Tardy,and J. L. Romet-Lemonne, “Adoptive immunotherapy ofovarian carcinoma,” Gynecologic Oncology, vol. 86, no. 1, pp.102–103, 2002.

[169] M. H. Kershaw, J. A. Westwood, L. L. Parker et al., “A phaseI study on adoptive immunotherapy using gene-modifiedT cells for ovarian cancer,” Clinical Cancer Research, vol. 12,no. 20, pp. 6106–6115, 2006.

[170] D. Ritchie, L. Mileshkin, D. Wall et al., “In vivo tracking ofmacrophage activated killer cells to sites of metastatic ovariancarcinoma,” Cancer Immunology, Immunotherapy, vol. 56, no.2, pp. 155–163, 2007.

[171] A. Barber, A. Rynda, and C. L. Sentman, “Chimeric NKG2Dexpressing T cells eliminate immunosuppression and activateimmunity within the ovarian tumor microenvironment,”Journal of Immunology, vol. 183, no. 11, pp. 6939–6947, 2009.

[172] M. De Cesare, L. Sfondrini, M. Campiglio et al., “Ascitesregression and survival increase in mice bearing advanced-stage human ovarian carcinomas and repeatedly treatedintraperitoneally with CpG-ODN,” Journal of Immunother-apy, vol. 33, no. 1, pp. 8–15, 2010.

[173] T. J. Curiel, B. Barnett, I. Krysczek, P. Cheng, and W. Zou,“Regulatory T cells in ovarian cancer: biology and therapeu-tic potential,” Cancer Immunity, vol. 6, pp. 20–21, 2006.


Research Article

Engagement of the Mannose Receptor by TumoralMucins Activates an Immune Suppressive Phenotype inHuman Tumor-Associated Macrophages

P. Allavena,1 M. Chieppa,2, 3 G. Bianchi,4 G. Solinas,1, 5 M. Fabbri,6

G. Laskarin,7 and A. Mantovani1, 8

1 Deptartment Immunology & Inflammation, IRCCS Clinical Institute Humanitas, Rozzano, 20089 Milan, Italy2 Department of Translational Medicine, National Institute of Gastroenterology IRCCS “Saverio de Bellis”, Castellana Grotte,70013 Bari, Italy

3 Advanced Research Centre for Health (ARCHES), Enviroment and Space, Castellana Grotte, 70013 Bari, Italy4 Environmental Health Sciences Department, Mario Negri Institute, 20157 Milano, Italy5 Cardiovascular Research Institute, University of California, San Francisco, CA 94158, USA6 European Commission, Joint Research Centre Institute for Health and Consumer Protection Molecular Biology and Genomics,21020 (VA) Ispra, Italy

7 Department of Physiology and Immunology, University of Rijeka, 51000 Rijeka, Croatia8 Department of Translational Medicine, University of Milan, 20121 Milan, Italy

Correspondence should be addressed to P. Allavena, [email protected]

Received 2 August 2010; Revised 18 October 2010; Accepted 21 December 2010

Academic Editor: Nima Rezaei

Copyright © 2010 P. Allavena et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Tumor-Associated Macrophages (TAMs) are abundantly present in the stroma of solid tumors and modulate several importantbiological processes, such as neoangiogenesis, cancer cell proliferation and invasion, and suppression of adaptive immuneresponses. Myeloid C-type lectin receptors (CLRs) constitute a large family of transmembrane carbohydrate-binding receptorsthat recognize pathogens as well as endogenous glycoproteins. Several lines of evidence demonstrate that some CLRs can inhibitthe immune response. In this study we investigated TAM-associated molecules potentially involved in their immune suppressiveactivity. We found that TAMs isolated from human ovarian carcinoma samples predominantly express the CLRs Dectin-1, MDL-1,MGL, DCIR, and most abundantly the Mannose Receptor (MR). Components of carcinomatous ascites and purified tumoralmucins (CA125 and TAG-72) bound the MR and induced its internalization. MR engagement by tumoral mucins and by an agonistanti-MR antibody modulated cytokine production by TAM toward an immune-suppressive profile: increase of IL-10, absence ofIL-12, and decrease of the Th1-attracting chemokine CCL3. This study highlights that tumoral mucin-mediated ligation of theMR on infiltrating TAM may contribute to their immune suppressive phenotype.

1. Introduction

Among tumor-infiltrating leukocytes, Tumor-AssociatedMacrophages (TAMs) constitute a major subset [1–3]. Whilethe presence of T lymphocytes in tumor stroma is usuallycorrelated with more favourable prognosis of cancer patients[4, 5], in most studies the density of TAM is associated withrapid tumor progression [6–9]. TAMs are poorly cytotoxicagainst neoplastic cells and may actually favour tumor cellsurvival and proliferation by actively producing growth

factors for cancer and endothelial cells. They are also amajor source of proteolytic enzymes that degrade the extra-cellular matrix thus favouring the invasion of neoplastic cells[9, 10]. Further, TAM contributes to the evasion of tumorsfrom immune control by producing immune-suppressivecytokines such as IL-10 and TGFβ [2, 9].

Our group proposed that TAMs are M2-like polar-ized macrophages [11]. Along a conventional definition,macrophages activated in the presence of inflammatorymediators (e.g., LPS) and Th1 cytokines (e.g., IFNγ) are


defined as M1 or classically activated macrophages. Theseeffectors have high cytotoxic functions, produce immune-stimulatory cytokines, and are important cells for the defenseagainst intracellular pathogens and transformed cells [11–13]. On the other hand M2 macrophages, or alternativelyactivated macrophages, differentiate in microenvironmentsrich in anti-inflammatory mediators or Th2 cytokines (e.g.,IL-10, IL-4, IL-13); they have high scavenging activity, pro-duce several growth factors, activating the process of tissueregeneration, and suppress adaptive immune responses [2,12, 14, 15]. While these activities are of extreme importanceduring wound healing to return to the homeostatic state, inthe context of a growing tumor they are deleterious to thehost. Indeed several studies have found a strong correlationbetween high numbers of TAM, number of vessels, and lowerdisease-free survival [9, 10, 16–18].

In this study we investigated TAM-related mechanismsof immune escape and considered C-type lectin receptors(CLRs) as interesting candidates. CLRs are a large familyof structurally related transmembrane receptors which, byvirtue of their carbohydrate-recognition domains, bind highaffinity sugar moieties present on the surface of pathogensas well as of endogenous glycoproteins [19, 20]. Togetherwith the Toll-like receptor (TLR) family, CLRs expressed onmyeloid cells of the innate immunity constitute the majorsystem to sense the outer world [21–23]. Myeloid CLRs aresubdivided into two major families: the first including themannose receptor (MR, DEC206), ENDO180, DEC205, andPLA2 receptor, and the second including DC-SIGN, Dectin-1, Langerin, DCIR, MGL, and BDAC2 [24, 25].

Although the majority of studies have investigatedthe role of CLRs in the recognition, internalization, andclearance of pathogens through activation of innate immu-nity, other studies have clearly demonstrated that at leastsome receptors elicit anti-inflammatory/immune-suppres-sive responses, raising the hypothesis that pathogens mayexploit CLR binding and internalization ability to overcomeinnate immunity and survive within the host [22, 24, 25].For instance, M. tuberculosis binds DC-SIGN and the MRand inhibits the production of IL-12 [26–28]. We previouslyreported in mono-DC that cross-linking of the MR withan agonist antibody increased IL-10 production leadingto inhibition of IL-12 and defective Th1 differentiation[27]. Activation of BDAC2, expressed by plasmacytoid DC,as well as DCIR, downregulates the production of IFNα[29, 30]. DCIR also interferes with GM-CSF signalling[31]. The receptor MGL recognizes the isoform CD45RBexpressed by effector memory T cells and negatively influ-ences T cell receptor signalling [32]. Thus, several linesof evidence point to a role of at least some CLRs inthe restriction of inflammatory reactions and in homeo-stasis preservation [24, 25].

Of interest, CLRs recognize glycans expressed also onendogenous ligands. For instance, the carbohydrate sialylLewisX -type expressed on lymphatic endothelium is recog-nized by the MR and DC-SIGN [24, 33]. The latter alsobinds vascular adhesion molecules [23]. The MR recog-nizes selected hormones (thyroglobulin, luteotropin), matrixmolecules (chondroitin sulphate proteoglycans, collagen),

and enzymes (myeloperoxidase, lysosomal hydrolases) [20].DC-SIGN, MGL, and MR bind epithelial mucins [24, 27, 34–36]. The physiological significance of the recognition ofendogenous ligands by CLR is not fully characterized.

Previous studies on CLRs have been mainly performedwith in vitro generated macrophages and DC or with in vivomouse models of diseases. The aim of this study is to inves-tigate the expression of CLRs in TAM. Here we show thatTAM isolated from human ovarian carcinoma samples pre-dominantly expressed Dectin-1, MDL-1, MGL, DCIR, andmost abundantly the MR. Experiments demonstrated thatthe MR recognizes ligands present in carcinomatous ascitesand tumoral mucins such as CA125 and tumor-associatedglycoprotein- (TAG-) 72. Upon mucin-engagement of theMR, TAMs secrete higher levels of IL-10 and lower levelsof the T cell attracting chemokine CCL3. Thus, tumoralmucin-mediated activation of the MR on TAM triggers animmune-suppressive response which likely contributes totumor immune evasion.


2.1. Isolation of Human Tumor-Associated Macrophages(TAMs). Having obtained an informed consent, we col-lected carcinomatous ascites and/or tumor samples from27 patients with histologically confirmed ovarian tumors.TAMs were isolated from ascites by density Ficoll, and Percollgradients (Lonza, Italy) as described in [37]; TAMs fromsolid tumors were isolated by enzymatic digestion and Ficollgradient [38] and were further purified by plastic adherence(RPM1 1640 w/o FBS, 1h, 37◦C); adherent cells were usually80–90% CD68+ macrophages as assessed by flow cytometry.

Human in vitro differentiated macrophages were ob-tained by culture of monocytes with M-CSF (20 ng/mL)[37] for 6 days [37]. Myeloid DCs were differentiated frommonocytes with GM-CSF (50 ng/mL) and IL-4 (20 ng/mL)for 6 days [27].

2.2. Transcriptional Profile Analysis. TAMs from 7 differentpatients (5 from carcinomatous ascites and 2 from solidtumors) were used for the transcriptional profiling exper-iments. TAMs were either immediately processed or after18-h stimulation with LPS (100 ng/mL) (Sigma, Italy) or IL-10 (20 ng/mL) (Peprotech,Italy) (for 4 TAM preparations).Macrophages from the peritoneal free-fluid of nontumoralpatients (ovarian cysts) were collected during surgery from12 different patients, centrifuged over Ficoll and immediatelyprocessed for RNA (purity >90%). Total RNA was extractedfrom 5×106 cells using Trizol (Invitrogen Life Technologies),retrotranscribed and prepared for GeneChip hybridization aspreviously described [38]. Each TAM preparation was inde-pendently tested. Macrophages from nontumoral patientswere pooled to reach the minimum necessary amount for1 GeneChip. Fragmented cRNA was hybridized to HG-U133 Plus 2.0 GeneChips (Affymetrix) and then washed andscanned according to manufacturer’s guidelines. Expressionmeasures were computed using Robust Multiarray Average(RMA). Statistical differences were assessed by a moderated


t-test analysis performed using a Limma bioconductorpackage, and resulting P-values were adjusted using theBenjamini and Hochberg step-up method for controlling theFalse Discovery Rate (FDR). Genes were defined as regulatedwhen characterized by a fold of induction≥2 and an FDR P-value≤.05. Computations were conducted using the R statis-tics programming environment (http://www.r-project.org/).

2.3. Phenotype Analysis. Tumor macrophages were analysedby flow cytometry on FACS Canto (BD Bioscience, Milan,Italy). Cells were first incubated with PBS 1% HS (30 minutes4◦C) to block FcγR, washed and resuspended in FACS buffer(PBS 0.5% BSA, 0.05% NaN3). PE-mouse antihuman CD14(clone M5E2) was purchased from (BD Pharmingen, Italy).Three of mouse anti-human MR/CD206 were used withidentical results. Clone 19.2 was from BD Pharmingen; clonePAM-1 was previously characterized [27]; clone WE458 wasin-house generated and selected for reactivity against MR-transfected CHO cells.

2.4. Endocytosis Assay. Mannose receptor–mediated endocy-tosis was measured as the cellular uptake of FITC-dextranand quantified by flow cytometry. Approximately 2 × 105

cells per sample were incubated in media containing FITC-dextran (1 mg/mL) (molecular weight 40,000; Sigma) overa period of 60 min. After incubation, cells were washedtwice with phosphate-buffered saline (PBS) to remove excessdextran and fixed in cold 1% formalin. Endocytosis wasexpressed as fluorescence intensity, calculated as mean fluo-rescence intensity of positive cells at 37◦C- mean fluorescenceintensity of positive cells at 4◦C.

2.5. Immunohistochemistry. Human surgical samples ofovarian tumors were immediately frozen in OCT aftersurgical collection. Sections were stained with anti-CD206mAbs, followed by a goat antimouse secondary antibody(EnVision horseradish peroxidase rabbit/mouse, DakoCy-tomation). After a diaminobenzidine reaction (Liquid DAB+ Substrate Chromogen System, Dako Cytomation), sectionswere counterstained with hematoxylin (Mayer, DIAPATH).

2.6. Elisa. Cytokines were measured in supernatants of TAM,macrophages, and Dendritic Cells (DCs) by commerciallyavailable ELISA kits (IL-10, IL-12, CCL3) according tomanufacture’s instructions (R&D Systems, Space Import,Milan, Italy). Cells were pretreated (10 min. room tem-perature) with anti-CD206 (clone WE458, 2 ug/mL), ortumoral mucins (Sigma) TAG-72 (200 UI/mL) or CA125(200 UI/mL), prior to stimulation with LPS (1 ug/mL,24 hours). TAG-72 and CA125 contained less than 0.125endotoxin unit/mL as checked by Limulus amebocyte lysateassay (BioWhittaker, Walkersville, MD).

2.7. Statistical Analysis. Prism software (GraphPad) andMicrosoft Excel were used for all statistical analyses. Student’st tests were used to determine statistically significant differ-ences between experimental groups. P < .05 was consideredto be statistically significant.

Table 1: Affymetrix Gene Expression analysis of selected C-typelectin receptors in human Tumor-Associated Macrophages (TAMs).

Gene symbol/other names

Intensity◦Modulation§

LPS/IFNγModulation§

IL-10

MRC1/CD206CLEC13D

562 ± 76 0,1∗ 1,5∗

Dectin-1CLEC7A

482 ± 63 0,3∗ 1,4∗

MDL-1CLEC5A

468 ± 94 0,3∗ 0,7

DCIR CLEC4A 359 ± 37 0,7 1,1

MGL-1/CD301CLEC10A

258 ± 85 0,9 1,1

DCL-1/CD302CLEC13A

160 ± 42 0.9 1

ENDO-180/CD280CLEC13E

111 ± 72 0,8 1

DEC-205/CD205CLEC13B

77 ± 42 2,9∗ 0,7

DC-SIGN/CD209CLEC4L

40 ± 13 0,8 0,9

Langerin/CD207CLEC4K

42 ± 11 0,9 0,9

PLA2RCLEC13C

25 ± 6 1 1

◦Normalized intensity, §Fold over untreated TAM. Results are expressed asmedian values ± SE of 7 different TAM preparations and are presented asnormalized intensity signals: modulation of CLR genes after TAM treatmentwith LPS/IFNγor IL-10 for 18 hrs. Results are presented as fold overuntreated TAM, ∗P < .05 versus untreated. Experiment of CLR modulationby cytokines was performed in 4 TAM preparations.

3. Results

3.1. C-Type Lectin Receptor Gene Expression in Human TAM.To study the expression of CLRs in TAM we interrogated ourAffymetrix database performed with 7 different populationsof purified TAM isolated from human ovarian carcinoma (5from carcinomatous ascites and 2 from solid tumors). CLRgene expression levels from TAM of solid tumors or fromascitic fluids were similar and were considered together. Themost expressed CLR genes were the mannose receptor (Mrc1,CD206), Dectin1, DCIR MDL-1, and MGL-1 (Table 1). OtherCLR genes were expressed at very low level (e.g., DEC205,DC-SIGN, PLA2R).

Modulation of CLR expression by LPS/IFNγ or IL-10 wasperformed in 4 TAM samples. Exposure of TAM to LPS/IFNγinduced a different gene modulation with a prominentincrease of DEC205 (2.9-fold) and strong decrease of Mrc1(0.1) and of MDL-1 (0.3). In contrast, pretreatment withIL-10 upregulated Mrc1 by 1.5-fold (Table 1). It was ofinterest to compare the gene expression analysis of TAMwith normal tissue macrophages. We had the opportunityto test peritoneal free-fluid macrophages collected from


MRC123%

MDL-117%

MRC115%

DCIR14%

Dectin-119%

MGL-110%

DCL-110%

Dectin-131%

DCIR21% MDL-1

6%

MGL-13%

DCL-16%

Tumor macrophages Tissue macrophages

Figure 1: Schematic representation of the relative expression of the eleven CLRs shown in Table 1, in tumoral macrophages (TAM) fromovarian tumor samples and in nontumoral macrophages isolated from the peritoneal free-fluid of patients with benign diseases (ovariancysts).

nontumoral patients. As the amount of free-fluid andthe cellular content is usually very small, we pooled thesamples from 12 different subjects who underwent surgeryfor nonneoplastic diseases (ovarian cysts) and analyzedwith the same GeneChips (Affymetrix) used with TAM.Figure 1 shows that the relative expression of the elevenCLRs analyzed was similar between TAM and normal tissuemacrophages, though some differences were noted: Mrc1,MDL-1, and MGL-1 were higher in TAM, while Dectin1 andDCIR and DCL-1 were higher in normal macrophages.

On the basis of these findings we further investigated themannose receptor (MR) in human TAM.

3.2. Phenotype and Functional Activity of the MR in HumanTAM. To check for protein expression, 12 different prepa-rations of human TAM were purified from the ascitic fluidof patients with ovarian carcinoma and tested by flowcytometry. Figure 2(a) shows the results of each individualpreparation as percentage of MR (CD206) and of CD14,used as a pan-myeloid marker. MR expression was variableand ranged from 17% to 72% (median value 39%). Suchheterogeneity is likely due to the fact that the MR is anendocytic receptor that continuously shuttles from the cellmembrane to the early endosome compartments.

Immunohistochemistry of surgical samples of humanovarian cancer was performed with two different anti-CD206mAb. Macrophages infiltrating the tumor stroma showedstrong reactivity (Figure 2(b)); these results confirmed thatthe MR is expressed both by ascitic fluid macrophages andby TAM infiltrating solid tumors.

To evaluate MR ability to internalize soluble particles,we incubated TAM with FITC-Dextran, a known ligandof MR. TAM rapidly internalized FITC-Dextran over 60′

period, with a kinetic similar to that of normal macrophagesdifferentiated in vitro with M-CSF (Figure 3(a)). Receptorspecificity was checked by pretreating cells with a blockinganti-MR mAb, which resulted in significant inhibition ofinternalization (Figure 3(b)). Pretreatment of TAM withtumoral ascites (33% v/v) reduced by 50% FITC-Dextran

0

20

40

60

80

100

1 2 3 4 5 6 7 8 9 10 11 12

TAM

Posi

tive

cells

(%)

CD206CD14

(a)

Ovarian tumor 1 Ovarian tumor 2

Anti-CD206(Pam-1)

Anti-CD206(WE458)

(b)

Figure 2: Expression of the Mannose Receptor (MR) by humanTAM. (a) Flow cytometry analysis of purified preparations ofTAM from carcinoumatous ascites of patients with ovarian cancer.TAMs were stained with anti-CD206 mAb (clone PAM-1) orwith anti-CD14. Twelve different preparations were tested. (b)Immunohistochemistry of 2 tumor samples from ovarian cancertissues stained with anti-CD206 mAbs (clone PAM-1, upper panels;clone WE458, lower panels). Positive cells are brown stained(magnification: left panels 40 x, right panels 100 x).


0

200

400

600

800

10 20 30 40 50 60

Time (min)

FIT

C-d

extr

anu

ptak

e(M

FI)

TAM 1Macro

(a)

FIT

C-d

extr

anco

ntr

ol(%

)

0

20

40

60

80

100

Medium Dextran Anti-CD206 Ascites

P < .05

(b)

Figure 3: TAMs express a functional endocytic MR. (a) Endocytosisof FITC-Dextran by TAM (triangle) and by in vitro M-CSF-differentiated macrophages (square), evaluated as Mean Fluores-cence Intensity (MFI) by flow cytometry. Shown is a representativeexperiment of 4 performed. (b) FITC-Dextran endocytosis in TAMis significantly inhibited (P < .05 Student’s t tests) by pretreatmentwith MR-ligands: unconjugated Dextran (1 mg/mL), anti-CD206mAb (10 ug/mL); and 33%v/v cell-free ascitic fluid from ovariantumors. Results are expressed as % relative to values of FITC-Dextran uptake in control cells (medium) and are the mean +/−SD of 3 experiments with 3 different TAM preparations.

endocytosis, suggesting that ascitic fluids contained putativeligand(s) of the MR. To have further proof of this, weincubated TAM with ascitic fluids prior to staining with anti-CD206 mAb and analyzed in flow cytometry. Figure 4(a)shows that tumoral ascites did induce the internalizationof the MR from the surface of TAM and of normalmacrophages: the percentage of surface MR decreased by 60–80% while that of CD14 was unaffected (Figure 4(b)).

It is established that the MR can bind to MUC1mucin and to the tumoral mucin TAG-72 [35, 36, 39, 40].We therefore tested the mucin CA125 that is specificallyassociated with ovarian cancer. Pretreatment of TAM withTAG-72 or CA125 decreased MR expression, indicating thatalso CA125 is able to engage the MR and to induce itsinternalization (Figure 4(c)).

Notably, both unconjugated Dextran and the other MRligands tested did not completely block endocytosis or inhibitreceptor expression, most likely because—as mentioned

above—MR has high recycling ability. In support of this,we noticed that in vitro culture of TAM for 24 hours inthe absence of ascitic fluid (i.e., out of the original micro-environment), resulted in higher MR levels compared toTAMs that were immediately tested after isolation (notshown).

3.3. Mucin-Mediated Ligation of the MR Modulates CytokineProduction in Human TAM. We previously reported in den-dritic cells (DCs) that activation of the MR with an agonistmAb or with MUC1 induced a regulatory/immunosuppres-sive phenotype with a switch of cytokine production charac-terized by low IL-12 and high IL-10 [27]. Hence, we testedcytokine production of mucin-treated TAM. Figure 5(a)shows that all tested MR ligands (TAG-72, CA125, and anti-CD206) induced a significant increase of IL-10 in TAM aswell as in normal macrophages. By contrast, IL-12 secretionwas strongly decreased in normal macrophages (Figure 5(a)).TAMs, as already reported [41], are unable to produceIL-12 even after optimal stimulation with LPS and IFNγ(Figure 5(a)).

Macrophages and DC are a major source of chemokineswhich importantly amplify the immunological network byrecruiting immunocompetent cells at tumor tissues. Weinvestigated the production of the chemokine CCL3, whichrecruits Th1 and cytotoxic effector lymphocytes. Figure 5(b)shows that TAG-72 mucin strongly inhibited the secretion ofCCL3 by TAM and by in vitro generated macrophages andmono-DC.

Overall these results demonstrate that the MR expressedby TAM recognizes endogenous ligands present in the tumormicroenvironment, including the ovarian cancer specificmucin CA125. Mucin-induced MR engagement modu-lates the cytokine secretion of TAM toward an immune-suppressive phenotype: increase of IL-10, absence of IL-12 and decrease of CCL3. This cytokine profile is likely tocontribute to tumor immune escape.

4. Discussion

Very little is known about the expression and functionalrole of CLRs in myeloid cells infiltrating tumors. Thisstudy demonstrates that human TAMs express a numberof CLRs (e.g., Dectin-1, MDL-1, MGL, DCIR) and mostprominently the MR/CD206. Other receptors were notsignificantly expressed (e.g., DC-SIGN, DEC205, Langerin),in line with their preferential localization on dendritic andLangerhans cells. ENDO180, which shares similarities withthe MR, was also poorly represented, and this finding isconsistent with its higher expression in fibroblasts [42]. Wefocused our attention on the MR. Although it has longbeen known that TAMs bear this receptor—and actually thisevidence served as paradigm of their M2-like polarization—no functional characterization of MR-positive TAM has everbeen provided.

We found that both TAMs from solid tumors andthose from the ascitic fluid associated to advanced ovariancancer, have high membrane expression of the MR. Levels of


Medium33% ascites

0

20

40

60

80

100Po

siti

vece

lls(%

)

TAM 1 TAM 2 Macro 1 Macro 2

CD206

(a)

Medium33% ascites

0

20

40

60

80

100

Posi

tive

cells

(%)

TAM 1 TAM 2 Macro 1 Macro 2

CD14

(b)

CD

206

(MFI

)co

ntr

ol(%

)

Medium Dextran TAG72200 IU/mL

CA12520 IU/mL

CA125200 IU/mL

P < .05

0

20

40

60

80

100

(c)

Figure 4: Tumoral mucins induce the internalization of the MR: flow cytometry expression of the MR /CD206 (a) and CD14 (b). Twodifferent TAM preparations (TAM1 and TAM2) and 2 M-CSF-differentiated normal macrophages (Macro1 and Macro2) were pretreated (30min. room temperature) with 33% v/v ascitic fluid from ovarian tumor patients, prior to staining with anti-CD206 or CD14 mAbs. Resultsare shown as % of positive cells. (c) Purified TAMs were pretreated with unconjugated Dextran (1 mg/mL); mucin Tag-72 (200 IU/mL);mucin CA125 (20–200 IU/mL) prior to staining with anti-CD206 mAb. Results are shown as % relative to values of Mean fluorescenceIntensity (MFI) of CD206 in control cells (medium) and are the mean +/− SD of 4 experiments with 4 different TAM preparations (3 TAMpreparations for CA125). P < .05 (Student’s t-tests).

expression were modulated by the tumor microenvironment,as components present in the ascitic fluid were able to inducereceptor internalization.

The MR is one of the oldest CLR described in macro-phages [43, 44]. It is an endocytic and phagocytic receptorthat binds carbohydrate moieties on several pathogens suchas bacteria, fungi, parasites, and viruses and is thereforeconsidered a Pattern Recognition Receptor (PPR). However,it has become increasingly clear that the MR is importantlyinvolved in the silent clearance of circulating inflammatorymolecules and degraded matrix components. Mice deficientin the MR do not show increased susceptibility to infections[45, 46] but have elevated levels of lysosomal hydrolasesand other glycoproteins which raise up during inflammationand tissue remodelling [47, 48]. These in vivo experimentshighlighted its important role in the clearance of unwantedmolecules, especially for the MR localized at hepatic sinu-soids.

Not only MR appears to be dispensable for pathogenclearance but also it can negatively modulate pathogen-elicited immune responses. We and others have pre-viously reported that MR-ligation with ManLAM from

M. tubercolosis or with an agonist anti-MR antibody modu-lates cytokine production in human DC, with a shift fromhigh to low IL-12, increased IL-10 levels, and defectiveTh1 immune responses [26, 27]. These results have beenconfirmed in this study in tumor macrophages activatedwith agonist anti-MR mAbs. The mechanisms that accountfor the regulatory functions of the MR are not completelycharacterized. Unlike other CLRs, MR has no ITIM domain[25]. It has been shown that some CLRs may interfere withTLRs/NF-kB signalling [25, 49], and the MR can indeedphysically interact with TLR2 upon internalization [50]. Inaddition, Pathak et al. reported that mannan induced theupregulation of IRAK-M kinase, which was responsible forthe decreased production of proinflammatory cytokines byinhibiting TLR-signaling [51, 52].

A number of recent studies corroborated the hypothesisthat the MR is implicated in the maintenance of homeostasisand tolerance. Macrophages cocultured with mesenchymalstem cells have high MR expression and produce IL-10 [53].Royer et al. reported that allergens inducing Th2-polarizedresponses express MR-binding carbohydrate moieties; thereceptor contributed to T cell polarization as its silencing in


0

1

2

P < .05

Macro TAM

IL-1

2

ControlAnti-CD206

TAG-72CA125

0

1

2

3

4

IL-1

0(n

g)

P < .05

P < .05

Macro TAM

ControlAnti-CD206

TAG-72CA125

/mL

(ng

)/m

L

(a)

P < .05

P < .05

P < .05

DC Macro TAM0

10

20

30

40

50

CC

L3

(ng

)/m

L

(b)

Figure 5: Modulation of cytokine production by tumoral mucin-engagement of the MR. (a) Purified TAM preparations or normal in vitrodifferentiated macrophages were pretreated (10 min. room temperature) with anti-CD206 (clone WE458, 2 ug/mL), TAG-72 (200 UI/mL), orCA125 (200 UI/mL) prior to stimulation with LPS (1 ug/mL, 24 hrs). Levels of IL-10 (upper panels) and IL-12 (lower panels) were measuredin supernatants by ELISA. Results are mean +/− SE of 6 different TAM preparations for IL-10 and 3 for IL-12 (P < .05, Student’s t-tests).(b) Purified TAM, normal in vitro differentiated macrophages, and mono-DC were pretreated (10 min. room temperature) with TAG-72(200 UI/mL) prior to stimulation with LPS (1 ug/mL, 24 hrs). Levels of CCL3 were measured in supernatants by ELISA. Results are mean+/− SE of 4 different TAM preparations, 3 for macrophages, and mono-DC. P < .05 (Student’s t-tests).

DC strongly impaired Th2 development [54]. Macrophageslocalized at sites where inflammation could be particularlyharmful are usually strongly MR-positive (e.g., alveolarmacrophages and brain microglia). Further, at the maternal-foetal interface the presence of immune cells is important topreserve tolerance as well as for active remodelling of uterinevessels. Decidual macrophages are a major source of IL-10and IDO [55] and express high levels of the MR [40]. MRrecognizes several endogenous ligands and acts as a bridgebetween innate immunity and homeostasis [25, 56]. Forinstance, circulating hormones, like lutropin and thyrotropinare bound by the MR cystein-rich domain [57]. Collagen isanother MR-ligand and the receptor may serve importantscavenger functions [58].

In the context of a tumor microenvironment, wherehighly glycosylated molecules such as mucins are present[59] CLRs encounter several putative ligands. MGL and DC-SIGN recognize cancer-specific glycosylation changes of themucin MUC1, in particular the carbohydrate sialyl LewisX

and the sialyl TN epitope [60]; MUC1 and TAG-72 bind alsothe MR [27, 35, 36]. We previously reported that mono-DCs differentiated in the presence of tumor cell-derivedmucins have a tolerogenic/regulatory cytokine profile [34].In the present study we extended this observation to tumormacrophages: TAMs bound and internalized both TAG-72and the ovarian cancer-associated mucin CA125 via the MR,indicating a specific recognition by this receptor. Further,these mucins interfered with the LPS-induced productionof IL-10 and of the chemokine CCL3. These results arein line with the observation that another tumoral mucin,the carcinoembryonic antigen (CEA) highly expressed bycolon cancer cells, binds DC-SIGN on DC and inducesincreased secretion of IL-10 and IL-6 [61]. Hence, evidenceis accumulating that CLR recognition of tumor glycansleads to the expression of the potent immunoregulatorycytokine IL-10. In the tumor microenvironment IL-10 hasdetrimental effects on immune responses as it promotes thepolarization of M2 macrophages inhibits the differentiation


of Th1 lymphocytes while favouring that of Treg [62].Interestingly, a recent study showed that distinct TAMsubsets can be distinguished on the basis of differentialexpression of MHC II molecules. TAMs characterized byMHC IIlow and suppressive activity on T cell proliferationhave higher expression of the MR [63].

In addition, MR is expressed by endothelial cells oflymphatic vessels [64] and it has been demonstrated to beimplicated in the dissemination of tumor cells along lym-phatics [33]. Recently, Arteta et al. reported that MR-positiveliver sinusoidal vessels also support hepatic metastasis ofcolon cancer cells by a mechanism that involves IL-1-inducedupregulation of the MR [65].

Thus, while under physiological conditions the regula-tory effect of CLRs on innate immunity cells is finalized tothe preservation of homeostasis, in pathological conditionssuch as cancer, CLR activity may hamper the activation ofa protective immune response and actually favour tumorspread.

In conclusion, we have demonstrated that the MR onhuman TAM can be engaged by mucins present in the tumormicroenvironment. This interaction further enhances theirimmunosuppressive phenotype and can be considered asanother mechanism of tumor immune evasion.

Acknowledgments

The authors would like to thank Professor C. Mangioni andDr. M. Signorelli (Hospital S. Gerardo, Monza, Italy), forproviding biological samples from patients and for exper-imental support, and Dr. Manuela Nebuloni, University ofMilan, for immunohistochemistry. This work was supportedby Associazione Italiana Ricerca Cancro (AIRC) Italy to P.Allavena and A. Mantovani, and grants from Ministry ofHealth and Istituto Superiore Sanita Italy (Project Oncology2006 and Alleanza Contro il Cancro).

References

[1] P. Allavena, A. Sica, C. Garlanda, and A. Mantovani, “The Yin-Yang of tumor-associated macrophages in neoplastic progres-sion and immune surveillance,” Immunological Reviews, vol.222, no. 1, pp. 155–161, 2008.

[2] G. Solinas, G. Germano, A. Mantovani, and P. Allavena,“Tumor-associated macrophages (TAM) as major players ofthe cancer-related inflammation,” Journal of Leukocyte Biology,vol. 86, no. 5, pp. 1065–1073, 2009.

[3] C. E. Lewis and J. W. Pollard, “Distinct role of macrophages indifferent tumor microenvironments,” Cancer Research, vol. 66,no. 2, pp. 605–612, 2006.

[4] J. Galon, A. Costes, F. Sanchez-Cabo et al., “Type, density,and location of immune cells within human colorectal tumorspredict clinical outcome,” Science, vol. 313, no. 5795, pp. 1960–1964, 2006.

[5] L. Laghi, P. Bianchi, E. Miranda et al., “CD3+ cells at theinvasive margin of deeply invading (pT3-T4) colorectal cancerand risk of post-surgical metastasis: a longitudinal study,” TheLancet Oncology, vol. 10, no. 9, pp. 877–884, 2009.

[6] L. Bingle, N. J. Brown, and C. E. Lewis, “The role of tumour-associated macrophages in tumour progression: implications

for new anticancer therapies,” Journal of Pathology, vol. 196,no. 3, pp. 254–265, 2002.

[7] J. W. Pollard, “Tumour-educated macrophages promotetumour progression and metastasis,” Nature Reviews Cancer,vol. 4, no. 1, pp. 71–78, 2004.

[8] J. Condeelis and J. W. Pollard, “Macrophages: obligate partnersfor tumor cell migration, invasion, and metastasis,” Cell, vol.124, no. 2, pp. 263–266, 2006.

[9] A. Mantovani, P. Allavena, A. Sica, and F. Balkwill, “Cancer-related inflammation,” Nature, vol. 454, no. 7203, pp. 436–444,2008.

[10] J. A. Joyce and J. W. Pollard, “Microenvironmental regulationof metastasis,” Nature Reviews Cancer, vol. 9, no. 4, pp. 239–252, 2009.

[11] A. Mantovani, S. Sozzani, M. Locati, P. Allavena, and A. Sica,“Macrophage polarization: tumor-associated macrophages asa paradigm for polarized M2 mononuclear phagocytes,”Trends in Immunology, vol. 23, no. 11, pp. 549–555, 2002.

[12] S. Gordon and P. R. Taylor, “Monocyte and macrophageheterogeneity,” Nature Reviews Immunology, vol. 5, no. 12, pp.953–964, 2005.

[13] F. O. Martinez, L. Helming, and S. Gordon, “Alternative activa-tion of macrophages: an immunologic functional perspective,”Annual Review of Immunology, vol. 27, pp. 451–483, 2009.

[14] S. Goerdt and C. E. Orfanos, “Other functions, other genes:alternative activation of antigen- presenting cells,” Immunity,vol. 10, no. 2, pp. 137–142, 1999.

[15] A. Mantovani, A. Sica, and M. Locati, “Macrophage polariza-tion comes of age,” Immunity, vol. 23, no. 4, pp. 344–346, 2005.

[16] J. E. Talmadge, M. Donkor, and E. Scholar, “Inflammatory cellinfiltration of tumors: Jekyll or Hyde,” Cancer and MetastasisReviews, vol. 26, no. 3-4, pp. 373–400, 2007.

[17] C. Steidl, T. Lee, S. P. Shah et al., “Tumor-associatedmacrophages and survival in classic Hodgkin’s lymphoma,”The New England Journal of Medicine, vol. 362, no. 10, pp. 875–885, 2010.

[18] A. Zumsteg and G. Christofori, “Corrupt policemen: inflam-matory cells promote tumor angiogenesis,” Current Opinion inOncology, vol. 21, no. 1, pp. 60–70, 2009.

[19] W. I. Weis, M. E. Taylor, and K. Drickamer, “The C-type lectinsuperfamily in the immune system,” Immunological Reviews,vol. 163, pp. 19–34, 1998.

[20] E. P. McGreal, J. L. Miller, and S. Gordon, “Ligand recognitionby antigen-presenting cell C-type lectin receptors,” CurrentOpinion in Immunology, vol. 17, no. 1, pp. 18–24, 2005.

[21] C. A. Janeway Jr. and R. Medzhitov, “Innate immune recog-nition,” Annual Review of Immunology, vol. 20, pp. 197–216,2002.

[22] M. J. Robinson, D. Sancho, E. C. Slack, S. LeibundGut-Landmann, and C. R. Sousa, “Myeloid C-type lectins in innateimmunity,” Nature Immunology, vol. 7, no. 12, pp. 1258–1265,2006.

[23] T. B. H. Geijtenbeek, S. J. van Vliet, A. Engering, B. A. ’T Hart,and Y. van Kooyk, “Self- and nonself-recognition by C-typelectins on dendritic cells,” Annual Review of Immunology, vol.22, pp. 33–54, 2004.

[24] J. J. Garcıa-Vallejo and Y. Van Kooyk, “Endogenous ligandsfor C-type lectin receptors: the true regulators of immunehomeostasis,” Immunological Reviews, vol. 230, no. 1, pp. 22–37, 2009.

[25] T. B. H. Geijtenbeek and S. I. Gringhuis, “Signalling throughC-type lectin receptors: shaping immune responses,” NatureReviews Immunology, vol. 9, no. 7, pp. 465–479, 2009.


[26] J. Nigou, C. Zelle-Rieser, M. Gilleron, M. Thurnher, andG. Puzo, “Mannosylated lipoarabinomannans inhibit IL-12production by human dendritic cells: evidence for a negativesignal delivered through the mannose receptor,” Journal ofImmunology, vol. 166, no. 12, pp. 7477–7485, 2001.

[27] M. Chieppa, G. Bianchi, A. Doni et al., “Cross-linking of themannose receptor on monocyte-derived dendritic cells acti-vates an anti-inflammatory immunosuppressive program,”Journal of Immunology, vol. 171, no. 9, pp. 4552–4560, 2003.

[28] T. B. H. Geijtenbeek, S. J. Van Vliet, E. A. Koppel et al.,“Mycobacteria target DC-SIGN to suppress dendritic cellfunction,” Journal of Experimental Medicine, vol. 197, no. 1,pp. 7–17, 2003.

[29] A. Dzionek, Y. Sohma, J. Nagafune et al., “BDCA-2, anovel plasmacytoid dendritic cell-specific type II C-typelectin, mediates antigen capture and is a potent inhibitor ofinterferon α/β induction,” Journal of Experimental Medicine,vol. 194, no. 12, pp. 1823–1834, 2001.

[30] F. Meyer-Wentrup, D. Benitez-Ribas, P. J. Tacken et al., “Tar-geting DCIR on human plasmacytoid dendritic cells results inantigen presentation and inhibits IFN-α production,” Blood,vol. 111, no. 8, pp. 4245–4253, 2008.

[31] N. Fujikado, S. Saijo, T. Yonezawa et al., “Dcir deficiencycauses development of autoimmune diseases in mice due toexcess expansion of dendritic cells,” Nature Medicine, vol. 14,no. 2, pp. 176–180, 2008.

[32] S. J. van Vliet, S. I. Gringhuis, T. B. H. Geijtenbeek, and Y. vanKooyk, “Regulation of effector T cells by antigen-presentingcells via interaction of the C-type lectin MGL with CD45,”Nature Immunology, vol. 7, no. 11, pp. 1200–1208, 2006.

[33] F. Marttila-Ichihara, R. Turja, M. Miiluniemi et al., “Macro-phage mannose receptor on lymphatics controls cell traffick-ing,” Blood, vol. 112, no. 1, pp. 64–72, 2008.

[34] P. Monti, B. E. Leone, A. Zerbi et al., “Tumor-derived MUC1mucins interact with differentiating monocytes and induce IL-10IL-12 regulatory dendritic cell,” Journal of Immunology, vol.172, no. 12, pp. 7341–7349, 2004.

[35] E. M. Hiltbold, A. M. Vlad, P. Ciborowski, S. C. Watkins, andO. J. Finn, “The mechanism of unresponsiveness to circulatingtumor antigen MUC1 is a block in intracellular sorting andprocessing by dendritic cells,” Journal of Immunology, vol. 165,no. 7, pp. 3730–3741, 2000.

[36] A. Rughetti, I. Pellicciotta, M. Biffoni et al., “Recombinanttumor-associated MUC1 glycoprotein impairs the differenti-ation and function of dendritic cells,” Journal of Immunology,vol. 174, no. 12, pp. 7764–7772, 2005.

[37] G. Solinas, S. Schiarea, M. Liguori et al., “Tumor-conditionedmacrophages secrete migration-stimulating factor: a newmarker for M2-polarization, influencing tumor cell motility,”Journal of Immunology, vol. 185, no. 1, pp. 642–652, 2010.

[38] A. Saccani, T. Schioppa, C. Porta et al., “p50 nuclear factor-κB overexpression in tumor-associated macrophages inhibitsM1 inflammatory responses and antitumor resistance,” CancerResearch, vol. 66, no. 23, pp. 11432–11440, 2006.

[39] V. Apostolopoulos, N. Barnes, G. A. Pietersz, and I. F. C.McKenzie, “Ex vivo targeting of the macrophage mannosereceptor generates anti-tumor CTL responses,” Vaccine, vol.18, no. 27, pp. 3174–3184, 2000.

[40] G. Laskarin, K. Cupurdija, V. Sotosek Tokmadzic et al., “Thepresence of functional mannose receptor on macrophages atthe maternal-fetal interface,” Human Reproduction, vol. 20, no.4, pp. 1057–1066, 2005.

[41] A. Sica, A. Saccani, B. Bottazzi et al., “Autocrine pro-duction of IL-10 mediates defective IL-12 production and

NF-κB activation in tumor-associated macrophages,” Journalof Immunology, vol. 164, no. 2, pp. 762–767, 2000.

[42] E. K. Thomas, M. Nakamura, D. Wienke, C. M. Isacke, A.Pozzi, and P. Liang, “Endo180 binds to the C-terminal regionof type I collagen,” The Journal of Biological Chemistry, vol.280, no. 24, pp. 22596–22605, 2005.

[43] R. A. Ezekowitz and S. Gordon, “Surface properties of acti-vated macrophages: sensitized lymphocytes, specific antigenand lymphokines reduce expression of antigen F4/80 and FCand mannose/fucosyl receptors, but induce Ia,” Advances inExperimental Medicine and Biology, vol. 155, pp. 401–407,1982.

[44] C. Tietze, P. Schlesinger, and P. Stahl, “Mannose-specificendocytosis receptor of alveolar macrophages: demonstrationof two functionally distinct intracellular pools of receptor andtheir roles in receptor recycling,” Journal of Cell Biology, vol.92, no. 2, pp. 417–424, 1982.

[45] S. D. Swain, S. J. Lee, M. C. Nussenzweig, and A. G. Harmsen,“Absence of the macrophage mannose receptor in mice doesnot increase susceptibility to pneumocystis carinii infection invivo,” Infection and Immunity, vol. 71, no. 11, pp. 6213–6221,2003.

[46] S. J. Lee, N. Y. Zheng, M. Clavijo, and M. C. Nussenzweig,“Normal host defense during systemic candidiasis in mannosereceptor-deficient mice,” Infection and Immunity, vol. 71, no.1, pp. 437–445, 2003.

[47] S. J. Lee, S. Evers, D. Roeder et al., “Mannose receptor-mediated regulation of serum glycoprotein homeostasis,”Science, vol. 295, no. 5561, pp. 1898–1901, 2002.

[48] Y. Mi, S. D. Shapiro, and J. U. Baenziger, “Regula-tion of lutropin circulatory half-life by the mannose/N-acetylgalactosamine-4-SO4 receptor is critical for implanta-tion in vivo,” Journal of Clinical Investigation, vol. 109, no. 2,pp. 269–276, 2002.

[49] W. Harnett and M. M. Harnett, “Helminth-derivedimmunomodulators: can understanding the worm producethe pill?” Nature Reviews Immunology, vol. 10, no. 4, pp.278–284, 2010.

[50] S. D. Tachado, J. Zhang, J. Zhu, N. Patel, M. Cushion, and H.Koziel, “Pneumocystis-mediated IL-8 release by macrophagesrequires coexpression of mannose receptors and TLR2,”Journal of Leukocyte Biology, vol. 81, no. 1, pp. 205–211, 2007.

[51] S. K. Pathak, S. Basu, A. Bhattacharyya, S. Pathak, M. Kundu,and J. Basu, “Mycobacterium tuberculosis lipoarabinoman-nan-mediated IRAK-M induction negatively regulates toll-like receptor-dependent interleukin-12 p40 production inmacrophages,” The Journal of Biological Chemistry, vol. 280,no. 52, pp. 42794–42800, 2005.

[52] J. Zhang, S. D. Tachado, N. Patel et al., “Negative regulatoryrole of mannose receptors on human alveolar macrophageproinflammatory cytokine release in vitro,” Journal of Leuko-cyte Biology, vol. 78, no. 3, pp. 665–674, 2005.

[53] J. Kim and P. Hematti, “Mesenchymal stem cell-educatedmacrophages: a novel type of alternatively activated macro-phages,” Experimental Hematology, vol. 37, no. 12, pp. 1445–1453, 2009.

[54] P.-J. Royer, M. Emara, C. Yang et al., “The mannose receptormediates the uptake of diverse native allergens by dendriticcells and determines allergen-induced T cell polarizationthrough modulation of IDO Activity,” Journal of Immunology,vol. 185, no. 3, pp. 1522–1531, 2010.

[55] T. Nagamatsu and D. J. Schust, “The immunomodulatoryroles of macrophages at the maternal-fetal interface,” Repro-ductive Sciences, vol. 17, no. 3, pp. 209–218, 2010.


[56] U. Gazi and L. Martinez-Pomares, “Influence of the mannosereceptor in host immune responses,” Immunobiology, vol. 214,no. 7, pp. 554–561, 2009.

[57] C. Leteux, W. Chai, R. W. Loveless et al., “The cysteine-rich domain of the macrophage mannose receptor is amultispecific lectin that recognizes chondroitin sulfates Aand B and sulfated oligosaccharides of blood group Lewisand Lewis(x) types in addition to the sulfated N-glycans oflutropin,” Journal of Experimental Medicine, vol. 191, no. 7, pp.1117–1126, 2000.

[58] L. Martinez-Pomares, D. Wienke, R. Stallion et al., “Carbohy-drate-independent recognition of collagens by the macro-phage mannose receptor,” European Journal of Immunology,vol. 36, no. 5, pp. 1074–1082, 2006.

[59] D. W. Kufe, “Mucins in cancer: function, prognosis andtherapy,” Nature Reviews Cancer, vol. 9, no. 12, pp. 874–885,2009.

[60] C. A. Aarnoudse, J. J. G. Vallejo, E. Saeland, and Y. Van Kooyk,“Recognition of tumor glycans by antigen-presenting cells,”Current Opinion in Immunology, vol. 18, no. 1, pp. 105–111,2006.

[61] M. Nonaka, B. Y. Ma, R. Murai et al., “Glycosylation-dependent interactions of C-type lectin DC-SIGN with col-orectal tumor-associated Lewis glycans impair the functionand differentiation of monocyte-derived dendritic cells,” Jour-nal of Immunology, vol. 180, no. 5, pp. 3347–3356, 2008.

[62] A. O’Garra and P. Vieira, “Regulatory T cells and mechanismsof immune system control,” Nature Medicine, vol. 10, no. 8,pp. 801–805, 2004.

[63] K. Movahedi, D. Laoui, C. Gysemans et al., “Different tumormicroenvironments contain functionally distinct subsets ofmacrophages derived from Ly6C(high) monocytes,” CancerResearch, vol. 70, no. 14, pp. 5728–5739, 2010.

[64] H. Irjala, K. Alanen, R. Grenman, P. Heikkila, H. Joensuu, andS. Jalkanen, “Mannose receptor (MR) and common lymphaticendothelial and vascular endothelial receptor (CLEVER)-1 direct the binding of cancer cells to the lymph vesselendothelium,” Cancer Research, vol. 63, no. 15, pp. 4671–4676,2003.

[65] B. Arteta, N. Lasuen, A. Lopategi, B. Sveinbjornsson, B.Smedsrød, and F. Vidal-Vanaclocha, “Colon carcinoma cellinteraction with liver sinusoidal endothelium inhibits organ-specific antitumor immunity through interleukin-1-inducedmannose receptor in mice,” Hepatology, vol. 51, no. 6, pp.2172–2182, 2010.


Research Article

Identification of Two Novel HLA-A∗0201-RestrictedCTL Epitopes Derived from MAGE-A4

Zheng-Cai Jia,1 Bing Ni,1 Ze-Min Huang,1 Yi Tian,1 Jun Tang,2 Jing-Xue Wang,1

Xiao-Lan Fu,1 and Yu-Zhang Wu1

1 Department of Immunology, Third Military Medical University, Chongqing 400038, China2 Department of Dermatology, Southwest Hospital, Third Military Medical University, Chongqing 400038, China

Correspondence should be addressed to Yu-Zhang Wu, [email protected]

Received 29 June 2010; Revised 8 October 2010; Accepted 5 December 2010

Academic Editor: Graham Ogg

Copyright © 2010 Zheng-Cai Jia et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

MAGE-A antigens belong to cancer/testis (CT) antigens that are expressed in tumors but not in normal tissues except testis andplacenta. MAGE-A antigens and their epitope peptides have been used in tumor immunotherapy trials. MAGE-A4 antigen isextensively expressed in various histological types of tumors, so it represents an attractive target for tumor immunotherapy. Inthis study, we predicted HLA-A∗0201-restricted cytotoxic T lymphocyte (CTL) epitopes of MAGE-A4, followed by peptide/HLA-A∗0201 affinity and complex stability assays. Of selected four peptides (designated P1, P2, P3, and P4), P1 (MAGE-A4286−294,KVLEHVVRV) and P3 (MAGE-A4272−280, FLWGPRALA) could elicit peptide-specific CTLs both in vitro from HLA-A∗0201-positive PBMCs and in HLA-A∗0201/Kb transgenic mice. And the induced CTLs could lyse target cells in an HLA-A∗0201-restricted fashion, demonstrating that the two peptides are HLA-A∗0201-restricted CTL epitopes and could serve as targets fortherapeutic antitumoral vaccination.

1. Introduction

The melanoma antigen genes family A (MAGE-A) consistsof 12 closely related genes (MAGE-A1 to A12) located inthe q28 region of chromosome X [1, 2]. MAGE-A-encodedantigens (MAGE-A) belong to cancer/testis (CT) antigens.These antigens are expressed in various histological types ofcarcinomas, but not in normal tissues with the exception oftestis and placenta [3–8]. Although testis expresses MAGE-A antigens and placenta also expresses some of them [9],testis and placenta do not express MHC class I moleculesand therefore cannot be attacked by cytotoxic T lymphocytes(CTLs) specific for these antigens. Thus, MAGE-A antigensare appealing targets for antitumor immunotherapy. Anumber of clinical trials of therapeutic vaccination havebeen performed, based on these antigens and their epitopepeptides. In some clinical trials executed with short peptides,tumor regressions have been observed in a minority ofpatients [10–12].

Of the MAGE-A family, MAGE-A4 is one of genes thatare abundantly expressed by many tumors of different

histological types, such as urothelial carcinoma, bladdercancer, lung cancer, ovarian neoplasm, esophageal squamouscell carcinoma, and oral squamous cell carcinoma [4, 13–17]. Up to now, at least four variants have been found forthis gene. The four variants encode the same protein referredto MAGE-A4. MAGE-A4 is found to interact with the liveroncoprotein gankyrin and suppress the tumorigenic activityof gankyrin [18]. Its carboxyl-terminal fragment of 107amino acids induces p53-dependent and p53-independentapoptosis in human cells [19]. Moreover, the expressionof MAGE-A4 may increase caspase-3 activity and promotetumor cell death [14].

Recent studies have identified several antigenic peptidespresented by HLA class I molecules, including HLA-A24restricted MAGE-A4143–151 (NYKRCFPVI) peptide, HLA-A1restricted MAGE-A4169–177 (EVDPASNTY) peptide, HLA-B37 restricted MAGE-A4156–163 (SESLKMIF) peptide, andHLA-A∗0201 restricted MAGE-A4230–239 (GVYDGREHTV)peptide [20–23]. It is reported that a polyepitope vaccinetargeted to one antigen may elicit strong antigen-specificCTLs to protect against tumor challenge and almost each


epitope in the polyepitope can induce specific CTL immuneresponse [24, 25]. On the other hand, about 50% ofCaucasians and Asians express HLA-A∗0201 [23]. So theidentification of more HLA-A∗0201-restricted epitopes forMAGE-A4 is likely to provide alternative candidates for thefuture of clinical trials with defined antigenic peptides andfacilitate the design of antitumor vaccines with high efficacy.

To identify epitopes capable of inducing MAGE-A4-specific HLA-A∗0201-restricted CTLs, we first predictedHLA-A∗0201-restricted epitopes of MAGE-A4 and mea-sured HLA-A∗0201 binding capacity of the candidate epi-tope peptides. We then induced MAGE-A4-specific CTLsfrom HLA-A∗0201 peripheral blood mononuclear cells(PBMCs) with these candidate peptides to seek CTL epitopesfrom MAGE-A4 antigen, followed by validation for in vivoimmunogenicity.

2. Material and Methods

2.1. Cell Lines and Animals. The HLA-A∗0201-expressinghuman tumor cells T2 (deficient in TAP1 and TAP2 trans-porters), MCF-7 (breast cancer; MAGE-A-negative [26]),and BB7.2 hybridoma producing anti-HLA-A2 monoclonalantibody (mAb) were purchased from American TypeCulture Collection (ATCC, USA). Human melanoma cellsLB1751-MEL expressing MAGE-A and HLA-A∗0201 werekindly provided by Dr. F. Brasseur (Ludwig Institute forCancer Research, Brussels, Belgium).

HLA-A∗0201/Kb transgenic mice (6–8 weeks old) werepurchased from the Jackson Laboratory (USA). Animalexperiments were performed in accordance with the guide-lines of the Animal Care and Use Committee of ThirdMilitary Medical University.

2.2. Epitope Prediction and Peptide Synthesis. The MAGE-A4protein sequence was analyzed for 9-amino acid long pep-tides, which could potentially bind to HLA-A∗0201 mole-cule, using the computer-based epitope prediction programsBIMAS (http://www-bimas.cit.nih.gov/molbio/hla bind/in-dex.shtml) and SYFPEITHI (http://www.syfpeithi.de/Scripts/MHCServer.dll/EpitopePrediction.htm). The selected can-didate peptides and control peptides (HBcAg18–27 andOVA257–264) were synthesized by Fmoc chemistry (Sangon,China) and purified by HPLC to a purity of >95%.

2.3. Affinity Measurement of Peptide for HLA-A∗0201. Theaffinity of peptides for HLA-A∗0201 was measured asdescribed previously [27]. Briefly, T2 cells were incubatedwith various concentrations of each peptide and 3 μg/mLhuman β2m in serum-free RPMI 1640 medium at 37◦C for16 h. Then, the cells were washed and stained with antiHLA-A2 mAb and FITC-labeled goat anti-mouse IgG. The expres-sion of HLA-A∗0201 on T2 cells was determined with FACSCalibur flow cytometer (Becton Dickinson, USA). For eachpeptide concentration, the percent mean fluorescence index(% MFI) increase of HLA-A∗0201 molecule was calculatedas follows: % MFI increase = [(MFI with the given peptide−MFI without peptide)/(MFI without peptide)] × 100.

2.4. Assessment of Peptide/HLA-A∗0201 Complex Stability. Aspreviously described [28], T2 cells (106/mL) were incubatedovernight with 100 μM of each peptide in serum-free RPMI1640 medium supplemented with 100 ng/mL human β2mat 37◦C. Then, they were washed to remove free peptidesand incubated with 10 μg/mL of Brefeldin A (Sigma-Aldrich,USA) for 1 h to block newly synthesized HLA-A∗0201molecules to be expressed on cell surface, washed andincubated at 37◦C for 0, 2, 4, 6, or 8 h. Subsequently, thecells were stained with the anti-HLA-A2 mAb to evaluate theexpression of HLA-A∗0201 molecules.

2.5. Plasmid Construction and Cell Transfection. The mam-malian expression plasmid pCI-MAGEA4, which containsthe encoding sequence of MAGE-A4, was constructed asdescribed below. Total RNA was extracted from LB1751-MEL cells using TRIZOL reagent (Invitrogen, USA). First-strand cDNA was synthesized and PCR was performed usingHigh Fidelity PrimeScript RT-PCR Kit (TaKaRa, Dalian,China) and primers (forward, 5′-TGCCCTGACCAGAGT-CATCAT-3′; reverse, 5′-ACAGAGTGAAGAATGGGCCT-3′), according to the manufacturer’s instructions. The ampli-fied products were inserted into pMD18-T plasmid (TaKaRa,Dalian, China) and the plasmid cloned into cDNA sequenceof MAGE-A4 was selected and identified by restrictionendonuclease digestion and sequencing. And then the encod-ing sequence of MAGE-A4 was amplified from the selectedplasmid above and cloned into Nhe I/Mlu I sites of pCI-neo plasmid (Promega, Beijing, China) using primers 5′-TCTAGCTAGCATGTCTTCTGAGCAGAAGAGTCAGC-3′

(forward) and 5′-CCTACGACGCGTTCAGACTCCCTC-TTCCTCCTCTAAC-3′ (reverse).

To establish a cell line expressing both HLA-A∗0201 andMAGE-A4, MCF-7 cells were transfected with plasmid pCI-MAGEA4 using Lipofectamine 2000 (Invitrogen, USA) andthen selected with G418. The expression of MAGE-A4 in theestablished cell line (designated MCF-7A4) was confirmed byreverse transcription-PCR and western blotting.

2.6. Induction of CTLs from Human PBMCs. PBMCs wereisolated from the buffy coat of heparinized whole bloodsamples of healthy HLA-A∗0201 donors by density gra-dient centrifugation on the Ficoll-Paque PREMIUM (GEHealthcare Bio-Sciences AB, Uppsala, Sweden). The effectorlymphocytes and dendritic cells (DCs) were prepared by ourpublished method [29]. All donors signed written, informedconsent to provide whole blood samples used in the study.Approval of the study was obtained from the relevant ethicalcommittees and was in accordance with the Declaration ofHelsinki.

2.7. Cytotoxicity Assay. Three to five days after the finalstimulation, the cytotoxic activity of the effector cellswas evaluated by a lactate dehydrogenase release assayusing Cytotox 96 Non-Radioactive Cytotoxicity Assay kit(Promega, USA) [30]. In brief, 1 × 104 target cells (LB1751-MEL, MCF-7 or MCF-7A4) in 50 μL RPMI 1640 containing5% fetal calf serum (FCS) was placed in the wells of a 96-well


round-bottom plate, then 50 μL of various concentrations ofeffector cells was added at different effector to target (E/T)ratios (50/1, 25/1, and 12.5/1). After 4 h incubation at 37◦C,the supernatant was collected to assay lactate dehydrogenase(LDH) release by OD490 measurement according to themanufacturer’s instructions. Experiments were performed intriplicates and the percentage of lysis was calculated as %Lysis = [(experimental LDH release − effector spontaneousLDH release − target spontaneous LDH release)/(targetmaximum LDH release − target spontaneous LDH release)]× 100.

2.8. Analysis of In Vivo Immunogenicity. HLA-A∗0201/Kb

mice were immunized with 100 μg of various peptides pre-pared in IFA or IFA emulsion without peptide as a control.After 10 days, mice were sacrificed and splenocytes werecultured for 5 days with 10 units/mL recombinant murineinterleukin-2 (rmIL-2) and 2 μg/mL peptide. And then,effector cells were counted and tested for cytotoxic activityin a cytotoxicity assay.

2.9. Statistical Analysis. Statistical analyses were performedusing the variance test and Student’s t-test. A difference wasconsidered significant at the conventional level of P < .05.

3. Results

3.1. Prediction of HLA-A∗0201-Restricted CTL Epitopes. TheHLA-A∗0201-restricted CTL epitopes of MAGE-A4 werepredicted using BIMAS software. The top four rankingpeptides with BIMAS scores were selected and then verifiedusing the program SYFPEITHI. As shown in Table 1, all ofthe four peptides had high SYFPEITHI scores. Thus, theywere selected as the candidate eptitope peptides.

3.2. Affinity of Candidate Epitope Peptides for HLA-A∗0201Molecule. We then evaluated the binding affinity of thesecandidate epitope peptides for HLA-A∗0201 molecule invitro using a T2-cell-peptide binding test. Figure 1 showedthat P3 had highest affinity for HLA-A∗0201 molecule andP2 was lower affinity peptide, while P1 and P4 had lowestaffinity. The negative control peptide did not increase theexpression of HLA-A∗0201 molecule on the T2 cell surfaceat all indicated peptide concentrations.

Because a stable peptide-MHC complex is very impor-tant for the induction of an antigen-specific CTL immuneresponse [29, 31, 32], we further investigated the capacityof candidate epitope peptides to stabilize the HLA-A∗0201molecule (Table 2). The results indicated that P3 exhibitedhighest stabilization capacity (DC50 > 8 h) and P4 was weakstabilizer of HLA-A∗0201 molecule (DC50 < 2 h). P2 hada binding affinity higher than that of P1 (Figure 1), butP2 stabilized HLA-A∗0201 molecule more weakly than P1(DC50 < 2 h and 4–6 h, resp.).

3.3. In Vitro Induction of Peptide-Specific CTLs. To studywhether these candidate epitope peptides can induce thegeneration of peptide-specific CTLs in vitro, PBMCs from

1 3 5 10 20 50 100

Peptide concentration (μM)

0

50

100

150

200

250

MFI

incr

ease

(%)

P1P2P3

P4HBcAg18–27

OVA257–264

Figure 1: Binding affinity of peptides for HLA-A∗0201 molecule.T2 cells were incubated with indicated concentrations of thepeptides in serum-free RPMI 1640 medium supplied with 3 μg/mLhuman β2m at 37◦C for 16 h. And then the cells were stainedwith anti-HLA-A2 mAb and FITC-labeled goat antimouse IgG. Theexpression of HLA-A∗0201 on T2 cells was determined with FACSCalibur flow cytometer. The peptides HBcAg18–27 and OVA257–264

were taken as positive control and negative control, respectively.Each sample was measured in three replicates and the experimentwas repeated three times.

3 HLA-A∗0201 individuals were prepared and stimulatedwith peptide-pulsed autologous DCs and PBMCs succes-sively. The cytotoxic activity of the stimulated PBMCs(effector cells) was evaluated using a cytotoxicity assay.The data from one representative donor was shown inFigure 2. The results showed that P1 and P3-stimulatedPBMCs could significantly lyse the target cells LB1751-MEL. However, similar to the irrelevant peptide HBcAg18–27,P2- and P4-stimulated PBMCs could not lyse the targetcells (Figure 2(a)). After blocking HLA-A∗0201 moleculeson the surface of LB1751-MEL cells with anti-HLA-A2mAb, the lysis of LB1751-MEL cells by the effector cellswas significantly abrogated (Figure 2(b)). Moreover, P1-and P3-primed effector cells could also lyse MCF-7A4 cellsexpressing both HLA-A∗0201 and MAGE-A4 (Figure 2(c)),but not MAGE-A4-negative MCF-7 cells (Figure 2(d)). Thesimilar results were obtained when the other two donors weretested with these peptides (data not shown).

3.4. In Vivo Induction of Peptide-Specific CTLs in HLA-A∗0201/Kb Transgenic Mice. Finally, we investigated whetherthe peptides P1 and P3 could also elicit CTL immuneresponses in vivo. The HLA-A∗0201/Kb transgenic mice wereinoculated once with the two peptides, respectively. Tendays later, the splenocytes were harvested and stimulated invitro with the corresponding peptide. The cytotoxicity assayshowed that the splenocytes from the P1- and P3-inoculatedmice could lyse target cells LB1751-MEL, but the splenocytesfrom the IFA-inoculated mice could not lyse the target cellsafter stimulated in vitro with P1 or P3. When anti-HLA-A2


Table 1: Predicted HLA-A∗0201-restricted CTL epitopes by BIMAS and SYFPEITHI methods.

Peptide Sequence PositionBIMAS SYFPEITHI

Score Rank Score Rank

P1 KVLEHVVRV 286–294 743 1 25 4

P2 ALLEEEEGV 309–317 517 2 27 3

P3 FLWGPRALA 272–280 189 3 21 9

P4 ALPTTISFT 71–79 94 4 20 10

Table 2: HLA-A∗0201 stabilization capacity of candidate epitope peptides.

Peptide% MFI increase at indicated time point (h)

DC50a

0 2 4 6 8

P1 126.16 95.74 74.98 60.29 49.07 4–6

P2 170.07 84.13 57.09 42.34 31.19 <2

P3 285.66 230.85 192.71 170.05 150.56 >8

P4 98.80 48.73 30.79 21.01 18.52 <2

HBcAg18–27 243.26 194.51 161.13 148.39 131.08 >8(a)

Half-time of the peptide/HLA-A∗0201 complex.

mAb was added, anti-HLA-A2 mAb inhibited the peptides-induced splenocytes from killing the targets (Figure 3).

4. Discussion

Tumor-specific immunotherapy is an appealing strategy intreating tumors. The identification of tumor-associated anti-gens (TAAs) and their epitopes has boosted the developmentof the strategy. TAAs are composed of five major groups:cancer/testis (CT) antigens (e.g., MAGE, BAGE, GAGE, NY-ESO-1, and SSX [33–35]), mutated antigens (e.g., MUM-1,p53, and beta-catenin [36, 37]), overexpressed antigens (e.g.,RCAS1, Survivin, and Her2/neu [38–40]), oncofetal antigens(e.g., Immature laminin receptor and CEA [41, 42]), anddifferentiation or lineage antigens (e.g., tyrosinase, Melan-A/MART-1, gp100, TRP-1, and TRP-2 [43, 44]). BecauseCTLs play a key role in antitumor immune responses [45]and CT antigens are expressed in many tumors but not innormal tissues except testis and placenta, the identification ofCTL epitopes derived from CT antigens is very important forthe studies of antitumor vaccines based on defined epitopepeptides.

In this study, we predicted the HLA-A∗0201-restrictedCTL epitopes of the tumor antigen MAGE-A4 with a com-bination of BIMAS and SYFPEITHI programs and selectedfour peptides (P1, P2, P3, and P4) as candidates basedon immunogenicity score. Then we examined the bindingaffinity of these peptides for HLA-A∗0201 and peptide/HLA-A∗0201 complex stability. The results showed that P3 hadboth highest binding affinity for HLA-A∗0201 and strongestcapacity of stabilizing complex among the four peptides.P2 had intermediate binding affinity, but it exhibited weakstabilization capacity. On the other hand, P1 bound weaklyto HLA-A∗0201 molecules, but it could form more stablecomplexes with HLA-A∗0201 molecules than P2 and P4.Recent studies indicate that peptide/MHC complex stabilityis an important parameter for distinguishing immunogenic

peptides from nonimmunogenic peptides, and that a stablepeptide/MHC complex may facilitate the formation of thesynapses between T cells and antigen-presenting cells (APCs)and warrant the full T cell activation through sustainedsignaling [28, 31, 32, 46]. Our data suggested that P1 and P3could be promising epitope candidates.

An in vitro CTL induction assay confirmed that P1and P3 could effectively prime peptide-specific CTLs thatcould lyse HLA-A∗0201+MAGE-A4+ target cells in anHLA-A∗0201-restricted fashion, but P2- and P4-stimulatedPBMCs could not lyse the target cells. At the same time,P1- and P3-stimulated PBMCs could also lyse the HLA-A∗0201+ MCF-7A4 cells transected with MAGE-A4 gene butnot MAGE-A4-negative MCF-7 cells. The in vivo immuno-genicity analysis found that P1 and P3 could also inducespecific CTL immune responses in vivo. These data show thatP1 and P3 are HLA-A∗0201-restricted CTL epitopes and thetwo peptides can induce peptide-specific CTLs both in vitroand in vivo which recognize endogenously processed MAGE-A4 antigen.

Noticeably, by aligning the sequences of MAGE-A, wefound that P3 (FLWGPRALA) is shared by MAGE-A1, -A4,and -A8 and that P3 is highly homologous to two peptidesFLWGPRALI (shared by MAGE-A2 and -A6) and FLWG-PRALV (shared by MAGE-A3 and -A12) with just one aminoacid different at the carboxyl-terminus. It has been reportedthat the interaction between epitope peptide and MHC classI molecule mainly depends on anchor residues P2 and P9 inthe nonapeptide, and the P3–P8 segment of the nonapeptideepitope contributes to the peptide/TCR interaction [28, 47–49]. The peptide FLWGPRALV is a known HLA-A∗0201restricted epitope of MAGE-A3, but it is not efficiently pro-cessed by tumor cells [50]. Therefore, we guess that P3 mightbe a common HLA-A∗0201-restricted CTL epitope amongmost of MAGE-A family members and have a potentialapplication in peptide-mediated immunotherapy, becauseabove 80% of all tumors express at least one MAGE-A


12.5 25 50

E/T ratio

0

10

20

30

40

50

60

70Ly

sis

(%)

P1P2P3

P4HBcAg18–27

(a)

12.5 25 50

E/T ratio

0

10

20

30

40

50

60

70

Lysi

s(%

)

P1P3

P1 + HLA-A2 mAbP3 + HLA-A2 mAb

(b)

12.5 25 50

E/T ratio

0

10

20

30

40

50

60

70

Lysi

s(%

)

P1P3HBcAg18–27

(c)

12.5 25 50

E/T ratio

0

10

20

30

40

50

60

70Ly

sis

(%)

P1P3HBcAg18–27

(d)

Figure 2: Induction of peptide-specific CTLs from human PBMCs. PBMCs from healthy HLA-A∗0201 donors were first stimulatedwith peptide-pulsed autologous DCs and then restimulated with peptide-pulsed autologous PBMCs. After three to five days of the finalstimulation, the stimulated PBMCs were used as effector cells to detect their cytotoxic activity against tumor cells at the indicated E/T ratiosin a cytotoxicity assay. The irrelevant peptide HBcAg18–27 was taken as negative control. Each sample was measured in three replicates andthe experiment was repeated three times. (a) P1, P2, P3, and P4-stimulated PBMCs mediated lysis of LB1751-MEL cells. (b) P1- and P3-stimulated PBMCs mediated lysis of LB1751-MEL cells (P1 and P3) and LB1751-MEL cells which surface HLA-A∗0201 molecules wereblocked with ant- HLA-A∗0201 mAb (P1 + HLA-A2 mAb and P3 + HLA-A2 mAb). (c) P1- and P3-stimulated PBMCs mediated lysis ofMCF-7A4 cells. (d) P1- and P3-stimulated PBMCs mediated lysis of MCF-7 cells.

antigen. In addition, the P1 epitope (KVLEHVVRV) isshared by MAGE-A4 and MAGE-A8. The peptide seems tohave homology to the HLA-A∗0201-restricted CTL epitopeMAGE-A1278–286 (KVLEYVIKV) [51], but the two epitopesexist TCR ligand sequence dissimilarity with difference at theP5, P7, and P8 residues. It remains unclear if the two epitopesKVLEHVVRV and KVLEYVIKV have the same specificity.

In conclusion, our results demonstrate that P1 (MAGE-A4286–294, KVLEHVVRV) and P3 (MAGE-A4272–280,

FLWGPRALA) derived from MAGE-A4 are HLA-A∗0201-restricted CTL epitopes, which can be endogenouslypresented to the surface of HLA-A∗0201+MAGE-A4+ tumorcells. The identification of the two epitopes might providealternative candidates for the studies of tumor-therapeuticvaccines based on defined antigenic peptides. Currently, weare investigating if P3-stimulated PBMCs can also recognizeand kill the HLA-A∗0201+ tumor cells expressing one ofother members of MAGE-A family.


12.5 25 50

E/T ratio

0

10

20

30

40

50

60

70

Lysi

s(%

)

P1P3IFA + P1

IFA + P3P1 + HLA-A2 mAbP3 + HLA-A2 mAb

Figure 3: In vivo induction of peptide-specific CTLs. The HLA-A∗0201/Kb transgenic mice were immunized with the peptidesP1 and P3 prepared in IFA, respectively. Another group of micewere immunized with IFA without peptide as negative control.Mice were sacrificed 10 days after immunization and splenocyteswere stimulated in vitro for 5 days with 2 μg/mL peptide and10 units/mL rmIL-2 to expand them as effector cells. A cytotoxicityassay was used to evaluate the lysis of LB1751-MEL cells by peptide-stimulated splenocytes from corresponding peptide-immunizedmice (P1 and P3), P1- or P3-stimulated splenocytes from IFA-immunized mice (IFA+P1 and IFA+P3), and the lysis of LB1751-MEL cells, which surface HLA-A∗0201 molecules were blocked withant-HLA-A∗0201 mAb, by peptide-immunized mice (P1 + HLA-A2mAb and P3 + HLA-A2 mAb). The experiment was repeated threetimes.

Acknowledgments

This work was supported by Natural Science Foundationof Chongqing (no. 2007BB5009). Z. C. Jia and B. Nicontributed equally to the work.

References

[1] E. De Plaen, K. Arden, C. Traversari et al., “Structure,chromosomal localization, and expression of 12 genes of theMAGE family,” Immunogenetics, vol. 40, no. 5, pp. 360–369,1994.

[2] U. C. Rogner, K. Wilke, E. Steck, B. Korn, and A. Poustka,“The melanoma antigen gene (MAGE) family is clustered inthe chromosomal band Xq28,” Genomics, vol. 29, no. 3, pp.725–731, 1995.

[3] A. A. Jungbluth, K. J. Busam, D. Kolb et al., “Expression ofMAGE-antigens in normal tissues and cancer,” InternationalJournal of Cancer, vol. 85, no. 4, pp. 460–465, 2000.

[4] A. Bergeron, V. Picard, H. LaRue et al., “High frequency ofMAGE-A4 and MAGE-A9 expression in high-risk bladdercancer,” International Journal of Cancer, vol. 125, no. 6, pp.1365–1371, 2009.

[5] M. Otte, M. Zafrakas, L. Riethdorf et al., “MAGE-A geneexpression pattern in primary breast cancer,” Cancer Research,vol. 61, no. 18, pp. 6682–6687, 2001.

[6] J. R. Tsai, I. W. Chong, Y. H. Chen et al., “Differentialexpression profile of MAGE family in non-small-cell lungcancer,” Lung Cancer, vol. 56, no. 2, pp. 185–192, 2007.

[7] J. Li, Y. Yang, T. Fujie et al., “Expression of BAGE, GAGE, andMAGE Genes in Human Gastric Carcinoma,” Clinical CancerResearch, vol. 2, no. 9, pp. 1619–1625, 1996.

[8] N. Yamashita, H. Ishibashi, K. Hayashida et al., “Highfrequency of the MAGE-1 gene expression in hepatocellularcarcinoma,” Hepatology, vol. 24, no. 6, pp. 1437–1440, 1996.

[9] A. A. Jungbluth, W. A. Silva, K. Iversen et al., “Expression ofcancer-testis (CT) antigens in placenta,” Cancer Immunity, vol.7, p. 15, 2007.

[10] F. Tanaka, N. Haraguchi, K. Isikawa, H. Inoue, and M.Mori, “Potential role of dendritic cell vaccination with MAGEpeptides in gastrointestinal carcinomas,” Oncology Reports,vol. 20, no. 5, pp. 1111–1116, 2008.

[11] N. van Baren, M. C. Bonnet, B. Dreno et al., “Tumoral andimmunologic response after vaccination of melanoma patientswith an ALVAC virus encoding MAGE antigens recognized byT cells,” Journal of Clinical Oncology, vol. 23, no. 35, pp. 9008–9021, 2005.

[12] M. Marchand, N. Van Baren, P. Weynants et al., “Tumorregressions observed in patients with metastatic melanomatreated with an antigenic peptide encoded by gene MAGE-3and presented by HLA-A1,” International Journal of Cancer,vol. 80, no. 2, pp. 219–230, 1999.

[13] P. Sharma, Y. Shen, S. Wen et al., “Cancer-testis antigens:expression and correlation with survival in human urothelialcarcinoma,” Clinical Cancer Research, vol. 12, no. 18, pp. 5442–5447, 2006.

[14] T. Peikert, U. Specks, C. Farver, S. C. Erzurum, and S. A. A.Comhair, “Melanoma antigen A4 is expressed in non-smallcell lung cancers and promotes apoptosis,” Cancer Research,vol. 66, no. 9, pp. 4693–4700, 2006.

[15] E. Yakirevich, E. Sabo, O. Lavie, S. Mazareb, G. C. Spagnoli,and M. B. Resnick, “Expression of the MAGE-A4 and NY-ESO-1 cancer-testis antigens in serous ovarian neoplasms,”Clinical Cancer Research, vol. 9, no. 17, pp. 6453–6460, 2003.

[16] V. Quillien, J. L. Raoul, D. Heresbach, B. Collet, L. Toujas,and F. Brasseur, “Expression of MAGE genes in esophagealsquamous-cell carcinoma,” Anticancer Research, vol. 17, no. 1A, pp. 387–391, 1997.

[17] J. Ries, S. Schultze-Mosgau, F. Neukam, E. Diebel, andJ. Wiltfang, “Investigation of the expression of melanomaantigen-encoding genes (MAGE-A1 to -A6) in oral squamouscell carcinomas to determine potential targets for gene-basedcancer immunotherapy,” International Journal of Oncology,vol. 26, no. 3, pp. 817–824, 2005.

[18] T. Nagao, H. Higashitsuji, K. Nonoguchi et al., “MAGE-A4interacts with the liver oncoprotein gankyrin and suppressesits tumorigenic activity,” The Journal of Biological Chemistry,vol. 278, no. 12, pp. 10668–10674, 2003.

[19] T. Sakurai, K. Itoh, H. Higashitsuji et al., “A cleaved form ofMAGE-A4 binds to Miz-1 and induces apoptosis in humancells,” The Journal of Biological Chemistry, vol. 279, no. 15, pp.15505–15514, 2004.

[20] S. Ottaviani, D. Colau, and P. Van Der Bruggen, “A newMAGE-4 antigenic peptide recognized by cytolytic T lym-phocytes on HLA-A24 carcinoma cells,” Cancer Immunology,Immunotherapy, vol. 55, no. 7, pp. 867–872, 2006.

[21] T. Kobayashi, C. Lonchay, D. Colau, N. Demotte, T. Boon,and P. van der Bruggen, “New MAGE-4 antigenic peptiderecognized by cytolytic T lymphocytes on HLA-A1 tumorcells,” Tissue Antigens, vol. 62, no. 5, pp. 426–432, 2003.


[22] Y. Zhang, V. Stroobant, V. Russo, T. Boon, and P. Van derBruggen, “A MAGE-A4 peptide presented by HLA-B37 isrecognized on human tumors by cytolytic T lymphocytes,”Tissue Antigens, vol. 60, no. 5, pp. 365–371, 2002.

[23] M. T. Duffour, P. Chaux, C. Lurquin, G. Cornelis, T. Boon,and P. Van Der Bruggen, “A MAGE-A4 peptide presented byHLA-A2 is recognized by cytolytic T lymphocytes,” EuropeanJournal of Immunology, vol. 29, no. 10, pp. 3329–3337, 1999.

[24] J. Duraiswamy, M. Sherritt, S. Thomson et al., “TherapeuticLMP1 polyepitope vaccine for EBV-associated Hodgkin dis-ease and nasopharyngeal carcinoma,” Blood, vol. 101, no. 8,pp. 3150–3156, 2003.

[25] A. Scardino, M. Alimandi, P. Correale et al., “A polyepitopeDNA vaccine targeted to Her-2/ErbB-2 elicits a broad rangeof human and murine CTL effectors to protect against tumorchallenge,” Cancer Research, vol. 67, no. 14, pp. 7028–7036,2007.

[26] S. Kayser, I. Watermann, C. Rentzsch, T. Weinschenk, D.Wallwiener, and B. Guckel, “Tumor-associated antigen pro-filing in breast and ovarian cancer: mRNA, protein or Tcell recognition?” Journal of Cancer Research and ClinicalOncology, vol. 129, no. 7, pp. 397–409, 2003.

[27] H. W. Nijman, J. G. A. Houbiers, M. P. M. Vierboom et al.,“Identification of peptide sequences that potentially trig-ger HLA-A2.1-restricted cytotoxic T lymphocytes,” EuropeanJournal of Immunology, vol. 23, no. 6, pp. 1215–1219, 1993.

[28] S. Tourdot, A. Scardino, E. Saloustrou et al., “A generalstrategy to enhance immunogenicity of low-affinity HLA-A2.1-associated peptides: implication in the identification ofcryptic tumor epitopes,” European Journal of Immunology, vol.30, no. 12, pp. 3411–3421, 2000.

[29] Y. Tang, Z. Lin, B. Ni et al., “An altered peptide ligand for naıvecytotoxic T lymphocyte epitope of TRP-2 (180–188) enhancedimmunogenicity,” Cancer Immunology, Immunotherapy, vol.56, no. 3, pp. 319–329, 2007.

[30] M. L. Shinohara, M. Jansson, E. S. Hwang, M. B. F. Werneck,L. H. Glimcher, and H. Cantor, “T-bet-dependent expressionof osteopontin contributes to T cell polarization,” Proceedingsof the National Academy of Sciences of the United States ofAmerica, vol. 102, no. 47, pp. 17101–17106, 2005.

[31] J. E. Slansky, F. M. Rattis, L. F. Boyd et al., “Enhanced antigen-specific antitumor immunity with altered peptide ligands thatstabilize the MHC-peptide-TCR complex,” Immunity, vol. 13,no. 4, pp. 529–538, 2000.

[32] S. H. van der Burg, M. J. W. Visseren, R. M. P. Brandt, W.M. Kast, and C. J. M. Melief, “Immunogenicity of peptidesbound to MHC class I molecules depends on the MHC-peptide complex stability,” Journal of Immunology, vol. 156,no. 9, pp. 3308–3314, 1996.

[33] P. Dalerba, E. Frascella, B. Macino et al., “MAGE, BAGEand GAGE gene expression in human rhabdomyosarcomas,”International Journal of Cancer, vol. 93, no. 1, pp. 85–90, 2001.

[34] M. Bolli, E. Schultz-Thater, P. Zajac et al., “NY-ESO-1/LAGE-1 coexpression with MAGE-A cancer/testis antigens: a tissuemicroarray study,” International Journal of Cancer, vol. 115,no. 6, pp. 960–966, 2005.

[35] N. R. Dos Santos, R. Torensma, T. J. De Vries et al.,“Heterogeneous expression of the SSX cancer/testis antigensin human melanoma lesions and cell lines,” Cancer Research,vol. 60, no. 6, pp. 1654–1662, 2000.

[36] C. Castelli, L. Rivoltini, G. Andreola, M. Carrabba, N.Renkvist, and G. Parmiani, “T-cell recognition of melanoma-associated antigens,” Journal of Cellular Physiology, vol. 182,no. 3, pp. 323–331, 2000.

[37] E. V. Fedoseyeva, F. Boisgerault, N. G. Anosova et al., “CD4+

T cell responses to self- and mutated p53 determinants duringtumorigenesis in mice,” Journal of Immunology, vol. 164, no.11, pp. 5641–5651, 2000.

[38] Y. Ito, H. Yoshida, K. Nakano et al., “Overexpression of humantumor-associated antigen, RCAS1, is significantly linked todedifferentiation of thyroid carcinoma,” Oncology, vol. 64, no.1, pp. 83–89, 2003.

[39] S. M. Schmidt, K. Schag, M. R. Muller et al., “Survivin isa shared tumor-associated antigen expressed in a broad varietyof malignancies and recognized by specific cytotoxic T cells,”Blood, vol. 102, no. 2, pp. 571–576, 2003.

[40] Z. Latif, A. D. Watters, I. Dunn, K. M. Grigor, MA.Underwood, and J. Bartlett, “HER2/neu overexpression in thedevelopment of muscle-invasive transitional cell carcinoma ofthe bladder,” British Journal of Cancer, vol. 89, no. 7, pp. 1305–1309, 2003.

[41] B. Friedrichs, S. Siegel, M. Kloess et al., “Humoral immuneresponses against the immature laminin receptor protein showprognostic significance in patients with chronic lymphocyticleukemia,” Journal of Immunology, vol. 180, no. 9, pp. 6374–6384, 2008.

[42] T. Itoh, Y. Ueda, I. Kawashima et al., “Immunotherapy ofsolid cancer using dendritic cells pulsed with the HLA-A24-restricted peptide of carcinoembryonic antigen,” CancerImmunology, Immunotherapy, vol. 51, no. 2, pp. 99–106, 2002.

[43] T. J. de Vries, A. Fourkour, T. Wobbes, G. Verkroost, D. J.Ruiter, and G. N. P. Van Muijen, “Heterogeneous expressionof immunotherapy candidate proteins gp100, MART-1, andtyrosinase in human melanoma cell lines and in humanmelanocytic lesions,” Cancer Research, vol. 57, no. 15, pp.3223–3229, 1997.

[44] T. J. de Vries, D. Trancıkova, D. J. Ruiter, and G. N. P.Van Muijen, “High expression of immunotherapy candidateproteins gp100, MART-1, tyrosinase and TRP-1 in uvealmelanoma,” British Journal of Cancer, vol. 78, no. 9, pp. 1156–1161, 1998.

[45] V. Appay, D. C. Douek, and D. A. Price, “CD8+ T cell efficacyin vaccination and disease,” Nature Medicine, vol. 14, no. 6, pp.623–628, 2008.

[46] J. B. Huppa, M. Gleimer, C. Sumen, and M. M. Davis,“Continuous T cell receptor signaling required for synapsemaintenance and full effector potential,” Nature Immunology,vol. 4, no. 8, pp. 749–755, 2003.

[47] K. C. Parker, M. Shields, M. DiBrino, A. Brooks, and J.E. Coligan, “Peptide binding to MHC class I molecules:implications for antigenic peptide prediction,” ImmunologicResearch, vol. 14, no. 1, pp. 34–57, 1995.

[48] K. C. Parker, M. A. Bednarek, L. K. Hull et al., “Sequencemotifs important for peptide binding to the human MHC classI molecule, HLA-A2,” Journal of Immunology, vol. 149, no. 11,pp. 3580–3587, 1992.

[49] D. R. Madden, D. N. Garboczi, and D. C. Wiley, “The antigenicidentity of peptide-MHC complexes: a comparison of theconformations of five viral peptides presented by HLA-A2,”Cell, vol. 75, no. 4, pp. 693–708, 1993.

[50] I. Miconnet, C. Servis, J. C. Cerottini, P. Romero, and F.Levy, “Amino acid identity and/or position determines theproteasomal cleavage of the HLA-A∗ 0201-restricted peptidetumor antigen MAGE-3,” The Journal of Biological Chemistry,vol. 275, no. 35, pp. 26892–26897, 2000.

[51] S. Pascolo, M. Schirle, B. Guckel et al., “A MAGE-A1 HLA-A∗0201 Epitope identified by mass spectrometry,” CancerResearch, vol. 61, no. 10, pp. 4072–4077, 2001.


Review Article

Tumor Antigen-Dependent and Tumor Antigen-IndependentActivation of Antitumor Activity inT Cells by a Bispecific Antibody-Modified Tumor Vaccine

Philippe Fournier1 and Volker Schirrmacher1, 2, 3

1 German Cancer Research Center (DKFZ), Im Neuenheimer Feld 280, 69120 Heidelberg, Germany2 Tumor Immunology Program, DKFZ, Im Neuenheimer Feld 280, 69120 Heidelberg, Germany3 Medical Centre for Immunology and Oncology (IOZK), 50674 Cologne, Germany

Correspondence should be addressed to Volker Schirrmacher, [email protected]

Received 1 July 2010; Accepted 14 December 2010

Academic Editor: Stuart Mannering

Copyright © 2010 P. Fournier and V. Schirrmacher. This is an open access article distributed under the Creative CommonsAttribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work isproperly cited.

New approaches of therapeutic cancer vaccination are needed to improve the antitumor activity of T cells from cancer patients. Westudied over the last years the activation of human T cells for tumor attack. To this end, we combined the personalized therapeutictumor vaccine ATV-NDV—which is obtained by isolation, short in vitro culture, irradiation, and infection of patient’s tumorcells by Newcastle Disease Virus (NDV)—with bispecific antibodies (bsAbs) binding to this vaccine and introducing anti-CD3(signal 1) and anti-CD28 (signal 2) antibody activities. This vaccine called ATV-NDV/bsAb showed the unique ability to reactivatea preexisting potentially anergized antitumor memory T cell repertoire. But it also activated naive T cells to have antitumorproperties in vitro and in vivo. This innovative concept of direct activation of cancer patients’ T cells via cognate and noncognateinteractions provides potential for inducing strong antitumor activities aiming at overriding T cell anergy and tumor immuneescape mechanisms.

1. Introduction

For decades, treatment of cancer has focused primarily onsurgery, chemotherapy, and radiation. Despite significantadvances by the introduction of new chemotherapeuticagents and also recently by the clinical introduction of mono-clonal antibodies, major limitations of such treatments keepthe inability to eliminate the last tumor cell. The offspringof those tumor cells that were not destroyed by the first-linetreatment may have a selective advantage, leaving the patientwith a recurrence of cancer that is often widespread and resis-tant to further chemotherapy or radiotherapy. Therefore,more effective therapies are needed. Immunotherapy basedon antitumor immune memory is a new modality for cancertreatment. It holds great promise for affecting in a positiveway cancer patients’ survival with minimal toxicity.

For over 200 years, active immunotherapy approacheshave been used to successfully prevent numerous infectious

diseases such as smallpox. These active immunotherapyconcepts are now being applied to develop therapeuticcancer vaccines with the intention of treating existingtumors and/or preventing tumor recurrence. In principle,anticancer vaccination (e.g., with autologous tumor cells,peptide vaccines, dendritic cells, idiotypic antibodies, andvirus-based vaccines) is a meaningful additional approachfor treatment of cancer [1].

Tumor antigens (TAs) of patients’ tumor cannot berecognized directly by the patients’ T cells. They need firstto be processed and to be properly presented by specializedcells that are known as professional antigen-presenting cells(APCs) such as dendritic cells (DCs). TA-presenting DCsthen migrate to lymph nodes, where they induce immunityin TA-specific naive T cells. This results in the differentiationinto effector T cells—mainly CD8+ cytolytic T cells (CTLs),which are capable of destroying tumor cells expressing thetumor antigen. The response also leads to the generation


of TA-specific memory T cells which provide immuneprotection against tumor recurrence.

Despite promising results of cancer vaccination obtainedin animal tumor models, results of published vaccine trialsreveal only a weak clinical response rate with less than 1%for active specific immunization procedures in colorectalcancer patients [2]. At the Surgery Branch of the NationalCancer Institute (Bethesda, Maryland, USA), an objectiveresponse rate of only 2.6% was reported [3] with variouscancer vaccines, even though about 50% of the vaccinatedpatients had developed CTL killer cells able to specificallyrecognize and kill tumor cells in vitro.

Difficulties met by vaccination approaches to cancertreatment have been attributed to tumor immune avoidancemechanisms [4]. Tumors employ many escape strategiesin order to evade immune attack. These strategies includedownregulation of MHC molecules in order to hide fromimmune recognition [5], expression of inhibitory factors andimmunosuppressive cytokines [6–8], including TGF-β [9,10], IL-10 [11], and recruitment of regulatory immune cellsCD4+CD25+FoxP3+ Tregs [12], Tr1 cells [13], tolerogenicDCs, and myeloid suppressor cells, including immaturemacrophages, granulocytes, DCs, and other myeloid cellsat earlier stages of differentiation [14, 15]. These immuneavoidance mechanisms employed by tumors render theimmune system tolerant. This may be responsible for tumorimmune evasion as many of the tolerance mechanisms thatprevent autoimmunity are the same as employed by tumorsto prevent immune destruction [16, 17].

In order to develop an effective immunotherapy strategyfor metastatic cancer, new approaches are required that notonly can create and enhance tumor-specific immunity butcan also counteract the ability of the tumor to evade immunedestruction. To this end, T cells of the cancer patients need tobe educated to attack tumor cells. Naive CD8+ T cells requiretwo distinct signals for activation: signal 1 is provided byengagement of the TCR with its cognate ligand, and signal2 is provided by interaction of costimulatory receptors withtheir respective ligands on the APCs [18, 19]. Memory CD8+T cells, which have been primed to TA, are often anergic andneed to be properly reactivated in order to be able to destroythe tumor cells.

The design of an efficient antitumor vaccine may beinfluenced by an important paradigm shift in the field ofimmunology regarding the regulation of immunity. A newconcept has emerged that proposes that the regulation ofimmunity and tolerance is not only determined by thespecificity of immune T cells as previously thought butalso by the context in which the antigens are presentedto the immune system [20, 21]. The implications are that,in the absence of appropriate inflammatory reactions, theself- (tumor) antigens presented by APCs will not leadto T cell activation. Since tumors can also produce anti-inflammatory cytokines, they are capable of influencing theimmune response by preventing an inflammatory response.

Therefore, successful antitumor immunity will developonly in situations where DCs are processing TAs in thepresence of an inflammatory microenvironment (“dangersignals”) which is potent enough to also downregulate

tumor-mediated immunosuppressive cytokine production.The magnitude and duration of the immune responsewill be dependent on the extent and quality of the localinflammatory response and will be contained by a variety ofexisting tolerogenic mechanisms.

Previous attempts at developing therapeutic cancer vac-cines have demonstrated that it is possible to elicit specificimmunity against self-tumor antigens [2, 3]. Recent insightson how immunity and tolerance are regulated indicate thatthe failure of these vaccines in the clinic may be related to theabsence of sufficient danger and T cell costimulation signalsat the time when tumor antigens are processed by DCs.

In this paper, we highlight some in vitro and in vivoobservations made during the evaluation of a tumor vaccinethat we developed in our laboratory. The tumor vaccine ofthe second generation, modified with bsAb, will be shownto be capable to reactivate memory T cells and to activatenonspecifically naive T cells against the tumor.

2. The Autologous NDV-Based Tumor Vaccine

Over the last 10 years, we have developed and evaluatedan autologous tumor vaccine which is first modified byvirus infection and which later was modified further byattachment of bispecific antibodies (see Figure 1). The aimwas to activate with such a vaccine potentially anergizedTA-specific memory T cells and to activate in additionnonspecifically naive T cells to overcome tumor escapevariants that may lack TA expression. For virus infection,we chose the avian paramyxovirus Newcastle Disease Virus(NDV) [22]. NDV is one of five species of viruses thatare under clinical evaluation [23]. It is a negative strandRNA virus with interesting antineoplastic and immune-stimulating properties [23, 24]. Most remarkable is itscapacity to induce strong type I interferon responses by viralprotein [24] and RNA [25]. Detection of foreign RNA inthe cytoplasm by RIG-I induces an innate antiviral programthat initiates the transcription of RNA-responsive genes.The responses involve a multimodal machinery of generegulation by the Interferon Regulatory Factor (IRF) familyof transcription factors [26] and link innate and adaptiveimmunity [27]. There are 2 generations of NDV-basedtumor vaccine: the ATV-NDV and ATV-NDV/bsAb.

2.1. First-Generation Vaccine: ATV-NDV. The virus-mod-ified tumour vaccine developed by us for human applica-tion consists of virus-infected intact viable and irradiatedautologous tumor cells (see Figure 1). The tumor cellinfection by NDV is designed to provide the necessarydanger signals to elicit antitumor immunity. This strategy isbased on preclinical studies in metastatic animal tumours.Antimetastatic effects were observed after local postoperativevaccination with NDV-infected autologous tumour cells[28]. The vaccination activated a tumour-line specific T cell-mediated immune response, which also protected against asecond challenge with the same tumour line [29].

Tumor cell infection by NDV was found also inhumans to be an efficient and safe way to produce anautologous tumor vaccine (first-generation ATV-NDV) with


vaccine of the 2nd generation has

NDV

1st generation vaccine (ATV-NDV)

NDV :

universal anchor molecule

humans

Prolongation of survivalamong cancer patients forvarious tumor entities [31-37]

2nd generation vaccine (ATV-NDV /bsAb)

vaccine for cross-linking of CD3 and

CD3 complex ( 1)

(signal

signal

2)

Advantages:

attachment to the vaccine

cancer cell which issusceptible to NDV infection)

delivery of T cell activatingsignals

attachment of the bsAbs

cells in vitro and inxenotransplanted mice in vivo

with the ATV-NDV tumor

been initiated in colorectalcancer patients and the clinicalresults are pending .

bsHN-CD3

Anti-HN

Anti-HN Anti-CD3

Anti-CD28

bsHN-CD28

HN

TA

cells

and interacts with TCR-associated

the costimulatory molecule CD28

future: a clinical phase I/II trial

CD28 molecules at the surface of T

Bispecific Abs :

expressed at the surface of the vaccine

(i) Tumor selective replication(ii) Introduction of viral HN as

(iii) High safety profile for

(i) Studies in mouse models

(ii) Preclinical studies

(iii) Clinical studies

(i) Production as recombinant proteins(ii) Attachment of antibodies to the

(iii) bsHN-CD3 binds to HN of NDV

(iv) bsHN-CD28 binds to HN and to

(i) Easy production(ii) Safety(iii) Monovalency(iv) BsAb active only upon

(v) Broad applicability (any

(vi) Controlled dosage and

(vii) Easy protocol for the specific

(viii) Preclinical studies with human

[28, 29]

[30]

Tumor cell(107cells)

(ix) Clinical studies planned for the

Figure 1: Principles of the NDV-based tumor vaccine of the first and second generation and status of the art (for more details, see the maintext).

pleiotropic immuno-stimulatory properties [30]. Promisingresults based on the prolongation of survival among cancerpatients with various tumor entities have already beenreported from several clinical phase II trials, including breastcancer [31, 32], colon carcinoma [33, 34], Head and NeckSquamous Cell Carcinoma (HNSCC) [35], and glioblastoma[36]. The antitumor clinical efficacy has been shown alsoin a randomized study performed among colon carcinomapatients operated for liver metastasis [37].

2.2. Second-Generation Vaccine: ATV-NDV/bsAb. The useof NDV during the generation of the ATV-NDV tumorvaccine presents the advantage that, upon infection, the viralhemagglutinin-neuraminidase (HN) protein is expressed onthe vaccine cells and can serve as universal anchor moleculefor the binding of new ligand proteins. The ATV-NDV canthen be combined easily with bispecific antibodies (bsAbs)binding to the HN protein for introducing anti-CD3 andanti-CD28 antibodies at the surface of the tumor vaccineto obtain the NDV-based tumor vaccine of the secondgeneration (see Figure 1). This novel strategy is designed tointensify T cell activation via agonistic anti-CD3 and/or anti-CD28 single-chain antibody reagents (scFv). When subop-timal amounts of anti-CD3 are employed, the combinationof TA, anti-CD3, and anti-CD28 should help to intensifyTA-induced signal 1 and HN-induced costimulatory signal

2 [38, 39]. Because of virus infection, the ATV-NDV/bsAbvaccine provides a highly inflammatory environment for Tcells in the presence of tumor cells. The presence of cellsurface bsAbs binding to CD3 and CD28 receptor moleculesserves among other effects for augmenting signal strength inT cells to override anergic states of TA-specific T cells. Signalintensity and duration (strength) of TCR stimulation has animpact on setting the balance between adaptive responsesand immunopathology [40] and influences induction of Tcell activation or anergy [41]. Signal strength, timing, andtuning are also important for T cell costimulation (signal2). Combining optimal signals 1 and 2 at the surface ofthe tumor vaccine is expected to generate strong antitumoradults. We showed over the last years that this vaccine actson TA-specific memory T cells but also on naive T cells viaTA-dependent and TA-indepentent pathways.

3. Reactivation of TA-Specific Memory T Cellsfrom Cancer Patients upon OptimalCombination of bsHN-CD3 and bsHN-CD28with the Vaccine ATV-NDV

The ATV-NDV tumor vaccine of the second generationwas capable of reactivating anergic T cells from tumor-draining lymph nodes of cancer patients. This was revealed


via a modified short-term IFN-γ ELISpot assay which weestablished for reactivation of cancer-reactive memory Tcells. As shown in Figure 2(A), from four different head andneck squamous cell carcinoma (HNSCC) patients, primarytumor cells were expanded in vitro, and, from each tumorline, autologous tumor vaccines were prepared as a vaccine(through irradiation, infection by NDV, and loading withthe bsAbs). These vaccines were finally combined with Tcells isolated from tumor-draining lymph nodes from thecorresponding HNSCC patients. The data from Figure 2(B)show a strong IFN-γ response only in that group whereT cells were stimulated with autologous TAs together withanti-CD3 and anti-CD28 signals. These responses requiredHNSCC-derived TAs. Another tumor line (the heterologouspromonocytic U937 cell line) modified following the sameway could not reactivate a short-term memory response fromthe patients’ T cells (Figure 2 (B), lower part).

We conclude

(i) that tumor-reactive memory T cells from draininglymph nodes of HNSCC patients could not beactivated with the first-generation vaccine ATV-NDV,

(ii) that the cells, however, could become efficientlyreactivated using the same vaccine together withoptimized signals 1 and 2, and

(iii) that a similarly modified vaccine from an unrelatedtumor cell line was not capable to elicit such amemory T cell response.

The fact that the ATV-NDV vaccine without bsAbs wasnot capable to reactivate a response might be explained byan anergic state of the tumor-draining lymph node-derivedT cells.

4. Activation of Antitumor Activity from Naive TCells by the Second-Generation Vaccine

The tumor vaccine ATV-NDV/bsAb was shown to havealso an effect on naive T cells which were obtained fromtotal T cells of normal healthy donors by removal ofthe CD45RO+ cells. This was observed by analyzing theactivation of the naive T cells towards tumor cells after 6 daysof incubation with various vaccines. Figure 3(A) shows CD25and CD45RO expression on naive T cells upon coincubationwith various NDV-based tumor vaccines. 1 × 105 naive Tcells from a normal healthy donor were labeled with CFSE.They were then coincubated with 1 × 104 irradiated MCF-7 cells which were modified with 100 HU NDV Ulster andwith suboptimal bsHN-CD3 (500 pg/well) and 84.4 U/wellbsHN-CD28 (d). Naive T cells, which came in contact onlywith the vaccine (a) or with the vaccine loaded with each ofthe constructs separately (b and c) served as controls. Aftersix days, the cells were harvested, blocked with Endobulin,and then stained with anti-CD25-APC (a) or anti-CD45RO-PE (b) before being analyzed on the FACS. Dead cellswere excluded by propidium iodide staining. The red boxesindicate actively dividing (CFSE low) cells expressing CD25or CD45RO. The frequencies are indicated as numbers in %(adapted from [42]). We observed a strong activation only

when both bsAbs (bsHN-CD3 and bsHN-CD28) were addedto the tumor vaccine (see Figure 3(A), (d)). Figure 3(B)shows that tumor growth inhibition induced in naive T cellsupon coincubation with various NDV-based tumor vaccines.In order to test the effectiveness of the new vaccine strategyto activate in vitro naive T cells, an in vitro assay that wecalled tumor neutralization assay (TNA) was developed. Itconsists of an adherent tumor cell monolayer (here humanMCF-7 breast cancer cells) in a cell culture plate in whichγ-irradiated NDV-modified vaccine cells, fusion proteins,and naive T cells of a healthy donor as effector cells arebrought into contact. During the incubation period of five toseven days, the vaccine cells activate the effector cells whichincrease their cytotoxic potential. During the effector phase,the “bystander” tumor cells are lysed, or their growth isinhibited. This results in a decrease in the number of livetumor cells in monolayer (when compared to the controls).After removal of the nonadhering remaining effector cells,the number of surviving tumor cells can be quantified withMTS as dye reagent for measuring the amount of viabletumor cells per well. We observed that the tumor vaccinewith the 2 bound bsAbs bsHN-CD3 and bsHN-CD28 was byfar the most efficient in this assay among all the combinationsof tumor vaccine and bsAbs tested (Figure 3(B)). We alsoobserved that the extent of T cell-mediated tumor growthinhibition in vitro depended on an optimal amount of bsHN-CD3 and bsHN-CD28 present at the surface of the tumorvaccine (data not shown, see [42]). Titration curves revealedfor each of the recombinant proteins, upon attachment tothe vaccine, a low dose optimum, of T cell stimulating orcostimulating activity [42].

An important aspect of T cell effector activity relates toits duration. To test this, we activated purified T cells onceby coculture with the second-generation vaccine, performedthe TNA, and then repeatedly transferred the T cells fromthe suspension above the destroyed monolayer onto freshtumor monolayers and followed their destruction thereafter.In these ways, live adherent MCF-7 monolayers were coin-cubated in 96-well plates with MCF-7-NDV vaccine cells,purified T cells, a suboptimal dose of bsHN-CD3 (1 μgper 107 vaccine cells; signal 1), and one of the followingcostimulatory fusion proteins: bsHN-CD28 (signal 2a) ortsHN-IL-2-CD28 (signal 2a-2b), each at a concentrationof 84 U per 107 vaccine cells (passage 0). T cells activatedin the presence of suboptimal bsHNCD3 alone (negativecontrols) showed no tumor growth inhibition in this assay(data not shown). For serial passages, the same conditionswere performed in parallel using a 6-well TNA format withidentical protein concentrations (TNA passage plates). After7 days, the TNA test plates were developed using MTS asreagent to obtain the value of tumor growth inhibition asdescribed in [43]. The cells from the TNA passage plateswere harvested, washed, and transferred for another 3 daysonto fresh MCF-7 monolayers, either in 96-well TNA testplates (passage 1) or in 6-well TNA passage plates for anotherround (passage 2). The results (Figure 3(C)) revealed a muchlonger duration of bystander antitumor activity in T cellsactivated by the vaccine bound bsHN-CD3 and bsHN-CD28(group b) than in T cells activated by the vaccine alone


HNO patients

Primary tumor

Tumor vaccine

PBMC

Coculture for40 hours

T cells

Tumor cell culture

CD28CD3

(A)

CD28CD28

CD28

CD3

CD28

CD3

CD3

CD3

35 87 39 116

(a) (b) (c) (d)

(e) (f)

Mean =

(g) (h)

Tce

lls

Vaccine (autologous with TA for memory T cells of patient (red), U937 without right TA (green))

TA

Patient no. 1

Patient no. 1

Patient no. 2

Patient no. 3

Patient no. 4

(B)

Figure 2: Reactivation of tumor-reactive memory T cells of cancer patients (here HNO cancer patients) by the ATV-NDV tumor vaccineof the 2nd generation. (A) Protocol for the ex vivo stimulation of T cells from cancer patients by the NDV-based tumor vaccine of thesecond generation in an autologous setting. Purified T cells isolated from tumor-draining lymph nodes (LN) of HNSCC patients weretested for their capacity to be restimulated and to produce IFN-γ upon contact with various combination of the ATV-NDV tumor vaccine(generated from the autologous tumor) and bsAbs molecules. For that, they were coincubated for 40 hours with the indicated autologousvaccine formulations (tumor) (n = 4). For specificity control, the unrelated human promonocytic tumor cell line U937 (U937) was modifiedidentically and used to reactivate the patients’ T cells. (B) IFN-γ ELISPOT results from A. Mean: mean number from triplicates of 4 patientsspot forming T cells per 1 million cells. Each patient’s T cells were stimulated with four different formulations of either autologous (a–d) orheterologous (e–h) vaccine. T cells of patients no. 2, no. 3, and no. 4 also did not react to the heterologous vaccine (data not shown) (adaptedfrom [42]).


CD

45R

OC

D25

(a) (b) (c)

CFSE

100

100

101

101

102

102

103

103

104

104

100

100

101

101

102

102

103

103

104

104

100

100

101

101

102

102

103

103

104

100

101

102

103

104

100

101

102

103

104

104

100

100

101

101

102

102

103

103

104

104

100 101 102 103 104

100

101

102

103

104

100 101 102 103 104

100 101 102 103 104

100

101

102

103

104

100 101 102 103 104

(d)

CD3 CD3CD28 CD28

2

1.5

9.9

9.3

4.7

4.2

65.1

63.9

Coculture of naive T cells with

(A)

TG

I(%

)

0

20

40

60

80

100

(B)

TG

I(%

)

0

20

40

60

80

100

0

20

40

60

80

100

0

20

40

60

80

100

1 2 3Passage number

1 2 3Passage number

1 2 3Passage number

(a) (b) (c)

CD3CD3

CD28

Coculture of naive T cells with

CD28 + IL-2

(C)

Figure 3: Activation of naive T cells towards tumor cells by the NDV-based tumor vaccine combined with bsAbs in different combinations.

(group a). We have constructed also a trispecific recombinantfusion protein (tsHN-IL-2-CD28) in which the cytokineinterleukin-2 was linked in between the anti-HN and theanti-CD28 scFv binding sites [42]. Since we had observedthat vaccine-bound IL-2 can deliver costimulatory signalvia CD25 (the IL-2 receptor alpha chain) to T cells, we

were then interested to test whether the combination ofcostimulatory signals delivered via anti-CD28 and via IL-2might have an advantage. As can be seen from the results inFigure 3(C), group c, this type of modified vaccine inducedantitumor effecter activity of the longest duration. Expressedquantitatively, a simultaneous introduction of costimulatory


signals via CD28 (signal 2a) and IL-2 receptor (signal 2b)led to a 20–40% increased bystander antitumor activity inpassage 1 (days 7–10) and passage 2 (days 10–13) whencompared to a tumor vaccine with only one costimulatorysignal (for more details, see [42]).

5. Human PBMC Preactivated In Vitro bythe NDV-Based Tumor Vaccine ofthe Second Generation Shows In VivoTherapeutic Efficiency upon AdoptiveTransfer to Tumor-Bearing Mice

To study possible immunotherapeutic effects of vaccine-activated PBMC in vivo, we used a NOD/SCID mouse modelwhich allows outgrowth of human MCF-7 breast cancercells [44]. For therapy, PBMCs from a healthy donor werepreactivated with an MCF-7-NDV tumor vaccine loaded ornot with bsHN-CD3 and bsHN-CD28 [42]. The preactivatedcells were then transferred via intraperitoneal injections intomice with established MCF-7 xenotransplants (Figure 4(A)).PBS-treated control mice showed progressive tumor growth,as did tumors from mice treated with PBMC preactivatedwith MCF-7-NDV tumor vaccine. The tumors of mice,however, treated with PBMCs, which were preactivated withthe NDV-based tumor vaccine of the second generation,started to regress 51 days after adoptive cell transfer andshowed at day 93 much smaller tumor diameters whencompared to tumors of mice treated with the NDV-basedtumor vaccine of the first generation (i.e., without the 2bsAbs) (see Figure 4(B)). Notably, treatment of mice withsupernatants from PBMC cultures preactivated with theNDV-based tumor vaccine of the second generation didnot lead to tumor regression [42]. Immunohistochemicalstainings of tumors that underwent regression revealed aheavy infiltration by both human CD4+ and CD8+ T cells(Figure 4(B)), a certain percentage of which were in anactivated state (CD69 positive) and exhibited a memoryphenotype (CD45RO positive) [42]. In contrast, tumorsections from mice treated with PBMC preactivated with theMCF-7-NDV tumor vaccine (Figure 4(B)) or with PBS (datanot shown) revealed no tumor-infiltrating T cells [42].

In summary, human naive T cells upon stimulation withthe second-generation tumor vaccine and upon transfer invivo can—after a certain lag period—infiltrate human tumortissue and mediate tumor regression.

6. Conclusion and Perspectives

The main question we address in this paper: how can weexploit maximal antitumor activity from T cells of cancerpatients through antitumor vaccination?

Over the last decades, many types of vaccines have beendeveloped towards this goal, including synthetic peptides,“naked” DNA, dendritic cells, or recombinant viruses. Ourapproach is based on the use of patient-derived tumor cellswhich are irradiated, modified by infection with NDV, andcoupled with bispecific Abs in order to introduce new ligandsfor T cell activation at the surface of the tumor vaccine.

We observed that the NDV-based autologous tumorvaccine of the second generation can be used for reactivationof apparently anergic memory T cells from tumor-draininglymph nodes of individual cancer patients. The danger andcostimulatory signals introduced to the vaccine throughvirus infection as well as additional signal 1 and 2 at thesurface of the tumor vaccine appear necessary in cancerpatients whose T cells exhibit a high degree of unrespon-siveness to stimulation to TA. The underlying mechanismexplaining such antitumor activity is suggested to rely ondirect presentation of autologous TAs (cognate interactionswith memory T cells (see Figure 5)) and on augmentation ofsignal intensity by bsAbs.

The absence of any response in the short-time Interferon-γ Elispot (see Figure 2, lower part) during the coincubationof patient T cells and heterologous bsAb-modified tumorcells might be explained by the absence of autologous TA.

What is also interesting is the capacity of the NDV-basedtumor vaccine of the 2nd generation to induce antitumoractivity in naive T cells. Such activation of naive T cellsrequires, however, a longer time period than that of memoryT cells. The proposed mechanism is direct T cell activationvia noncognate interactions with the vaccine/bsAb leadingto induction of strong bystander antitumor activity (seeFigure 5 and Table 1). These latter observations suggest thatT cells can be activated to exert antitumor activity by such anew tumor vaccine in a TA- and MHC-independent pathway,similar to cells of the innate immune system. This newmechanism may become an important safeguard againsttumor immune escape.

The cells of the adaptive immune system utilize somati-cally rearranged receptors to recognize antigens. By contrast,cells of the innate immune system primarily use germline-encoded receptors to defend against infected or transformedcells. Interestingly, cells of the adaptive immune systemcan express also some of these germline-encoded receptors[45].

Can T cells, in similarity to cells of the innate immunesystem, sense danger? The first published “danger model” ofimmunity [46] proposed only one mechanism for immunerecognition of danger that perceived by DCs upon releaseof cellular contents following necrosis of a diseased cell inits neighbourhood. This model predicts a superior effect ofa lytic as opposed to a nonlytic virus in the treatment oftumors, because tumor cells necrotically destroyed by thevirus would be phagocytosed and perceived as dangerous byDCs. The DCs would (i) process TAs, (ii) become activated,and (iii) present processed TA peptides to T cells for cognateinteraction and immune response induction.

Another recent model [47] suggests that T lymphocytesthemselves correlate danger signals to antigen. This hypoth-esis associates danger also with nonlytic viruses (as NDVUlster) if these are upregulating danger signals in their hostcells. Such an event will quickly cause its host cells to bekilled by the immune system. Killed infected tumor cellsare likely to result in TA being presented by DC along withpotent costimulation. Recently, it was shown that dsRNA inthe apoptotic bodies of virus-infected dead cells is recognizedby CD8 alpha+ DCs that have high expression of toll-like


(a) Induction of a tumor through s.c.injection of tumor cells

NOD/SCID mice with a human breast

(b) Activation of human PBMC throughincubation with the tumor vaccine

cells(human PBMC)

Tumor vaccine

(c) Adoptive transfer of activated T cells in tumor-bearing mice

i.p.

Activated effector cells(human PBMC)

of the tumorgrowth over time:

CD28CD3

tumor (MCF-7) “xenograft”

(c) Followup

Effector

(A)Vaccine used for the activation of the PBMC

02468

1012

Tum

ordi

amet

erat

day

93(m

m)

His

toch

emis

try

(CD

8st

ain

ing)

CD28CD3

(a) (b)

xx x

xx x

xx x

xx x

x x

x x

(B)

Figure 4: In vivo antitumor activities of human PBMC preactivated in vitro by the NDV-based tumor vaccine of the second generation. (A)Experimental protocol. MCF-7 tumor cells were injected subcutaneously into NOD/SCID mice (n = 8 per group). The mice were kept untilpalpable tumors were established (6–8 mm in diameter) (a). Then, at days 1, 4, and 7, the mice were treated by i.p. injection of 107 PBMCwhich were preactivated ex vivo for 3 days by the NDV-based tumor vaccine of the first or second generation obtained using the MCF-7 cellline (b). Tumor growth was then monitored over time (c). (B) Tumor diameter and representative immunohistochemistry images of tumortissue sections. Tumor-bearing mice were treated with PBMC activated with MCF7-NDV (left) or with PBMC preactivated with MCF-7-NDV loaded with bsHN-CD3 and bsHN-CD28 (right) and were analyzed over time for tumor growth, and data at day 93 is represented(top). At this time point, some mice were sacrificed, and the tumor sections were stained with mAbs against the human CD8 antigen andanalyzed by fluorescence microscopy (bottom). Scale bar, 100 μm (adapted from [42]).

receptor 3 (TLR-3) [48]. This promotes cross-priming of Tcells to virus-infected cells [49].

Recent data support a role for CD8+ T cells ininnate immune responses, independent of TCR specificity.Under certain circumstances, antigen non-specific TCR-independent responses of CD8+ T cells play a beneficial role

in controlling tumors [50]. Marsland et al. [51] showed thatinnate signals driven by DC can compensate for the absenceof PKC-Φ, which is important for TCR signalling, duringCD8+ T cell effector and memory responses in vivo.

The discovery of toll-like receptors (TLRs) and morerecently of cytosolic innate immune sensors such as


T cellsfrom

cancer patients

Memory:(primed,

possibly anergic)

Naive:(not primed)

(a) Direct presentation

Cognateinteractions

(c) Direct presentation

interactions

human PBMC antitumor activity that can be

Evidence

See results

See results

[42]

[42]

T cells

Newly activatedT cells

Non-TA specific

Anti-HN

Anti-HN

HNAnti-CD3

Anti-CD28

ATV-NDV

NDV

-bsAb Vaccine

- Breaking tolerance

- Inducing through co-incubation with

+ Additional signals

from Figure 4

from Figure 2

from Figure 3[42 ,43]

TA specific

(b) Cross-presentation

Noncognate

Effector/memory

See [43] and resultsobserved in vitro and in vivo

Figure 5: Cognate and noncognate antitumor mechanisms induced by the NDV-based tumor vaccine of the second generation (ATV-NDV/bsAbs). The different mechanisms of T cell activation against tumor suggested explaining the efficacy of the ATV-NDV/bsAb ininducing antitumor activity in T cells are highlighted. (a) Direct antigen presentation of TA to memory T cells. Addition of danger signalsvia NDV infection of the tumor cells during the elaboration of the tumor vaccine induces a reactivation of the antitumor reactivity in TA-specific memory T cells [32]. This mechanism is intensified by the addition of the 2 bsAbs introducing binding activities to CD3 and CD28at the surface of the tumor vaccine (see Figure 2). + signal intensification via bsAb (memory T cells). (b) Cross-presentation via DCs in vivo(naive T cells). (c) Direct polyclonal activation this pathway is highlighted in this manuscript.

Table 1: Summary of the types of interactions, mechanisms, characteristics of the immune responses induced by the ATV-NDV/bsAb vaccinein memory and naive T cells and the eventual danger of autoimmunity (see the main text for more details).

Vaccine + bsAbsInteractions

T cells-tumor cellsMechanisms Response characteristics Autoimmunity

TA + bsCD3 + bsCD28 CognateRestimulation (memory)/TAcross-presentation via DCs (naive)⇒ TA-specific T cell activation

Long-term effectsystemic

Low risk

bsCD3 + bsCD28 NoncognateProvision of inflammatory anddanger signals⇒ Non-TA-specific Tcell activation

Transient, local High risk

RIG-I-like receptors (RLRs) and nod-like receptors (NLRs)as components that recognize structures of “danger” such aspathogens has greatly advanced understanding of how innateimmune response can be triggered and can prime antigen-specific adaptive immunity. Cytotoxic CD8+ T cells and Thelper 1 (Th1) cells are central to effective immune responsesagainst tumor. However, clinical trials with cancer vaccineshave shown the weakness of responses observed among thetreated cancer patients, although cytotoxic activities specificfor the targeted TA were detected in vivo [3].

A key determinant in the induction of a strong andefficient cellular immune response against tumors seems to

be, in addition to a broad repertoire of TA presented by thetumor cells used as vaccine, the recognition of the TA as“nonself.” Such TAs have been termed “unique” (individuallytumor specific) and have a potential to serve as tumorrejection antigen. The superiority of autologous tumorvaccines among tumor vaccines evaluated in randomizedclinical studies [52] suggests that unique tumor antigensare indeed particularly important in generating responsiveT cells for a therapeutic effect. Restimulation of TA-specificmemory T cells and activation of naive T cells may explainthe strong antitumor potential of the NDV-based and bsAb-modified tumor vaccine.


CD8+ T cells have been implicated not only in antigen-dependent but also in antigen-independent antitumorresponses. Self-reactive CD8+ T cells expressing high levelsof NKG2D induce killing of target cells using a redirectedlysis assay [53]. Furthermore, activation of CD44hi CD8+ Tcells using IL-2 resulted in significantly higher levels of killingof a syngeneic target compared with CD44loCD8+ T cells[54]. Expression of the NKG2D ligand Rae-1δ resulted inincreased killing of the syngeneic targets. As the expressionof Rae-1δ is associated with tumors, it seems plausiblethat a subset of innate CD8+ T cells activated with IL-2could directly lyse tumors using NKG2D, independent ofTCR specificity. Another report characterized CD8+ T cellsin cancer patients receiving IL-12, showing that a subsetof CD8+ T cells expanded and displayed enhanced MHCnonrestricted cytotoxicity in vitro [55]. This suggests thatcytokines can stimulate subsets of CD8+ T cells, independentof TCR signalling, to lyse tumors. In this context, it isinteresting to see an example of clinical application: thetransfusion of intentionally mismatched donor lymphocytesin high-risk chemotherapy-resistant patients with metastaticsolid tumors and haematological malignancies [56].

Danger signals and T cell costimulation have greattherapeutic potential but need to be optimized and con-trolled. Unfortunately, these same types of innate antigen-independent responses might be harmful when they are self-reactive. The risk factor for induction of autoimmune disease(see Table 1) has been discussed in more detail in a recenteditorial [57].

In summary, the tumor vaccine ATV-NDV/bsAb showsdual activity: (i) it can activate TA-specific memory T cellsfrom cancer patients, and (ii) it can generate antitumor activ-ity from naive T cells. A strategy harnessing both arms of theimmune system—innate (based on noncognate interactions)and adaptive (relying on cognate interactions)—holds greatpotential for clinical applications in cancer patients.

Our findings open new ways of investigations to manip-ulate T cell activity against tumor cells and to exploit thefull power of the immune system through reactivation oftumor antigen-specific memory T cells and through de novoactivation of naive T cells. These could, for example, bebased on the combination of new recombinant immunos-timulatory molecules such as bi- or trispecific antibodies orimmunocytokines with vaccines which have already provenclinical effectivity.


The authors declared that no competing financial interestsexist.

References

[1] A. W. Silk and O. J. Finn, “Cancer vaccines: a promising cancertherapy against all odds,” Future Oncology, vol. 3, no. 3, pp.299–306, 2007.

[2] D. Nagorsen and E. Thiel, “Clinical and immunologicresponses to active specific cancer vaccines in human colorec-tal cancer,” Clinical Cancer Research, vol. 12, no. 10, pp. 3064–3069, 2006.


[4] L. Rivoltini, P. Canese, V. Huber et al., “Escape strategiesand reasons for failure in the interaction between tumourcells and the immune system: how can we tilt the balancetowards immune-mediated cancer control?” Expert Opinionon Biological Therapy, vol. 5, no. 4, pp. 463–476, 2005.

[5] C. Demanet, A. Mulder, V. Deneys et al., “Down-regulationof HLA-A and HLA-Bw6, but not HLA-Bw4, allospecificitiesin leukemic cells: an escape mechanism from CTL and NKattack?” Blood, vol. 103, no. 8, pp. 3122–3130, 2004.

[6] G. Filaci and N. Suciu-Foca, “CD8+ T suppressor cellsare back to the game: are they players in autoimmunity?”Autoimmunity Reviews, vol. 1, no. 5, pp. 279–283, 2002.

[7] H. Jonuleit and E. Schmitt, “The regulator T cell family: dis-tinct subsets and their interrelations,” Journal of Immunology,vol. 171, no. 12, pp. 6323–6327, 2003.

[8] A. Ben-Baruch, “Inflammation-associated immune suppres-sion in cancer: the roles played by cytokines, chemokines andadditional mediators,” Seminars in Cancer Biology, vol. 16, no.1, pp. 38–52, 2006.

[9] R. L. Elliott and G. C. Blobe, “Role of transforming growthfactor beta in human cancer,” Journal of Clinical Oncology, vol.23, no. 9, pp. 2078–2093, 2005.

[10] V. C. Liu, L. Y. Wong, T. Jang et al., “Tumor evasion ofthe immune system by converting CD4CD25 T cells intoCD4CD25 T regulatory cells: role of tumor-derived TGF-β,”Journal of Immunology, vol. 178, no. 5, pp. 2883–2892, 2007.

[11] J. Zhou, F. Ye, H. Chen, W. Lv, and N. Gan, “The expressionof interleukin-10 in patients with primary ovarian epithelialcarcinoma and in ovarian carcinoma cell lines,” Journal ofInternational Medical Research, vol. 35, no. 3, pp. 290–300,2007.

[12] W. Zou, “Immunosuppressive networks in the tumour envi-ronment and their therapeutic relevance,” Nature ReviewsCancer, vol. 5, no. 4, pp. 263–274, 2005.

[13] X. Zhang, H. Huang, J. Yuan et al., “CD48 dendritic cells primeCD4 T regulatory 1 cells to suppress antitumor immunity,”Journal of Immunology, vol. 175, no. 5, pp. 2931–2937, 2005.

[14] M. Terabe, S. Matsui, J.-M. Park et al., “Transforming growthfactor-beta production and myeloid cells are an effector mech-anism through which CD1d-restricted T cells block cytotoxicT lymphocyte-mediated tumor immunosurveillance: abro-gation prevents tumor recurrence,” Journal of ExperimentalMedicine, vol. 198, no. 11, pp. 1741–1752, 2003.

[15] Y. Nefedova, M. Huang, S. Kusmartsev et al., “Hyperactivationof STAT3 is involved in abnormal differentiation of dendriticcells in cancer,” Journal of Immunology, vol. 172, no. 1, pp. 464–474, 2004.

[16] G. J. Prud’homme and C. A. Piccirillo, “The inhibitoryeffects of transforming growth factor-beta-1 (TGF-β1) inautoimmune diseases,” Journal of Autoimmunity, vol. 14, no.1, pp. 23–42, 2000.

[17] N. Pakravan, A. T. Hassan, and Z. M. Hassan, “Naturallyoccurring self-reactive CD4+CD25+ regulatory T cells: uni-versal immune code,” Cellular & Molecular Immunology, vol.4, no. 3, pp. 197–201, 2007.

[18] P. Bretscher, “The two-signal model of lymphocyte activationtwenty-one years later,” Immunology Today, vol. 13, no. 2, pp.74–76, 1992.

[19] C. H. June, J. A. Bluestone, L. M. Nadler, and C. B. Thompson,“The B7 and CD28 receptor families,” Immunology Today, vol.15, no. 7, pp. 321–331, 1994.


[20] P. Matzinger, “Tolerance, danger, and the extended family,”Annual Review of Immunology, vol. 12, pp. 991–1045, 1994.

[21] C. A. Janeway Jr., “The immune system evolved to discrimi-nate infectious nonself from noninfectious self,” ImmunologyToday, vol. 13, no. 1, pp. 11–16, 1992.

[22] D. J. Alexander, “Newcastle disease and other Paramyxoviridaeinfections,” in Diseases of Poultry, B. W. Calnek, H. J. Barnes, C.W. Beard, L. McDougald, and J. Y. M. Saif, Eds., pp. 541–569,Iowa State University, Ames, Iowa, USA, 10th edition, 1997.

[23] M. Aghi and R. L. Martuza, “Oncolytic viral therapies—theclinical experience,” Oncogene, vol. 24, no. 52, pp. 7802–7816,2005.

[24] J. Zeng, P. Fournier, and V. Schirrmacher, “Induction ofinterferon-α and tumor necrosis factor-related apoptosis-inducing ligand in human blood mononuclear cells byhemagglutinin-neuraminidase but not F protein of Newcastledisease virus,” Virology, vol. 297, no. 1, pp. 19–30, 2002.

[25] P. Fournier, J. Zeng, and V. Schirrmacher, “Two ways toinduce innate immune responses in human PBMCs: paracrinestimulation of IFN-α responses by viral protein or dsRNA,”International Journal of Oncology, vol. 23, pp. 673–680, 2003.

[26] T. Taniguchi and A. Takaoka, “The interferon-α/β systemin antiviral responses: a multimodal machinery of generegulation by the IRF family of transcription factors,” CurrentOpinion in Immunology, vol. 14, no. 1, pp. 111–116, 2002.

[27] A. Le Bon and D. F. Tough, “Links between innate andadaptive immunity via type I interferon,” Current Opinion inImmunology, vol. 14, no. 4, pp. 432–436, 2002.

[28] R. Heicappell, V. Schirrmacher, P. Von Hoegen, T. Ahlert,and B. Appelhans, “Prevention of metastatic spread by post-operative immunotherapy with virally modified autologoustumor cells. I. Parameters for optimal therapeutic effects,”International Journal of Cancer, vol. 37, no. 4, pp. 569–577,1986.

[29] V. Schirrmacher and R. Heicappell, “Prevention of metastaticspread by postoperative immunotherapy with virally modifiedautologous tumor cells. II. Establishment of specific systemicanti-tumor immunity,” Clinical and Experimental Metastasis,vol. 5, no. 2, pp. 147–156, 1987.

[30] V. Schirrmacher, C. Haas, R. Bonifer, T. Ahlert, R. Gerhards,and C. Ertel, “Human tumor cell modification by virusinfection: an efficient and safe way to produce cancer vaccinewith pleiotropic immune stimulatory properties when usingNewcastle disease virus,” Gene Therapy, vol. 6, no. 1, pp. 63–73, 1999.

[31] T. Ahlert, W. Sauerbrei, G. Bastert et al., “Tumor-cell numberand viability as quality and efficacy parameters of autologousvirus-modified cancer vaccines in patients with breast orovarian cancer,” Journal of Clinical Oncology, vol. 15, no. 4, pp.1354–1366, 1997.

[32] V. Schirrmacher, “Clinical trials of antitumor vaccination withan autologous tumor cell vaccine modified by virus infection:improvement of patient survival based on improved antitu-mor immune memory,” Cancer Immunology, Immunotherapy,vol. 54, no. 6, pp. 587–598, 2005.

[33] D. Ockert, V. Schirrmacher, N. Beck et al., “Newcastledisease virus-infected intact autologous tumor cell vaccine foradjuvant active specific immunotherapy of resected colorectalcarcinoma,” Clinical Cancer Research, vol. 2, no. 1, pp. 21–28,1996.

[34] V. Schirrmacher, T. Ahlert, T. Probstle et al., “Immunizationwith virus-modified tumor cells,” Seminars in Oncology, vol.25, no. 6, pp. 677–696, 1998.

[35] J. Karcher, G. Dyckhoff, P. Beckhove et al., “Antitumorvaccination in patients with head and neck squamous cellcarcinomas with autologous virus-modified tumor cells,”Cancer Research, vol. 64, no. 21, pp. 8057–8061, 2004.

[36] H. H. Steiner, M. M. Bonsanto, P. Beckhove et al., “Antitumorvaccination of patients with glioblastoma multiforme: a pilotstudy to assess feasibility, safety, and clinical benefits,” Journalof Clinical Oncology, vol. 22, no. 21, pp. 4272–4281, 2004.

[37] T. Schulze, W. Kemmner, J. Weitz, K. D. Wernecke, V.Schirrmacher, and P. M. Schlag, “Efficiency of adjuvant activespecific immunization with Newcastle disease virus modifiedtumor cells in colorectal cancer patients following resectionof liver metastases: results of a prospective randomized trial,”Cancer Immunology, Immunotherapy, vol. 58, no. 1, pp. 61–69,2009.

[38] C. Haas, C. Ertel, R. Gerhards, and V. Schirrmacher, “Intro-duction of adhesive and costimulatory immune functionsinto tumor cells by infection with Newcastle Disease Virus,”International Journal of Oncology, vol. 13, no. 6, pp. 1105–1115, 1998.

[39] C. C. Termeer, V. Schirrmacher, E. B. Brocker, and J.C. Becker, “Newcastle disease virus infection induces B7-1/B7-2-independent T-cell costimulatory activity in humanmelanoma cells,” Cancer Gene Therapy, vol. 7, no. 2, pp. 316–323, 2000.

[40] M. A. Nolte, R. W. Van Olffen, K. P. J. M. Van Gisbergen,and R. A. W. Van Lier, “Timing and tuning of CD27-CD70interactions: the impact of signal strength in setting thebalance between adaptive responses and immunopathology,”Immunological Reviews, vol. 229, no. 1, pp. 216–231, 2009.

[41] T. Yamamoto, M. Hattori, and T. Yoshida, “Induction of T-cell activation or anergy determined by the combination ofintensity and duration of T-cell receptor stimulation, andsequential induction in an individual cell,” Immunology, vol.121, no. 3, pp. 383–391, 2007.

[42] M. Aigner, M. Janke, M. Lulei, P. Beckhove, P. Fournier, andV. Schirrmacher, “An effective tumor vaccine optimized forcostimulation via bispecific and trispecific fusion proteins,”International Journal of Oncology, vol. 32, no. 4, pp. 777–789,2008.

[43] C. Haas, M. Lulei, P. Fournier, A. Arnold, and V. Schirrmacher,“A tumor vaccine containing anti-CD3 and anti-CD28 bispe-cific antibodies triggers strong and durable antitumor activityin human lymphocytes,” International Journal of Cancer, vol.118, no. 3, pp. 658–667, 2006.

[44] M. Feuerer, P. Beckhove, L. Bai et al., “Therapy of humantumors in NOD/SCID mice with patient-derived reactivatedmemory T cells from bone marrow,” Nature Medicine, vol. 7,no. 4, pp. 452–458, 2001.

[45] G. M. Barton and R. Medzhitov, “Control of adaptiveimmune responses by Toll-like receptors,” Current Opinion inImmunology, vol. 14, no. 3, pp. 380–383, 2002.

[46] P. Matzinger, “Tolerance, danger, and the extended family,”Annual Review of Immunology, vol. 12, pp. 991–1045, 1994.

[47] C. Forden, “Do T lymphocytes correlate danger signals toantigen?” Medical Hypotheses, vol. 62, no. 6, pp. 898–906,2004.

[48] T. Kawai and S. Akira, “Pathogen recognition with Toll-likereceptors,” Current Opinion in Immunology, vol. 17, no. 4, pp.338–344, 2005.

[49] O. Schulz, S. S. Diebold, M. Chen et al., “Toll-like receptor 3promotes cross-priming to virus-infected cells,” Nature, vol.433, no. 7028, pp. 887–892, 2005.


[50] R. E. Berg and J. Forman, “The role of CD8+ T cells in innateimmunity and in antigen non-specific protection,” CurrentOpinion in Immunology, vol. 18, no. 3, pp. 338–343, 2006.

[51] B. J. Marsland, C. Nembrini, N. Schmitz et al., “Innate signalscompensate for the absence of PKC-θ during in vivo CD8+T cell effector and memory responses,” Proceedings of theNational Academy of Sciences of the United States of America,vol. 102, no. 40, pp. 14374–14379, 2005.

[52] P. Fournier and V. Schirrmacher, “Randomized clinical studiesof anti-tumor vaccination: state of the art in 2008,” ExpertReview of Vaccines, vol. 8, no. 1, pp. 51–66, 2009.

[53] S. Dhanji, S. J. Teh, D. Oble, J. J. Priatel, and H. S. Teh, “Self-reactive memory-phenotype CD8+ T cells exhibit both MHC-restricted and non-MHC-restricted cytotoxicity: a role for theT-cell receptor and natural killer cell receptors,” Blood, vol.104, no. 7, pp. 2116–2123, 2004.

[54] S. Dhanji and H. S. Teh, “IL-2-activated CD8CD44 cellsexpress both adaptive and innate immune system receptorsand demonstrate specificity for syngeneic tumor cells,” Journalof Immunology, vol. 171, no. 7, pp. 3442–3450, 2003.

[55] J. A. Gollob, C. P. Schnipper, E. Orsini et al., “Characterizationof a novel subset of CD8+ T cells that expands in patientsreceiving interleukin-12,” Journal of Clinical Investigation, vol.102, no. 3, pp. 561–575, 1998.

[56] S. Slavin, A. Ackerstein, R. Or et al., “Immunotherapy in high-risk chemotherapy-resistant patients with metastatic solidtumors and hematological malignancies using intentionallymismatched donor lymphocytes activated with rIL-2: a phaseI study,” Cancer Immunology, Immunotherapy, vol. 59, no. 10,pp. 1511–1519, 2010.

[57] V. Schirrmacher and P. Fournier, “Danger signals in tumorcells: a risk factor for autoimmune disease?” Expert Review ofVaccines, vol. 9, no. 4, pp. 347–350, 2010.


Review Article

Monoclonal Antibodies for Non-Hodgkin’s Lymphoma:State of the Art and Perspectives

Giulia Motta, Michele Cea, Eva Moran, Federico Carbone, Valeria Augusti,Franco Patrone, and Alessio Nencioni

Department of Internal Medicine, University of Genoa, Room 221, V.le Benedetto XV 6, 16132 Genoa, Italy

Correspondence should be addressed to Alessio Nencioni, [email protected]

Received 1 July 2010; Revised 5 November 2010; Accepted 22 December 2010

Academic Editor: Scott Antonia

Copyright © 2010 Giulia Motta et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Monoclonal antibodies have been the most successful therapeutics ever brought to cancer treatment by immune technologies. Theuse of monoclonal antibodies in B-cell Non-Hodgkin’s lymphomas (NHL) represents the greatest example of these advances, asthe introduction of the anti-CD20 antibody rituximab has had a dramatic impact on how we treat this group of diseases today.Despite this success, several questions about how to optimize the use of monoclonal antibodies in NHL remain open. The bestadministration schedules, as well as the optimal duration of rituximab treatment, have yet to be determined. A deeper knowledgeof the mechanisms underlying resistance to rituximab is also necessary in order to improve the activity of this and of similartherapeutics. Finally, new antibodies and biological agents are entering the scene and their advantages over rituximab will have tobe assessed. We will discuss these issues and present an overview of the most significant clinical studies with monoclonal antibodiesfor NHL treatment carried out to date.

1. Introduction

In 1975, Kohler and Milstein heralded a new era in antibodyresearch with their discovery of hybridoma technology[1]. Mouse hybridomas were the first reliable source ofmonoclonal antibodies. Subsequently, the introduction ofrecombinant technologies, transgenic animals, and phagedisplay technology has modernized selection, humanizationand production of therapeutic antibodies. The use of mAbsin cancer treatment stems from the idea that these, becauseof their intrinsic specificity, could be used to selectively targetcancer cells based on the expression of one or more antigens.In such approaches, antibodies could be used alone or beconjugated to toxins, radioactive moieties, or enzymes inorder to achieve toxic concentrations of these agents in thecancerous tissues while sparing healthy organs.

Indeed, since their initial discovery, more than 20 mAbshave been approved by the US Food and Drug Admin-istration (FDA) for the treatment of several conditions,including several types of cancers. This success has openednew therapeutic perspectives and prompted research effortsaimed to improve their activity, select for those patients who

will most benefit from them, and, potentially, to expand theirtherapeutic indications. The anti-CD20 mAb rituximab isone of the best examples of this new class of therapeutics,since it has rapidly become a key part of the pharmacologicalschemes used to treat Non-Hodgkin’s lymphomas (NHLs).Moreover, due to its capacity to eliminate B lymphocytes, ithas recently been applied in immune-mediated disorders [2].

Here, we will focus on the use of rituximab in thetreatment of NHL, on the clinical issues associated with thistherapeutic, and on the most recent advances in the field oflymphoma immunotherapy.

2. Tumor Antigens in NHL

When designing a therapeutic approach for NHL, cancerimmunologists face the issue of selecting the best targetantigen. Tumor antigens are traditionally divided in tumor-specific antigens (proteins that are uniquely expressed bycancer cells) and tumor-associated antigens (molecules thatare expressed by cancer cells, although their expression isalso found on normal cells) [3]. Ideally, an immune response


B cell

CD20

Rituximab

CD20

CD20

CD20

P38 MAPK

ERK 1/2NF-κBAkt

Macrophage,monocyte or

natural killer cell

Cell lysis

Membraneattack

complex

Cell lysis

Antibody-dependent cell-mediatedcytotoxicity (ADCC)

Complement-dependentcytotoxicity (CDC)

Apoptosis

Figure 1: Schematic representation of the putative mechanisms mediating rituximab’s anticancer activity in NHL cells. The anti-CD20monoclonal antibody rituximab has several mechanisms of action, including antibody-dependent cellular cytotoxicity (ADCC), whichinvolves recruitment of effector cells, mediated by Fcγ receptors; complement-dependent cytotoxicity (CDC); apoptosis induction.

against tumor antigens should destroy tumor cells withoutdamaging normal cells. Thus, cancer-specific antigens wouldbe the first choice. Unfortunately, true cancer-specific anti-gens, such as new proteins resulting from fusion oncogenes,are not frequent in NHL. Another important issue is toensure that the chosen antigen does not mutate in a waythat allows cancer cells to avoid destruction by the immunesystem [3].

The cell surface protein CD20 is a 33-kDa proteinexpressed by mature B cells and most malignant B cells, butnot by pre-B cells or differentiated plasma cells [4–8]. Invitro studies have revealed that CD20 acts as a calcium ionchannel [9, 10], and may also activate intracellular signalingthrough its ability to associate with the B-cell receptor (BCR)[11]. Interestingly, CD20’s ability to induce cytosolic Ca2+

flux appears to be BCR dependent. Rituximab (Rituxan,Mabthera), is the first anti-CD20 monoclonal antibodyapproved by the Food and Drug Administration (FDA)(on November 26, 1994) for the treatment of relapsed orrefractory, CD20+ follicular lymphoma (FL). It is a chimericanti-CD20 antibody derived from the mouse mAb 2B8,targeting CD20 antigens, following replacement of the heavyand light chain constant regions with the correspondingregions of a human IgG1 mAb. Importantly, rituximabdepletes both malignant and normal CD20+ B lymphocytes[4, 12, 13].

3. Rituximab’s Mode of Action inLymphoma Cells

Although the exact in vivo mechanisms of action forrituximab are not fully understood, the mechanisms of B-cellkilling by this mAb have been exhaustively analyzed [14].

Briefly, the major mechanism of rituximab-induced B-cell depletion involves antibody-dependent cell-mediatedcytotoxicity (ADCC) and complement dependent cytotox-icity (CDC) [15]. Additionally rituximab was reported todirectly induce apoptosis, inhibit B-cell proliferation, and toenhance the cytotoxic activity of chemotherapeutic agents[16] (Figure 1).

Rituximab-induced CDC is triggered upon rituximabbinding to B cells with consequent initiation of the comple-ment cascade starting from C1 activation. This mechanismcauses osmotic lysis of neoplastic B cells [13, 14]. ADCC istriggered by the interaction between rituximab and the Fcreceptor of natural killer (NK) cells [13, 14]. Once activated,NK cells release small proteins, including perforin andgranzymes, which in turn form pores in the malignant B-cellmembrane, and thus induce apoptosis or osmotic cell lysis.Finally, recent data demonstrate the novel role of rituximabas a signal-inducing antibody, and as a chemosensitizingagent, capable of negative regulation of major survivalpathways [16].


Besides these mechanisms, rituximab’s activity appearsto be linked, at least in part, to its signaling via CD20.In this field, studies in B-NHL cell lines revealed sev-eral mechanisms involved in rituximab-mediated chemo/immunosensitization. Rituximab was shown to inhibit thep38 mitogen-activated protein kinase, nuclear factor-κB(NF-κB), extracellular signal-regulated kinase 1/2 (ERK 1/2),and Akt antiapoptotic survival pathways [17]. All of theseeffects result in upregulation of PTEN and of Raf kinaseinhibitor protein (RKIP) [18], in the downregulation ofantiapoptotic gene products, such as Bcl-2, Bcl-xL and Mcl-1, and, as a result, in chemo/immunosensitization [19].In addition, treatment with rituximab inhibits the overex-pressed transcription repressor Yin Yang 1 (YY1) [20]. YY1downregulates Fas and DR5 expression and its inhibitionleads to sensitization to Fas ligand and tumor necrosis factor-related apoptosis-inducing ligand- (TRAIL-) induced celldeath [21].

Interestingly, recent studies also show that rituximabstrongly affects BCR signaling [22]. Pretreatment of lym-phoma cells or healthy B-cells with rituximab results ina time-dependent inhibition of the BCR-signaling cascadeinvolving Lyn, Syk, PLCγ2, Akt, ERK, and calcium flux. Suchinhibitory effects by rituximab are associated with a decreasein raft-associated cholesterol, inhibition of BCR relocaliza-tion to lipid rafts, and BCR downregulation. Since BCRsignaling appears to be crucial for healthy and malignant Bcell survival and expansion [23–25], this mode of action ofrituximab could actually have an important role in mediatingits anticancer activity.

The relative importance of each mechanism of action ofrituximab is likely to vary with the type of tumor and thetype of treatments that are administered together with thismAb. CDC and ADCC appear to be important to targetleukemia/lymphoma cells circulating in the bloodstream[26]. Conversely, an immunological mechanism of actionseems to be less important in the presence of nodal andextranodal involvement.

4. Rituximab’s Applications inHematological Malignancies

We will discuss here the current therapeutic applications ofrituximab in indolent NHL, diffuse large B cell non-Hodgkinlymphoma (DLBCL), and in B-cell chronic lymphocyticleukemia (B-CLL). Although trials may have had endpointdefinitions that are not always identical, almost all definedcomplete response (CR) as the complete disappearance of thesymptoms and signs of lymphoma (including bone marrowclearing for >28 days), and partial response (PR) as a >50%decrease in the size or number of the lymphomas lesions,without any evidence of progressive disease for >28 days. CRand PR together represent the objective response (OR) rate[4, 27].

4.1. Follicular and Low Grade Lymphoma. Until the early90’s, the first-line therapy in symptomatic low-grade NHLwas chlorambucil and prednisone [42]. Subsequently, several

randomized trials showed the efficacy of rituximab in com-bination with other chemotherapeutic agents such as flu-darabine (R-F), fludarabine, and cyclophosphamide (R-FC),fludarabine, cyclophosphamide and mitoxantrone (FCM-R),cyclophosphamide, vincristine, and prednisone (R-CVP),CVP plus mitoxantrone (R-CNOP), fludarabine, dexam-ethasone, and mitoxantrone (R-FND) as well as CHOP (R-CHOP) [28, 43–45] (Table 1). The clinical response rates ofrituximab-containing regimens were encouraging, with anOR rates consistently around 95% and with a CR and PRrates ranging from 45% to 100%, and from 0% to 52%,respectively.

Importantly, clinical data on the benefit of rituximabcombined with chemotherapy has also become available inpatients with relapsed or refractory indolent B-cell NHL.Also here, the results are very encouraging, with OR of 81%for R-CVP, 97% for R-FC, 88% for R-CHOP, and 95% forFCM-R respectively [4, 46].

Finally, the efficacy of rituximab monotherapy in patientswith relapsed or refractory CD20-positive low-grade orfollicular lymphoma was examined in noncomparative mul-ticentre trials [33–35, 47–55]. The overall response rates were38%–48% after a 4-week therapy with rituximab, and 57%after 8 weeks of rituximab administration. CR rates rangingbetween 3 and 17% were recorded in these studies.

Remarkably, studies show that, in FL, sequential admin-istration of standard chemotherapy followed by rituximabinduces molecular clearance (as detected by PCR for theBcl-2/IgH rearrangement) in more than 70% of the patients[42, 43, 56]. The actual clinical impact of achieving amolecular response in FL still has to be determined,since long-term remissions have been reported also inpatients with persistently detectable Bcl-2/IgH rearrange-ment [57]. Moreover this rearrangement may occasion-ally be found in healthy peripheral blood lymphocytes[58]. In fact, a recent study by van Oers and coworkerssuggests that BCL-2/IgH polymerase chain reaction statusat the end of induction treatment would not be predic-tive for progression-free survival in relapsed/resistant FL[59]. Nonetheless, the above-mentioned studies supportthe efficacy of rituximab in FL, and indicate its poten-tial for treating minimal residual disease in this type ofdisorder.

In summary, the current guidelines for the treatmentof FL recommend that rituximab is administered incombination with standard chemotherapy in previouslyuntreated stage III–IV FL, and at first relapse (at a dosage of375 mg/m2 on day 1 of each chemotherapy cycle, for up toeight doses). Rituximab is recommended as a monotherapyfor stage III–IV chemoresistant FL, or at second (orsubsequent) relapse after chemotherapy (375 mg/m2 onceweekly for four doses) (http://www.ema.europa.eu/docs/enGB/document library/Summary of opinion/human/000165/WC500097025.pdf).

4.2. DLBCL. After the disappointing results obtained withthird-generation chemotherapy regimens in the UnitedStates, the CHOP regimen was reverted to as the standard of


Table 1: Principal clinical trials of chemotherapy plus Rituximab versus chemotherapy alone in NHL.

Lymphoma Subtype TreatmentPatients(no.)

% Overallresponse rate(P value)∗

Median Follow-up(mo.)

Reference

FollicularCVP versus

R-CVP321

57 versus81 (<.001)

53 Marcus et al. [28]

FollicularCHOP versus

R-CHOP428

90 versus96 (=.011)

18 Hiddemann et al. [29]

FollicularCHOP versus

R-CHOP465

72.3 versus85.1 (<.001)

39,4 van Oers et al. [30]

FollicularFCM versus

R-FCM176

71 versus95 (=.01)

26 Forstpointner et al. [31]

FollicularMCP versus

R-MCP201

75 versus92 (=.009)

47 Herold et al. [32]

relapsed/refractarylow grade

R 37 46 13,4 Maloney et al. [33]


R 30 47 19 Feuring-Buske et al. [34]


R 166 48 19,5 McLaughlin et al. [35]

DLBCLCHOP versus

R-CHOP399

63 versus76 (=.005)∗

24 Coiffier et al. [36]

DLBCLCHOP versus

R-CHOP824

84 versus93 (=.0001)∗∗

34 Pfreundschuh et al. [37]

DLBCLCHOP versus

R-CHOP632

57 versus67 (=.05)∗∗

42 Habermann et al. [38]

DLBCLCHOP versus

R-CHOP122

75 versus94 (=.0054)

18 Lenz et al. [39]

B-CLL FC versus R-FC 55258 versus69.9 (=.0034)

25 Robak et al. [40]

B-CLL FC versus R-FC 81782.5 versus87.2 (=.012)∗∗

37,7CLL8- German CLLStudy Group∗∗∗ [41]

∗CR, CR-unconfirmed, partial response.

∗∗CR rate.∗∗∗3-year OS.

therapy. The efforts to introduce rituximab in the treatmentof this aggressive hematological disease led to two essentialclinical trials: the Mabthera International trial (MinT) [37]and the Groupe d’Etude des lymphomes de l’Adulte study(GELA) [36]. The first one involved young, the latter elderly,DLBCL patients. In the multicenter study conducted byCoiffier and colleagues, therapy using rituximab combinedwith standard CHOP chemotherapy demonstrated a higherefficacy than CHOP alone, in terms of both event-freesurvival at 2 years (57% versus 38%, P < .001), overallsurvival at 2 years (70% versus 57%, P < .01), and CRrate (76% versus 63%, P < .01). Likewise, the MinTstudy showed an increased OS of the combined rituximab-adding regimen, compared to standard therapy, from 84%to 93%. These results led to FDA approval of rituximabin combination with CHOP chemotherapy for previouslyuntreated patients with DLBCL. Whether or not all patientsneed rituximab has been questioned. Studies from Franceand the American National Cancer Institute suggested thatthe benefit of rituximab would be observed in patients withtumors overexpressing Bcl-2. On the other hand, a recent

report from the French group shows benefit in both Bcl-2-positive and Bcl-2-negative lymphomas using the methodof competing risks [27, 60, 61]. Therefore, the question ofwhether molecular features should or will direct treatmentdecisions remains unanswered.

Finally, for recurrent DLBCL, the standard of care issalvage chemotherapy followed by high-dose chemotherapywith stem cell transplantation. Also in this setting, rituximabproved to be effective and has been incorporated into salvagechemotherapy regimens, since it may improve the overallresponse rate with ICE (ifosfamide, carboplatin, and etopo-side) and DHAP (dexamethasone, high-dose cytarabine, andcisplatin) [62].

In summary rituximab is approved for previouslyuntreated DLBCL patients in combination with CHOPchemotherapy and with salvage chemotherapy regimens inrelapsed/refractory patients. The recommended rituximabdosage is 375 mg/m2 on day 1 of each chemotherapy cycle,for up to eight doses (http://www.ema.europa.eu/docs/enGB/document library/Summary of opinion/human/000165/WC500097025.pdf).


4.3. Rituximab and Autologous Stem Cells Transplantationfor Advanced Stage DLBCL. Young high-risk patients withDLBCL achieving a complete remission after a completecourse of chemotherapy are likely to benefit from autologousstem cell transplantation (ASCT) [27, 63]. Several studies areassessing the role of rituximab as part of high-dose regimens(HDT) pre-ASCT in DLBCL because of its effectiveness,limited toxicity, and its ability to deplete B cells. In this field,in a 2-year study by Khouri and colleagues evaluating theefficacy and safety of high-dose rituximab in combinationwith high-dose BEAM and ASCT, the OS was 80% for thestudy group compared to 53% for the control group [64].Superior survival rates have also been reported for patientswho become PCR negative for BCL2/JH rearrangements inperipheral blood or bone marrow compared with those whoremain positive.

4.4. Rituximab Maintenance Therapy for FL and DLBCL.Despite the fact that rituximab used in combination withchemotherapy has been shown to prolong the survival ofpatients with NHL, residual lymphoma cells (which thenbecome responsible for disease relapses) frequently remain[65]. As a matter of fact, NHL relapses continue to bean important clinical issue. Therefore, several randomizedtrials have been conducted in order to analyze the benefitof rituximab maintenance treatment in NHL [66–68]. Thestudies that were done for FL adopted different schemes forinduction (rituximab 375 mg/m2 weekly × 4 in Ghielmini etal. and in Hainsworth et al.; CHOP or R-CHOP in van Oerset al.; fludarabine, cyclophosphamide, and mitoxantronewith or without rituximab in Forstpointner et al.) as wellas for the maintenance treatment (375 mg/m2 intravenouslyweekly for 4 weeks at six-month intervals in Hainsworthet al.; 375 mg/m2 intravenously weekly for 4 weeks forGhielmini et al.; 375 mg/m2 rituximab intravenously onceevery 3 months in van Oers et al.; 2 further courses of4-times-weekly doses of rituximab after 3 and 9 monthsin Forstpointner et al.). However, overall, they unequivo-cally show that rituximab maintenance increases event-freesurvival (EFS) and duration of response in indolent NHL.In Ghielmini et al., at a median followup of 35 months,the median EFS was 12 months in the no-maintenancegroup versus 23 months in the prolonged treatment arm(P = .02) [69]. The authors reported that the differencewas particularly notable in chemotherapy-naive patients (19versus 36 months; P = .009) and in patients responding toinduction treatment (16 versus 36 months; P = .004). In thestudy by van Oers et al., rituximab maintenance significantlyimproved EFS compared with observation (median, 3.7 yearsversus 1.3 years; P < .00), both after CHOP induction(P < .001) and R-CHOP (P = .003) [30]. The 5-yearoverall survival (OS) was 74% in the rituximab maintenancearm, and it was 64% in the observation arm (P = .07).Finally, also in the trial by Forstpointner and colleagues,response duration was significantly prolonged by rituximabmaintenance, with the median not being reached in thisevaluation versus an estimated median of 16 months in theobservation group (P = .001) [31]. This beneficial effect was

also observed when analyzing FL (P = .035) and mantle celllymphoma (P = .049) separately.

Unlike in indolent NHL, rituximab maintenance therapyin DLBCL has failed to demonstrate benefit in the publishedclinical trials [38].

In conclusion, the current guidelines recommend theuse of rituximab as a maintenance therapy only in relapsedor refractory follicular lymphoma responding to inductiontherapy with chemotherapy with or without rituximab.The recommended dosage of rituximab is 375 mg/m2

once every 3 months until disease progression or for amaximum of 2 years (http://www.ema.europa.eu/docs/enGB/document library/Summary of opinion/human/000165/WC500097025.pdf).

4.5. B-cell Chronic Lymphocytic Leukemia (B-CLL). B-CLLis a heterogeneous disorder with a variable course (i.e.,following diagnosis, survival ranges from months to decades)and risk factors such as age and performance status shouldbe considered when selecting the most appropriate treatmentoption [70].

Rituximab monotherapy is generally not associated withsustained responses in B-CLL, possibly reflecting alteredrituximab pharmacokinetics in patients with B-CLL [40, 70–73]. However, studies show that the addition of rituximabto fludarabine plus cyclophosphamide (FC) does improveclinical outcomes in B-CLL patients. The first study, knownas CLL8, was conducted by the German CLL Study Groupon 817 previously untreated B-CLL patients (ClinicalTri-als.gov number, NCT00281918). The second trial, known asREACH, enrolled 552 patients with relapsed or refractoryB-CLL following prior systemic therapy [40]. Both studiesshowed a benefit in terms of OS rates in the R-FC armversus FC arm (86% versus 73 % in the CLL8 trial and 54%versus 45% in the REACH). In addition, the benefit of addingrituximab to chemotherapy in B-CLL was shown by severalother trials [40, 74–77].

Interestingly, since rituximab plus FC represents thestandard treatment for B-CLL, clinical studies compared theconventional regimen to rituximab plus low-dose FC (i.e.,FCR-Lite) or to sequential FC and rituximab [49], sincethese alternative regimens are expected to be associated withless grade 3 or 4 neutropenia than the conventional R-FCregimen [5, 50].

The current international guidelines recommend thatchemoimmunotherapy regimens with R-FC are preferredas the first-line treatment for advanced CLL (stage II–IV)in patients without del(17p) who are aged <70 years oraged >70 years without significant comorbidities. Amongpatients with relapsed or refractory disease, those with a longresponse (i.e., >3 years) can be retreated with one of thefirst-line treatment options. Various chemoimmunotherapyoptions are suggested for patients with a short response(i.e., <2 years) (e.g., rituximab may be administered incombination with FC or with CHOP). The recommendeddosage of rituximab is 375 mg/m2 the day before startingchemotherapy, followed by 500 mg/m2 on day 1 of cycles2–6 (National Comprehensive Cancer Network. NCCN


clinical practice guidelines in oncology: non-Hodgkin’s lym-phoma).

5. Tolerability

Adverse events were reported in 84% of patients, receivingrituximab, during therapy or within the first 30 daysfollowing treatment [4, 35]. However, more than 95% ofthese events were described as mild to moderate in severity,of brief duration, and observed during the first infusion.The most common adverse effects were infusion-relatedreactions and lymphopenia. Ten percent of the patientsreported severe fever, chills, infections, or other adverseeffects. Serious adverse effects included severe infusion-related reactions, tumor lysis syndrome, mucocutaneousreactions, hypersensitivity reactions, cardiac arrhythmias,angina, and renal failure [4, 51].

These adverse events were less common during thesubsequent rituximab administrations. One possible hema-tological adverse event is the reduction in peripheral B-lymphocyte counts, which can last for up to 6 months witha recovery period of 9 to 12 months [4, 35]. Nevertheless,the risk of serious opportunistic infections appears to bemuch lower than that reported with conventional therapy[4]. Interestingly, Bedognetti and coworkers have recentlyevaluated the impact of rituximab on the effectiveness of anantiflu vaccine in patients who had previously been treatedwith this mAb [52]. Due to the fact that disease status mightaffect immune response, only NHL patients without evi-dence of disease, who had completed rituximab no less than6 months before the accrual, were selected for this evaluation.The study showed that patients who had previously receivedrituximab had a significantly lower seroconversion rate inresponse to the vaccine. Remarkably, while peripheral CD27-naıve B cells were present, Bedognetti et al. found a profounddepletion in CD27+B memory cells, which may well explainthe defective induction of antiflu immunity. Thus, concernsremain that patients who have been treated with the anti-CD20 mAb may be at risk for infections and that they mayneed careful monitoring.

6. Improving Rituximab Efficacy andOvercoming Resistance

Despite the expression of CD20 on their lymphoma cells,some patients exhibit primary resistance and do not respondwell to this targeted antibody therapy. Moreover, an initiallyresponsive lymphoma can subsequently become resistant torituximab (secondary/acquired resistance). Several mecha-nisms have been reported that have the potential to con-tribute to reductions in rituximab efficacy. The identificationof such mechanisms has allowed for the proposal of strategiesto overcome these issues, and thus achieve better in vivoactivity. Some of these mechanisms have been reviewedelsewhere [53]. Here, we will summarize some of themost recent and promising observations, and the relatedsuggestions for therapeutic interventions.

(i) Interfering with CD20 Downregulation/Shaving. Ini-tial in vitro observations suggested that CD20 wouldnot be downregulated in the presence of anti-CD20antibodies. Namely, the anti-CD20/CD20 complexwas found to remain at the cell surface long enough toensure cell killing by specific mechanisms. However,these observations may not be reproduced in in vivosettings. A recent report by Beers et al. showed thatrituximab is able to induce CD20 internalizationin a B-CLL mouse model [54]. Interestingly, theseauthors demonstrated that the degree of CD20 down-modulation correlates inversely with some types ofNHL’s susceptibility to rituximab. Namely, CLL andmantle cell lymphoma showed greater downmod-ulation of CD20 in response to rituximab thanFL and DLBCL did, and were less responsive totreatment. Previous reports by Beum et al. describeda “shaving reaction” in which mAb-CD20 complexeswere “shaved” off CLL cells, by phagocytes, as themalignant cells circulated [78]. Whether the observedreduction in CD20 levels actually reflects shaving,or rather antigen masking by rituximab, remainsunclear [79, 80]. Downregulation of CD20 access,irrespective of the underling cause, appears to be animportant mechanism affecting rituximab efficacy, asantigen loss by malignant cells will prevent rituximabactivity. New anti-CD20 mAbs (tositumomab-like)may be able to induce considerably less CD20down-modulation than rituximab, and thus possiblybe more effective (see below) [54]. It is also ofinterest that CD20 expression on lymphoma cellscan be increased with HDAC inhibitors, such asvalproic acid and romidepsin [81]. These were shownto transactivate the CD20 gene through promoterhyperacetylation and Sp1 recruitment. In line withthese premises, HDAC inhibitors potentiated theactivity of rituximab both in vitro and in vivo inmurine lymphoma models.

(ii) Targeting CD20 Transcript Variants Associated withResistance. Henry and coworkers have recently identi-fied an alternative CD20 transcript variant (ΔCD20)associated with resistance to rituximab [82]. Thisnovel, alternatively spliced CD20 variant encodes fora truncated 130 amino acid protein lacking largeparts of the four transmembrane domains, suggestingthat ΔCD20 is a nonanchored membrane protein.ΔCD20 expression was detected in B-cell leukemias,B-cell lymphomas, and activated B cells, but not inhealthy resting B cells. Finally, the authors went onto show that ΔCD20 is associated with resistanceto rituximab, although the mechanism whereby thisCD20 splice variant impairs the benefit of rituximabremains to be determined. The authors suggest that,given its selective expression in malignant (andactivated) B-cells, ΔCD20 could become a thera-peutic target, for instance for the development ofantilymphoma vaccines. Whether this approach willprove effective remains to be assessed.


(iii) Preventing NK Cell-Mediated ADCC Exhaustion.NK cell-mediated ADCC can be exhausted. Studiesshowed that NK cells can engage and kill 3-4 targetcells in 16 hours. Thereafter, cells become exhausted,possibly due to a reduction in the available levels ofperforin and granzyme B [83]. Indeed, incubationof NK cells with rituximab-coated target cells leadsto CD16 (FcγRIIIa) downregulation and to upreg-ulation of CD107a, a marker for degranulation andexhaustion [84, 85]. Finally, NK cell-mediated targetcell killing was shown to become less efficient inthe presence of high burdens of rituximab-opsonizedlymphoma cells [86]. Remarkably, IL-2 treatment canrestore NK cell-mediated ADCC. In line with thisconcept, Berdeja and coworkers found that systemicinterleukin-2 and adoptive transfer of lymphokine-activated killer cells improves antibody-dependentcellular cytotoxicity in patients with relapsed B-cell lymphoma treated with rituximab [86]. In thiscontext, recent studies showed that also complementcomponents, such as C3b, can inhibit NK-cell medi-ated killing of mAb-opsonized lymphoma cells [55].Importantly, C3 depletion by cobra venum factor, orthe related drug (HC3-1496), appears to effectivelyovercome this mechanism and improve the activity ofrituximab in lymphoma-bearing mice. Thus, overall,strategies aimed to improve NK cell activity couldhelp enhance the efficacy of rituximab and shouldtherefore be further investigated.

(iv) Enhancing CDC. Studies showed that also comple-ment can be depleted upon rituximab infusion in B-CLL patients [79]. Kennedy et al. found that freshfrozen plasma would then restore rituximab efficacy.More studies on this approach should be performedin order to confirm its viability. Another approach toenhance rituximab-induced CDC has recently beenproposed by Wang and colleagues [87]. These authorsobserved that many tumors, including lymphomas,upregulate the expression of CD46, an inhibitorycomplement receptor. As a means to overcomethis issue, they identify a recombinant adenovirustype 35 fiber knop protein (Ad35K++) which, whenincubated with lymphoma cells, leads to CD46downregulation and cooperates with rituximab ininducing CDC. In xenograft models with humanlymphoma cells, preinjection of Ad35K++ dramat-ically increased the efficacy of rituximab, suggest-ing that the Ad35K++-based approach has potentialimplications in mAb therapy of NHL. Finally, Satoand colleagues have recently reported the identifica-tion of a novel CDC-enhancing variant of rituximab(113F) [88]. Compared to rituximab, 113F appearedto mediate highly enhanced CDC against primaryCD20-expressing lymphoma cells in vitro. Moreover,these authors were able to establishe a human tumor-bearing NOD/Shi-scid-IL-2Rγ(null) mouse model,in which human complement functions as the CDC

mediator. Using this model, the authors demon-strated that 113F exerted significantly more potentantitumor effects than rituximab.

(v) Improving Phagocytosis Through CD47 Blockade.Chao and colleagues have recently shown thatmultiple B-cell NHL subtypes, including DLBCL,FL, and B-CLL, exhibit increased levels of CD47,a transmembrane protein which activates SIRP1ain phagocytic cells [89]. This results in initiationof a signal transduction cascade which leads tophagocytosis inhibition. These authors demonstratethat CD47 overexpression correlates with worseprognosis. Blocking anti-CD47 antibodies promotephagocytosis of NHL cells and cooperate with rit-uximab both in vitro and in vivo in murine NHLxenotransplant models. Again, whether this approachwill prove useful in humans remains to be assessed.

(vi) Topical IFN-α Delivery. Finally, Xuan and colleagueshave proposed an approach to target IFN-αmoleculesto lymphoma sites by constructing a fusion proteinconsisting of IFN-α and an anti-CD20 mAb [90].IFN-α has potent immunostimulatory properties andantiproliferative effects in some B-cell NHLs, but itssystemic administration is frequently associated tosignificant toxicity. The CD20-IFN-α fusion proteinsshowed efficient anticancer activity against an aggres-sive rituximab- resistant human CD20+ murinelymphoma (38C13-huCD20) and a human B-celllymphoma (Daudi). Further experimentation withthis administration method is warranted to assess itsapplicability in patients.

(vii) Rituximab Mutants with Proapoptotic Activity. Inorder to improve rituximab anticancer activity, Liand colleagues modulated the binding property ofthis mAb by introducing several point mutationsin its complementarity-determining regions [91].These authors found that the CDC potency ofsuch CD20 mAbs was independent of the off-rate.However, they were able to identify a rituximabtriple mutant (H57DE/H102YK/L93NR) with anextremely potent apoptosis-inducing activity. Thistriple mutant efficiently initiated both caspase-dependent and-independent apoptosis, and exhib-ited potent in vivo activity even in a rituximab-resistant lymphoma model. These modified versionsof rituximab hold promise as new therapeutic agentsfor B-cell lymphomas, although their efficacy inpatients still has to be assessed.

(viii) Combining Rituximab with Other mAbs. Rituximabactivity in NHL as a single agent is limited, espe-cially when administered to pretreated patients.However, combining rituximab with chemotherapydoes achieve significantly better outcomes thanchemotherapy alone. In addition, strategies to usetwo mAbs have also been proposed. Combinationssuch as anti-CD20 plus anti-CD22, anti-CD20 plus


anti-HLA-DR, anti-CD20 plus anti-TRAIL-R1, anti-CD20 plus anti-CD80 have been evaluated preclin-ically and/or clinically, showing enhanced antitu-mor activity both in vitro and in vivo [92–94].An interesting approach to achieve the benefit ofmultiple targeting in NHL consists of the genera-tion of multivalent antibodies using the so-namedDock-and-Lock (DNL) method, which enables site-specific self-assembly of two modular componentswith each other, resulting in a covalent structure withretained bioactivity [95]. Using this approach, Rossiand colleagues generated bispecific anti-CD20/CD22hexavalent antibodies with promising antilymphomaactivity in vitro and in vivo [96, 97]. Interest-ingly, in a recent study, these authors were ableto correlate the strong direct cytotoxicity of theanti-CD20/CD22 hexavalent antibodies, comparedto their bivalent parental antibodies, with theirincreased ability to upregulate PTEN, phospho-p38,and cyclin-dependent kinase inhibitors, such as p21,p27 and Kip2 [98].

7. Other mAbs for NHL

In addition to the above-mentioned strategies aiming toimprove rituximab activity, numerous research efforts haveled to new mAbs directed against different target antigensand to the development of radioimmunoconjugates. Themost promising newer therapeutics are listed below.

(i) Epratuzumab: a humanized IgG1 anti-CD22 anti-body. It induces ADCC and CDC in preclinicalstudies. Phase I/II studies demonstrated objectiveresponses in relapsed/refractory FL (24%) [99], andin DLBCL (15%) [100], without dose-limiting toxiceffects.

(ii) Galiximab: a primatised anti-CD80 (IgG1λ) mAbwith human constant regions and primate (cynomo-logus macaque) variable regions [101]. CD80 is acostimulatory molecule involved in regulating T-cellactivation. It is transiently expressed on the surface ofactivated B cells, dendritic cells, and T cells of healthyindividuals [102]. Additionally, a variety of lymphoidmalignancies constitutively express CD80, makingthis antigen a suitable target [103]. A phase-I/IIstudy showed that GALIXIMAB is able to enhancerituximab antitumor activity in previously untreatedNLH patients, with a response reported in 70% ofpatients [104].

(iii) Alemtuzumab (Campath): a humanized monoclonalantibody against CD52 (an antigen expressed bynormal and malignant B- and T-lymphocytes, mono-cytes, and NK cells). It is indicated for the treatmentof patients with B-CLL refractory to fludarabine(ORR of 56%) [105], for advanced-stage mycosisfungoides/Sezary syndrome [106], and for relapsedor refractory peripheral T-cell lymphomas [107,108]. Notably, although clinically effective, this mAb

induces a dramatic decrease in CD4+ and CD8+ Tlymphocytes and thus strongly increases the risk ofinfections.

(iv) Apolizumab (Hu1D10): a humanized anti-HLA-DRantibody that induces CDC, ADCC, and apoptosis.HLA class II antigens are expressed at the surfaceof professional antigen presenting cells, including Bcells. They are involved in antigen presentation andin promoting cell proliferation. Thus, mAbs againstHLA-DR inhibit B-cell proliferation and induceapoptosis through activation of the extrinsic apop-totic pathway. Recently, this type of approach hasshown promising results in B-cell malignances [109].Single agent therapy APOLIZUMAB in previouslyuntreated B-CLL patients showed an ORR of 83%[110]. Moreover, the combination of APOLIZUMABand rituximab in relapsed/refractory B-cell lym-phoma and B-CLL showed an ORR of 42% [111].

(v) Radioimmunotherapy: this type of treatment involvesthe administration of an antibody linked to aradioisotope. This approach permits the targetingof the radioactive isotopes to cancer tissues and isespecially interesting as it allows for killing neigh-boring cancer cells that either are inaccessible tothe antibody or express insufficient antigen for theantibody to bind in adequate quantities. Two anti-CD20 radioimmunoconjugates are approved for usein patients with relapsed or refractory follicular orlow-grade lymphoma:

(1) Yttrium-90: labelled ibritumomab tiuxetan(zevalin),

(2) iodine-131: labelled tositumomab (bexxar).

These therapeutics hold great promise for the treatmentof NHL and their usefulness has recently been confirmed byseveral clinical trials [112–122].

About 80% of patients with follicular or low-gradelymphomas respond to treatment with Zevalin, with 20 to30% achieving a CR. Interestingly, the duration of responseappears to exceed 3 years in about 25% of patients [123]. Thebenefit of adding a radioisotope to the antibody was con-firmed in a study enrolling patients with indolent NHL thatwere refractory to rituximab. In this study, Zevalin showeda 74% response rate and 15% of CR [115]. Additionally, ascompared to rituximab, Zevalin produces higher responserates among patients with follicular or low-grade lymphomawho have not previously received antibody-based treatments(ORR 80% versus 56%, P = .002; CR 30% versus 16%,P = .04 [115]). Finally Zevalin also appears to be effectiveagainst some diffuse large B-cell lymphomas, and mantle-cell lymphomas, when used in sequence with chemotherapy(ORR of 53% versus 19%; OS 22.4, versus 4.6, resp.) [117].

Similar results are obtained with Bexxar. In particularin patients with NHL refractory to standard chemotherapy,treatment with Bexxar resulted in CR in 20% of patients[124]. Additionally, in one study, 95% of patients withNHL had responses to 131I-labeled tositumomab used as


initial treatment, with 75% demonstrating CR [125]. Finally,the results of a recently completed study (ClinicalTrials.govnumber, NCT00006721), comparing CHOP followed by131I-labeled tositumomab to rituximab plus CHOP for theinitial treatment of FL, is predicted to redefine standardtherapy for this disorder.

Importantly, there are other radiolabeled immunother-apeutics for NHL that are currently under evaluation [112,114, 126–130]. These include

(a) LL2 anti-CD22, conjugated to either 131I or 90Y; Lym-I;

(b) anti-HLA-DR, conjugated to 90Y or 67Cu;

(c) rituximab, conjugated to 211At, 186Re, or 227Th;

(d) anti-CD19 mAb conjugated to 90Y.

8. Conclusions and Perspectives

Combining rituximab with chemotherapy has proven to bean effective treatment for both indolent and aggressive formsof NHL. The same type of treatment can be used in patientswith B-CLL, although its efficacy in this disorder appearsto be lower. In addition, it has also been demonstrated thatusing rituximab alone as a maintenance therapy improves theprognosis and extends disease-free survival in FL. Although astandard scheme for rituximab maintenance therapy has notbeen established yet, it is currently under investigation andthe ongoing studies will establish the most effective regimen.

For patients in which treatment with rituximab has notgiven the expected results, autologous stem cell transplan-tations have shown promise. It has been demonstrated thatusing a cycle of rituximab in association with stem celltransplantations and after it as maintenance therapy, yieldsbetter results than transplant alone.

Radiolabeled antibodies may be effective in rituximab-resistant and chemotherapy-resistant disease, but their clin-ical use is still limited when compared to that of unla-beled mAbs. Recent data suggest that sequential radioim-munotherapy after chemotherapy may have significant clin-ical value. Additionally, novel monoclonal antibodies areunder development. If these will prove to be more effectivethan rituximab will have to be assessed by randomizedcomparative trials.

Overall, the results obtained with antibody-based thera-peutics in NHL are clearly highly promising. They herald theadvent of therapeutic strategies based on targeted agents thatwill likely be more effective and, at the same time, less toxicthan traditional chemotherapy-based treatments.

Grant Support

Alessio Nencioni is supported by the Associazione Italianaper la Ricerca sul Cancro (AIRC) and by the University ofGenoa.

References

[1] G. Kohler and C. Milstein, “Continuous cultures of fusedcells secreting antibody of predefined specificity,” Nature, vol.256, no. 5517, pp. 495–497, 1975.

[2] N. Cooper and D. M. Arnold, “The effect of rituximabon humoral and cell mediated immunity and infection inthe treatment of autoimmune diseases,” British Journal ofHaematology, vol. 149, no. 1, pp. 3–13, 2010.

[3] O. J. Finn, “Cancer immunology,” New England Journal ofMedicine, vol. 358, no. 25, pp. 2704–2715, 2008.

[4] G. L. Plosker and D. P. Figgitt, “Rituximab: a review of itsuse in non-Hodgkin’s lymphoma and chronic lymphocyticleukaemia,” Drugs, vol. 63, no. 8, pp. 803–843, 2003.

[5] A. Demidem, T. Lam, S. Alas, K. Hariharan, N. Hanna, andB. Bonavida, “Chimeric anti-CD20 (idec-C2B8) monoclonalantibody sensitizes a B cell lymphoma cell line to cell killingby cytotoxic drugs,” Cancer Biotherapy and Radiopharmaceu-ticals, vol. 12, no. 3, pp. 177–186, 1997.

[6] L. M. Nadler, J. Ritz, and R. Hardy, “A unique cell surfaceantigen identifying lymphoid malignancies of B cell origin,”Journal of Clinical Investigation, vol. 67, no. 1, pp. 134–140,1981.

[7] M. Petryk and M. L. Grossbard, “Rituximab therapy of B-cellneoplasms,” Clinical Lymphoma, vol. 1, no. 3, pp. 186–194,2000.

[8] N. M. Almasri, R. E. Duque, J. Iturraspe, E. Everett, andR. C. Braylan, “Reduced expression of CD20 antigen asa characteristic marker for chronic lymphocytic leukemia,”American Journal of Hematology, vol. 40, no. 4, pp. 259–263,1992.

[9] J. P. Deans, H. Li, and M. J. Polyak, “CD20-mediatedapoptosis: signalling through lipid rafts,” Immunology, vol.107, no. 2, pp. 176–182, 2002.

[10] J. K. Bubien, L. J. Zhou, P. D. Bell, R. A. Frizzell, and T.F. Tedder, “Transfection of the CD20 cell surface moleculeinto ectopic cell types generates a Ca conductance foundconstitutively in B lymphocytes,” Journal of Cell Biology, vol.121, no. 5, pp. 1121–1132, 1993.

[11] C. A. Walshe, S. A. Beers, R. R. French et al., “Induction ofcytosolic calcium flux by CD20 is dependent upon B cellantigen receptor signaling,” Journal of Biological Chemistry,vol. 283, no. 25, pp. 16971–16984, 2008.

[12] A. J. Grillo-Lopez, E. Hedrick, M. Rashford, and M.Benyunes, “Rituximab: ongoing and future clinical develop-ment,” Seminars in Oncology, vol. 29, no. 1, pp. 105–112,2002.

[13] S. V. Onrust, H. M. Lamb, and J. A. Balfour, “Rituximab,”Drugs, vol. 58, no. 1, pp. 79–88, 1999.

[14] M. J. Glennie, R. R. French, M. S. Cragg, and R. P.Taylor, “Mechanisms of killing by anti-CD20 monoclonalantibodies,” Molecular Immunology, vol. 44, no. 16, pp. 3823–3837, 2007.

[15] J. L. Teeling, W. J. M. Mackus, L. J. J. M. Wiegman et al., “Thebiological activity of human CD20 monoclonal antibodies islinked to unique epitopes on CD20,” Journal of Immunology,vol. 177, no. 1, pp. 362–371, 2006.

[16] A. R. Jazirehi and B. Bonavida, “Cellular and molecular signaltransduction pathways modulated by rituximab (rituxan,anti-CD20 mAb) in non-Hodgkin’s lymphoma: implicationsin chemosensitization and therapeutic intervention,” Onco-gene, vol. 24, no. 13, pp. 2121–2143, 2005.


[17] B. Bonavida, “Rituximab-induced inhibition of antiapop-totic cell survival pathways: implications in chemo/immuno-resistance, rituximab unresponsiveness, prognostic and noveltherapeutic interventions,” Oncogene, vol. 26, no. 25, pp.3629–3636, 2007.

[18] A. R. Jazirehi, M. I. Vega, D. Chatterjee, L. Goodglick, and B.Bonavida, “Inhibition of the Raf-MEK1/2-ERK1/2 signalingpathway, BCL- down-regulation, and chemosensitization ofnon-Hodgkin’s lymphoma B cells by rituximab,” CancerResearch, vol. 64, no. 19, pp. 7117–7126, 2004.

[19] S. Alas and B. Bonavida, “Rituximab inactivates signaltransducer and activation of transcription 3 (STAT3) activityin B-non-Hodgkin’s lymphoma through inhibition of theinterleukin 10 autocrine/paracrine loop and results in down-regulation of Bcl-2 and sensitization to cytotoxic drugs,”Cancer Research, vol. 61, no. 13, pp. 5137–5144, 2001.

[20] M. I. Vega, A. R. Jazirehi, S. Huerta-Yepez, and B. Bonavida,“Rituximab-induced inhibition of YY1 and Bcl-xL expres-sion in ramos non-Hodgkin’s lymphoma cell line via inhi-bition of NF-kappaB activity: role of YY1 and Bcl-xL inFas resistance and chemoresistance, respectively,” Journal ofImmunology, vol. 175, no. 4, pp. 2174–2183, 2005.

[21] S. Huerta-Yepez, M. Vega, A. Jazirehi et al., “Nitric oxidesensitizes prostate carcinoma cell lines to TRAIL-mediatedapoptosis via inactivation of NF-κB and inhibition of Bcl-expression,” Oncogene, vol. 23, no. 29, pp. 4993–5003, 2004.

[22] S. Kheirallah, P. Caron, E. Gross et al., “Rituximab inhibitsB-cell receptor signaling,” Blood, vol. 115, no. 5, pp. 985–994,2010.

[23] L. Chen, S. Monti, P. Juszczynski et al., “SYK-dependenttonic B-cell receptor signaling is a rational treatment targetin diffuse large B-cell lymphoma,” Blood, vol. 111, no. 4, pp.2230–2237, 2008.

[24] L. Leseux, S. M. Hamdi, T. Al Saati et al., “Syk-dependentmTOR activation in follicular lymphoma cells,” Blood, vol.108, no. 13, pp. 4156–4162, 2006.

[25] R. M. Young, I. R. Hardy, R. L. Clarke et al., “Mouse modelsof non-hodgkin lymphoma reveal Syk as an importanttherapeutic target,” Blood, vol. 113, no. 11, pp. 2508–2516,2009.

[26] M. R. Smith, “Rituximab (monoclonal anti-CD20 antibody):mechanisms of action and resistance,” Oncogene, vol. 22, no.47, pp. 7359–7368, 2003.

[27] J. O. Armitage, “How I treat patients with diffuse large B-celllymphoma,” Blood, vol. 110, no. 1, pp. 29–36, 2007.

[28] R. Marcus, K. Imrie, P. Solal-Celigny et al., “Phase III studyof R-CVP compared with cyclophosphamide, vincristine,and prednisone alone in patients with previously untreatedadvanced follicular lymphoma,” Journal of Clinical Oncology,vol. 26, no. 28, pp. 4579–4586, 2008.

[29] W. Hiddemann, M. Kneba, M. Dreyling et al., “Front-line therapy with rituximab added to the combination ofcyclophosphamide, doxorubicin, vincristine, and prednisone(CHOP) significantly improves the outcome for patients withadvanced-stage follicular lymphoma compared with therapywith CHOP alone: results of a prospective randomized studyof the German Low-Grade Lymphoma Study Group,” Blood,vol. 106, no. 12, pp. 3725–3732, 2005.

[30] M. H. J. Van Oers, R. Klasa, R. E. Marcus et al., “Rituximabmaintenance improves clinical outcome of relapsed/resistantfollicular non-Hodgkin lymphoma in patients both with andwithout rituximab during induction: results of a prospectiverandomized phase 3 intergroup trial,” Blood, vol. 108, no. 10,pp. 3295–3301, 2006.

[31] R. Forstpointner, M. Unterhalt, M. Dreyling et al., “Main-tenance therapy with rituximab leads to a significant pro-longation of response duration after salvage therapy with acombination of rituximab, fludarabine, cyclophosphamide,and mitoxantrone (R-FCM) in patients with recurring andrefractory follicular and mantle cell lymphomas: results ofa prospective randomized study of the German Low GradeLymphoma Study Group (GLSG),” Blood, vol. 108, no. 13,pp. 4003–4008, 2006.

[32] M. Herold, A. Haas, S. Srock et al., “Rituximab addedto first-line mitoxantrone, chlorambucil, and prednisolonechemotherapy followed by interferon maintenance prolongssurvival in patients with advanced follicular lymphoma: anEast German study group hematology and oncology study,”Journal of Clinical Oncology, vol. 25, no. 15, pp. 1986–1992,2007.

[33] D. G. Maloney, A. J. Grillo-Lopez, C. A. White et al.,“IDEC-C2B8 (rituximab) anti-CD20 monoclonal antibodytherapy in patients with relapsed low-grade non-Hodgkin’slymphoma,” Blood, vol. 90, no. 6, pp. 2188–2195, 1997.

[34] M. Feuring-Buske, M. Kneba, M. Unterhalt et al., “IDEC-C2B8 (Rituximab) anti-CD20 antibody treatment in relapsedadvanced-stage follicular lymphomas: results of a phase-IIstudy of the German low-grade lymphoma study group,”Annals of Hematology, vol. 79, no. 9, pp. 493–500, 2000.

[35] P. McLaughlin, A. J. Grillo-Lopez, B. K. Link et al., “Ritux-imab chimeric anti-CD20 monoclonal antibody therapy forrelapsed indolent lymphoma: half of patients respond to afour-dose treatment program,” Journal of Clinical Oncology,vol. 16, no. 8, pp. 2825–2833, 1998.

[36] B. Coiffier, E. Lepage, J. Briere et al., “Chop chemotherapyplus rituximab compared with chop alone in elderly patientswith diffuse large-B-cell lymphoma,” New England Journal ofMedicine, vol. 346, no. 4, pp. 235–242, 2002.

[37] M. Pfreundschuh, L. Trumper, A. Osterborg et al., “CHOP-like chemotherapy plus rituximab versus CHOP-likechemotherapy alone in young patients with good-prognosisdiffuse large-B-cell lymphoma: a randomised controlled trialby the MabThera International Trial (MInT) Group,” LancetOncology, vol. 7, no. 5, pp. 379–391, 2006.

[38] T. M. Habermann, E. A. Weller, V. A. Morrison et al.,“Rituximab-CHOP versus CHOP alone or with maintenancerituximab in older patients with diffuse large B-cell lym-phoma,” Journal of Clinical Oncology, vol. 24, no. 19, pp.3121–3127, 2006.

[39] G. Lenz, M. Dreyling, E. Hoster et al., “Immunochemother-apy with rituximab and cyclophosphamide, doxorubicin,vincristine, and prednisone significantiy improves responseand time to treatment failure, but not long-term outcomein patients with previously untreated mantle cell lymphoma:results of a prospective randomized trial of the German LowGrade Lymphoma Study Group (GLSG),” Journal of ClinicalOncology, vol. 23, no. 9, pp. 1984–1992, 2005.

[40] T. Robak, A. Dmoszynska, P. Solal-Celigny et al., “Rit-uximab plus fludarabine and cyclophosphamide prolongsprogression-free survival compared with fludarabine andcyclophosphamide alone in previously treated chronic lym-phocytic leukemia,” Journal of Clinical Oncology, vol. 28, no.10, pp. 1756–1765, 2010.

[41] CLL8- German CLL Study Group.[42] E. Kimby, M. Bjorkholm, G. Gahrton et al., “Chloram-

bucil/prednisone vs. CHOP in symptomatic low-gradenon-Hodgkin’s lymphomas: a randomized trial from the


Lymphoma Group of Central Sweden,” Annals of Oncology,vol. 5, no. 2, pp. 67–71, 1994.

[43] M. S. Czuczman, R. Weaver, B. Alkuzweny, J. Berlfein, and A.J. Grillo-Lopez, “Prolonged clinical and molecular remissionin patients with low-grade or follicular non-Hodgkin’s lym-phoma treated with rituximab plus CHOP chemotherapy: 9-year follow-up,” Journal of Clinical Oncology, vol. 22, no. 23,pp. 4659–4664, 2004.

[44] M. S. Czuczman, A. Koryzna, A. Mohr et al., “Rituximab incombination with fludarabine chemotherapy in low-grade orfollicular lymphoma,” Journal of Clinical Oncology, vol. 23,no. 4, pp. 694–704, 2005.

[45] W. Hiddemann, C. Buske, M. Dreyling et al., “Treatmentstrategies in follicular lymphomas: current status and futureperspectives,” Journal of Clinical Oncology, vol. 23, no. 26, pp.6394–6399, 2005.

[46] E. Domingo-Domenech, E. Gonzalez-Borco, C. Estany, A.Sureda, J. Besalduch, and A. F. De Sevilla, “Combinedtreatment with anti CD20 (rituximab) and CHOP in relapsedadvanced-stage follicular lymphomas,” Haematologica, vol.87, no. 11, pp. 1229–1230, 2002.

[47] T. A. Davis, A. J. Grillo-Lopez, C. A. White et al., “Rituximabanti-CD20 monoclonal antibody therapy in non-Hodgkin’slymphoma: safety and efficacy of re-treatment,” Journal ofClinical Oncology, vol. 18, no. 17, pp. 3135–3143, 2000.

[48] T. A. Davis, C. A. White, A. J. Grillo-Lopez et al., “Single-agent monoclonal antibody efficacy in bulky non-Hodgkin’slymphoma: results of a phase II trial of rituximab,” Journal ofClinical Oncology, vol. 17, no. 6, pp. 1851–1857, 1999.

[49] N. Lamanna, J. G. Jurcic, A. Noy et al., “Sequential therapywith fludarabine, high-dose cyclophosphamide, and ritux-imab in previously untreated patients with chronic lympho-cytic leukemia produces high-quality responses: molecularremissions predict for durable complete responses,” Journalof Clinical Oncology, vol. 27, no. 4, pp. 491–497, 2009.

[50] T. Ahmadi and S. J. Schuster, “Variations on the fludarabine,cyclophosphamide, and rituximab combination in chroniclymphocytic leukemia therapy: what have we learned?”Journal of Clinical Oncology, vol. 27, no. 4, pp. 479–480, 2009.

[51] L. Kunkel, A. Wong, T. Maneatis et al., “Optimizing the use ofrituximab for treatment of B-cell non-Hodgkin’s lymphoma:a benefit-risk update,” Seminars in Oncology, vol. 27, no. 6,pp. 53–61, 2000.

[52] D. Bedognetti, G. Zoppoli, C. Massucco et al., “Patientstreated with rituximab-containing immunochemotherapyhave a significant and prolonged lack of humoral responseto influenza vaccine associated with a persistent depletion ofb memory cells,” in Blood (ASH Annual Meeting Abstracts),November 2009.

[53] B. D. Cheson and J. P. Leonard, “Monoclonal antibodytherapy for B-cell non-Hodgkin’s lymphoma,” New EnglandJournal of Medicine, vol. 359, no. 6, pp. 613–626, 2008.

[54] S. A. Beers, R. R. French, H. T. C. Chan et al., “Antigenicmodulation limits the efficacy of anti-CD20 antibodies:implications for antibody selection,” Blood, vol. 115, no. 25,pp. 5191–5201, 2010.

[55] S. Y. Wang, S. Veeramani, E. Racila et al., “Depletion ofthe C3 component of complement enhances the ability ofrituximab-coated target cells to activate human NK cells andimproves the efficacy of monoclonal antibody therapy in anin vivo model,” Blood, vol. 114, no. 26, pp. 5322–5330, 2009.

[56] A. Rambaldi, M. Lazzari, C. Manzoni et al., “Monitoringof minimal residual disease after CHOP and rituximab in

previously untreated patients with follicular lymphoma,”Blood, vol. 99, no. 3, pp. 856–862, 2002.

[57] C. G. A. Price, J. Meerabux, S. Murtagh et al., “The signifi-cance of circulating cells carrying t(14;18) in long remissionfrom follicular lymphoma,” Journal of Clinical Oncology, vol.9, no. 9, pp. 1527–1532, 1991.

[58] J. Limpens, R. Stad, C. Vos et al., “Lymphoma-associatedtranslocation t(14;18) in blood B cells of normal individuals,”Blood, vol. 85, no. 9, pp. 2528–2536, 1995.

[59] M. H. van Oers, E. Tonnissen, M. Van Glabbeke et al.,“BCL-2/IgH polymerase chain reaction status at the end ofinduction treatment is not predictive for progression-freesurvival in relapsed/resistant follicular lymphoma: results ofa prospective randomized EORTC 20981 phase III intergroupstudy,” Journal of Clinical Oncology, vol. 28, no. 13, pp. 2246–2252, 2010.

[60] N. Mounier, J. Briere, C. Gisselbrecht et al., “Rituximab plusCHOP (R-CHOP) overcomes bcl-2-associated resistance tochemotherapy in elderly patients with diffuse large B-celllymphoma (DLBCL),” Blood, vol. 101, no. 11, pp. 4279–4284,2003.

[61] N. Mounier, J. Briere, C. Gisselbrecht, F. Reyes, P. Gaulard,and B. Coiffier, “Estimating the impact of rituximab on bcl-2-associated resistance to CHOP in elderly patients withdiffuse large B-cell lymphoma,” Haematologica, vol. 91, no.5, pp. 715–716, 2006.

[62] T. Kewalramani, A. D. Zelenetz, E. E. Hedrick et al., “High-dose chemoradiotherapy and autologous stem cell transplan-tation for patients with primary refractory aggressive non-Hodgkin lymphoma: an intention-to-treat analysis,” Blood,vol. 96, no. 7, pp. 2399–2404, 2000.

[63] C. Haioun, E. Lepage, C. Gisselbrecht et al., “Survivalbenefit of high-dose therapy in poor-risk aggressive non-Hodgkin’s lymphoma: final analysis of the prospectiveLNH87-2 protocol—a groupe d’etude des lymphomes del’adulte study,” Journal of Clinical Oncology, vol. 18, no. 16,pp. 3025–3030, 2000.

[64] A. Nademanee and S. J. Forman, “Role of HematopoieticStem Cell Transplantation for Advanced-Stage Diffuse LargeCell B-Cell Lymphoma-B,” Seminars in Hematology, vol. 43,no. 4, pp. 240–250, 2006.

[65] N. L. Berinstein, “Principles of maintenance therapy,”Leukemia Research, vol. 30, no. 1, pp. S3–S10, 2006.

[66] J. D. Hainsworth, S. Litchy, H. A. Burris et al., “Rituximab asfirst-line and maintenance therapy for patients with indolentnon-Hodgkin’s lymphoma,” Journal of Clinical Oncology, vol.20, no. 20, pp. 4261–4267, 2002.

[67] J. D. Hainsworth, S. Litchy, D. W. Shaffer, L. Van Lackey,M. Grimaldi, and F. A. Greco, “Maximizing therapeutic ben-efit of rituximab: maintenance therapy versus re-treatmentat progression in patients with indolent non-Hodgkin’slymphoma—a randomized phase II trial of the Minnie PearlCancer Research Network,” Journal of Clinical Oncology, vol.23, no. 6, pp. 1088–1095, 2005.

[68] C. Buske and W. Hiddemann, “Rituximab maintenancetherapy in indolent NHL: a clinical review,” LeukemiaResearch, vol. 30, no. 1, pp. S11–S15, 2006.

[69] M. Ghielmini, S. F. H. Schmitz, S. B. Cogliatti et al., “Pro-longed treatment with rituximab in patients with follicularlymphoma significantly increases event-free survival andresponse duration compared with the standard weekly × 4schedule,” Blood, vol. 103, no. 12, pp. 4416–4423, 2004.


[70] K. A. Foon and M. J. Hallek, “Changing paradigms in thetreatment of chronic lymphocytic leukemia,” Leukemia, vol.24, no. 3, pp. 500–511, 2010.

[71] H. Schulz, S. K. Klein, U. Rehwald et al., “Phase 2 studyof a combined immunochemotherapy using rituximab andfludarabine in patients with chronic lymphocytic leukemia,”Blood, vol. 100, no. 9, pp. 3115–3120, 2002.

[72] N. Lamanna, M. Kalaycio, P. Maslak et al., “Pentostatin,cyclophosphamide, and rituximab is an active, well-toleratedregimen for patients with previously treated chronic lympho-cytic leukemia,” Journal of Clinical Oncology, vol. 24, no. 10,pp. 1575–1581, 2006.

[73] T. Robak, P. Smolewski, B. Cebula, O. Grzybowska-Izydorczyk, and J. Z. Błonski, “Rituximab plus cladribinewith or without cyclophosphamide in patients with relapsedor refractory chronic lymphocytic leukemia,” European Jour-nal of Haematology, vol. 79, no. 2, pp. 107–113, 2007.

[74] M. Hallek, G. Fingerle-Rowson, and A.-M. Fink, “First-line treatment with fludarabine (f), cyclophosphamide (c),and rituximab (r) (fcr) improves overall survival (os) inpreviously untreated patients (pts) with advanced chroniclymphocytic leukemia (cll): results of a randomized phase iiitrial on behalf of an international group of investigators andthe german cll study group,” Blood, vol. 114, 2009.

[75] M. J. Keating, S. O’Brien, M. Albitar et al., “Early results ofa chemoimmunotherapy regimen of fludarabine, cyclophos-phamide, and rituximab as initial therapy for chroniclymphocytic leukemia,” Journal of Clinical Oncology, vol. 23,no. 18, pp. 4079–4088, 2005.

[76] C. S. Tam, S. O’Brien, W. Wierda et al., “Long-term results ofthe fludarabine, cyclophosphamide, and rituximab regimenas initial therapy of chronic lymphocytic leukemia,” Blood,vol. 112, no. 4, pp. 975–980, 2008.

[77] W. Wierda, S. O’Brien, S. Wen et al., “Chemoimmunother-apy with fludarabine, cyclophosphamide, and rituximabfor relapsed and refractory chronic lymphocytic leukemia,”Journal of Clinical Oncology, vol. 23, no. 18, pp. 4070–4078,2005.

[78] P. V. Beum, M. A. Lindorfer, and R. P. Taylor, “Withinperipheral blood mononuclear cells, antibody-dependentcellular cytotoxicity of rituximab-opsonized Daudi cells ispromoted by NK cells and inhibited by monocytes due toshaving,” Journal of Immunology, vol. 181, no. 4, pp. 2916–2924, 2008.

[79] A. D. Kennedy, P. V. Beum, M. D. Solga et al., “Rituximabinfusion promotes rapid complement depletion and acuteCD20 loss in chronic lymphocytic leukemia,” Journal ofImmunology, vol. 172, no. 5, pp. 3280–3288, 2004.

[80] P. V. Beum, A. D. Kennedy, and R. P. Taylor, “Three newassays for rituximab based on its immunological activityor antigenic properties: analyses of sera and plasmas ofRTX-treated patients with chronic lymphocytic leukemia andother B cell lymphomas,” Journal of Immunological Methods,vol. 289, no. 1-2, pp. 97–109, 2004.

[81] R. Shimizu, J. Kikuchi, T. Wada, K. Ozawa, Y. Kano, and Y.Furukawa, “HDAC inhibitors augment cytotoxic activity ofrituximab by upregulating CD20 expression on lymphomacells,” Leukemia, vol. 24, no. 10, pp. 1760–1768, 2010.

[82] C. Henry, M. Deschamps, P. S. Rohrlich et al., “Identificationof an alternative CD20 transcript variant in B-cell malig-nancies coding for a novel protein associated to rituximabresistance,” Blood, vol. 115, no. 12, pp. 2420–2429, 2010.

[83] R. Bhat and C. Watzl, “Serial killing of tumor cells by humannatural killer cells—enhancement by therapeutic antibodies,”PLoS ONE, vol. 2, no. 3, article no. e326, 2007.

[84] J. A. Bowles and G. J. Weiner, “CD16 polymorphisms and NKactivation induced by monoclonal antibody-coated targetcells,” Journal of Immunological Methods, vol. 304, no. 1-2,pp. 88–99, 2005.

[85] L. Fischer, O. Penack, C. Gentilini et al., “The anti-lymphomaeffect of antibody-mediated immunotherapy is based on anincreased degranulation of peripheral blood natural killer(NK) cells,” Experimental Hematology, vol. 34, no. 6, pp. 753–759, 2006.

[86] J. G. Berdeja, A. Hess, D. M. Lucas et al., “Systemicinterleukin-2 and adoptive transfer of lymphokine-activatedkiller cells improves antibody-dependent cellular cytotoxicityin patients with relapsed B-cell lymphoma treated withrituximab,” Clinical Cancer Research, vol. 13, no. 8, pp. 2392–2399, 2007.

[87] H. Wang, Y. Liu, Z. Y. Li, X. Fan, A. Hemminki, andA. Lieber, “A recombinant adenovirus type 35 fiber knobprotein sensitizes lymphoma cells to rituximab therapy,”Blood, vol. 115, no. 3, pp. 592–600, 2010.

[88] F. Sato, A. Ito, T. Ishida et al., “A complement-dependentcytotoxicity-enhancing anti-CD20 antibody mediatingpotent antitumor activity in the humanized NOD/Shi-scid,IL- 2Rγnull mouse lymphoma model,” Cancer Immunology,Immunotherapy, vol. 59, no. 12, pp. 1791–1800, 2010.

[89] M. P. Chao, A. A. Alizadeh, C. Tang et al., “Anti-CD47antibody synergizes with rituximab to promote phagocytosisand eradicate non-Hodgkin lymphoma,” Cell, vol. 142, no. 5,pp. 699–713, 2010.

[90] C. Xuan, K. K. Steward, J. M. Timmerman, and S. L.Morrison, “Targeted delivery of interferon-alpha via fusionto anti-CD20 results in potent antitumor activity against B-cell lymphoma,” Blood, vol. 115, no. 14, pp. 2864–2871, 2010.

[91] B. Li, L. Zhao, H. Guo et al., “Characterization of a rituximabvariant with potent antitumor activity against rituximab-resistant B-cell lymphoma,” Blood, vol. 114, no. 24, pp. 5007–50515, 2009.

[92] R. Stein, Z. Qu, S. Chen et al., “Characterization of a newhumanized anti-CD20 monoclonal antibody, IMMU-106,and its use in combination with the humanized anti-CD22antibody, epratuzumab, for the therapy of non-Hodgkin’slymphoma,” Clinical Cancer Research, vol. 10, no. 8, pp.2868–2878, 2004.

[93] S. Maddipatla, F. J. Hernandez-Ilizaliturri, J. Knight, andM. S. Czuczman, “Augmented antitumor activity against b-cell lymphoma by a combination of monoclonal antibodiestargeting trail-r1 and cd20,” Clinical Cancer Research, vol. 13,no. 15, pp. 4556–4564, 2007.

[94] E. Tobin, G. L. DeNardo, N. Zhang, A. L. Epstein, C. Liu,and S. DeNardo, “Combination immunotherapy with anti-CD20 and anti-HLA-DR monoclonal antibodies inducessynergistic anti-lymphoma effects in human lymphoma celllines,” Leukemia and Lymphoma, vol. 48, no. 5, pp. 944–956,2007.

[95] E. A. Rossi, D. M. Goldenberg, T. M. Cardillo, W. J.McBride, R. M. Sharkey, and C. H. Chang, “Stably tetheredmultifunctional structures of defined composition made bythe dock and lock method for use in cancer targeting,”Proceedings of the National Academy of Sciences of the UnitedStates of America, vol. 103, no. 18, pp. 6841–6846, 2006.

[96] E. A. Rossi, D. M. Goldenberg, T. M. Cardillo, R. Stein, andC. H. Chang, “Hexavalent bispecific antibodies represent a


new class of anticancer therapeutics—1. Properties of anti-CD20/CD22 antibodies in lymphoma,” Blood, vol. 113, no.24, pp. 6161–6171, 2009.

[97] E. A. Rossi, D. M. Goldenberg, T. M. Cardillo, R. Stein,Y. Wang, and C. H. Chang, “Novel designs of multivalentanti-CD20 humanized antibodies as improved lymphomatherapeutics,” Cancer Research, vol. 68, no. 20, pp. 8384–8392, 2008.

[98] P. Gupta, D. M. Goldenberg, E. A. Rossi, and C.-H.Chang, “Multiple signaling pathways induced by hexavalent,monospecific, anti-CD20 and hexavalent, bispecific, anti-CD20/CD22 humanized antibodies correlate with enhancedtoxicity to B-cell lymphomas and leukemias,” Blood, vol. 116,no. 17, pp. 3258–3267, 2010.

[99] J. P. Leonard, M. Coleman, J. C. Ketas et al., “Phase I/II trial ofepratuzumab (humanized anti-CD22 antibody) in indolentnon-Hodgkin’s lymphoma,” Journal of Clinical Oncology, vol.21, no. 16, pp. 3051–3059, 2003.

[100] J. P. Leonard, M. Coleman, J. C. Ketas et al., “Epratuzumab,a humanized anti-CD22 antibody, in aggressive non-Hodgkin’s lymphoma: phase I/II clinical trial results,” ClinicalCancer Research, vol. 10, no. 16, pp. 5327–5334, 2004.

[101] A. Younes, K. Hariharan, R. S. Allen, and B. R. Leigh,“Initial trials of anti-CD80 monoclonal antibody (galiximab)therapy for patients with relapsed or refractory follicularlymphoma,” Clinical Lymphoma, vol. 3, no. 4, pp. 257–259,2003.

[102] F. A. Vyth-Dreese, T. A. M. Dellemijn, D. Majoor, andD. De Jong, “Localization in situ of the co-stimulatorymolecules B7.1, B7.2, CD40 and their ligands in normalhuman lymphoid tissue,” European Journal of Immunology,vol. 25, no. 11, pp. 3023–3029, 1995.

[103] J. M. Munro, A. S. Freedman, J. C. Aster et al., “In vivoexpression of the B7 costimulatory molecule by subsets ofantigen- presenting cells and the malignant cells of Hodgkin’sdisease,” Blood, vol. 83, no. 3, pp. 793–798, 1994.

[104] J. P. Leonard, J. W. Friedberg, A. Younes et al., “A phaseI/II study of galiximab (an anti-CD80 monoclonal antibody)in combination with rituximab for relapsed or refractory,follicular lymphoma,” Annals of Oncology, vol. 18, no. 7, pp.1216–1223, 2007.

[105] M. J. Keating, I. Flinn, V. Jain et al., “Therapeutic role ofalemtuzumab (Campath-1H) in patients who have failedfludarabine: results of a large international study,” Blood, vol.99, no. 10, pp. 3554–3561, 2002.

[106] J. Lundin, H. Hagberg, R. Repp et al., “Phase 2 study ofalemtuzumab (anti-CD52 monoclonal antibody) in patientswith advanced mycosis fungoides/Sezary syndrome,” Blood,vol. 101, no. 11, pp. 4267–4272, 2003.

[107] G. Enblad, H. Hagberg, M. Erlanson et al., “A pilot study ofalemtuzumab (anti-CD52 monoclonal antibody) therapy forpatients with relapsed or chemotherapy-refractory periph-eral T-cell lymphomas,” Blood, vol. 103, no. 8, pp. 2920–2924,2004.

[108] A. Gallamini, F. Zaja, C. Patti et al., “Alemtuzumab(Campath-1H) and CHOP chemotherapy as first-line treat-ment of peripheral T-cell lymphoma: results of a GITIL(Gruppo Italiano Terapie Innovative nei Linfomi) prospec-tive multicenter trial,” Blood, vol. 110, no. 7, pp. 2316–2323,2007.

[109] R. Stein, Z. Qu, S. Chen, D. Solis, H. J. Hansen, and D.M. Goldenberg, “Characterization of a humanized IgG4anti-HLA-DR monoclonal antibody that lacks effector cellfunctions but retains direct antilymphoma activity and

increases the potency of rituximab,” Blood, vol. 108, no. 8,pp. 2736–2744, 2006.

[110] P. Hillmen, A. B. Skotnicki, T. Robak et al., “Alemtuzumabcompared with chlorambucil as first-line therapy for chroniclymphocytic leukemia,” Journal of Clinical Oncology, vol. 25,no. 35, pp. 5616–5623, 2007.

[111] K. Dunleavy, T. White, N. Grant et al., “Phase 1 study of com-bination rituximab with apolizumab in relapsed/refractoryB-cell lymphoma and chronic lymphocytic leukemia,” vol.23, 2005, ASCO Meeting Abstract 6607.

[112] G. L. DeNardo, S. J. DeNardo, D. S. Goldstein et al.,“Maximum-tolerated dose, toxicity, and efficacy of I-Lym-1 antibody for fractionated radioimmunotherapy of non-Hodgkin’s lymphoma,” Journal of Clinical Oncology, vol. 16,no. 10, pp. 3246–3256, 1998.

[113] R. Schnell, M. Dietlein, K. Schomacker et al., “Yttrium-90 ibritumomab tiuxetan-induced complete remission ina patient with classical lymphocyte-rich Hodgkin’s lym-phoma,” Onkologie, vol. 31, no. 1-2, pp. 49–51, 2008.

[114] D. Ma, M. R. McDevitt, E. Barendswaard et al., “Radioim-munotherapy for model B cell malignancies using Y-labeled anti-CD19 and anti-CD20 monoclonal antibodies,”Leukemia, vol. 16, no. 1, pp. 60–66, 2002.

[115] T. E. Witzig, L. I. Gordon, F. Cabanillas et al., “Randomizedcontrolled trial of yttrium-90-labeled ibritumomab tiuxetanradioimmunotherapy versus rituximab immunotherapy forpatients with relapsed or refractory low-grade, follicular, ortransformed B-cell non-Hodgkin’s lymphoma,” Journal ofClinical Oncology, vol. 20, no. 10, pp. 2453–2463, 2002.

[116] T. E. Witzig, A. Molina, L. I. Gordon et al., “Long-termresponses in patients with recurring or refractory B-cell non-Hodgkin lymphoma treated with yttrium 90 ibritumomabtiuxetan,” Cancer, vol. 109, no. 9, pp. 1804–1810, 2007.

[117] F. Morschhauser, T. Illidge, D. Huglo et al., “Efficacy andsafety of yttrium-90 ibritumomab tiuxetan in patients withrelapsed or refractory diffuse large B-cell lymphoma notappropriate for autologous stem-cell transplantation,” Blood,vol. 110, no. 1, pp. 54–58, 2007.

[118] P. L. Zinzani, M. Tani, S. Fanti et al., “A phase II trial of CHOPchemotherapy followed by yttrium 90 ibritumomab tiuxetan(Zevalin) for previously untreated elderly diffuse large B-celllymphoma patients,” Annals of Oncology, vol. 19, no. 4, pp.769–773, 2008.

[119] S. J. Horning, A. Younes, V. Jain et al., “Efficacy and safetyof tositumomab and iodine-131 tositumomab (Bexxar) inB-cell lymphoma, progressive after rituximab,” Journal ofClinical Oncology, vol. 23, no. 4, pp. 712–719, 2005.

[120] M. S. Kaminski, M. Tuck, J. Estes et al., “131I-tositumomabtherapy as initial treatment for follicular lymphoma,” NewEngland Journal of Medicine, vol. 352, no. 5, pp. 441–449,2005.

[121] O. W. Press, J. M. Unger, R. M. Braziel et al., “Phase II trialof CHOP chemotherapy followed by tositumomab/iodine I-131 tositumomab for previously untreated follicular non-Hodgkin’s lymphoma: five-year follow-up of SouthwestOncology Group protocol S9911,” Journal of Clinical Oncol-ogy, vol. 24, no. 25, pp. 4143–4149, 2006.

[122] A. K. Gopal, J. M. Pagel, J. R. Fromm, S. Wilbur, and O.W. Press, “131I anti-CD45 radioimmunotherapy effectivelytargets and treats T-cell non-Hodgkin lymphoma,” Blood,vol. 113, no. 23, pp. 5905–5910, 2009.

[123] T. Witzig, C. Emmanouilides, A. Molina, L. Gordon, I.Gaston, and I. Flinn, “Yttrium-90 ibritumomab tiuxe-tan radioimmunotherapy (RIT) induces durable complete


responses (CR/CRu) in patients with relapsed or refractoryB-cell non-Hodgkin’s lymphoma (NHL),” Proceedings of theAmerican Society of Clinical Oncology, vol. 22, p. 597, 2003.

[124] M. S. Kaminski, A. D. Zelenetz, O. W. Press et al., “Pivotalstudy of iodine I 131 tositumomab for chemotherapy-refractory low-grade or transformed low-grade B-cell non-Hodgkin’s lymphomas,” Journal of Clinical Oncology, vol. 19,no. 19, pp. 3918–3928, 2001.

[125] M. S. Kaminski, M. Tuck, J. Estes et al., “131I-tositumomabtherapy as initial treatment for follicular lymphoma,” NewEngland Journal of Medicine, vol. 352, no. 5, pp. 441–449,2005.

[126] M. E. Juweid, E. Stadtmauer, G. Hajjar et al., “Pharma-cokinetics, dosimetry, and initial therapeutic results with I-and In-/Y-labeled humanized LL2 anti-CD22 monoclonalantibody in patients with relapsed, refractory non-Hodgkin’slymphoma,” Clinical Cancer Research, vol. 5, no. 10, pp.3292s–3303s, 1999.

[127] R. O’Donnell, G. DeNardo, D. Kukis et al., “67copper-2-iminothiolane-6-[p-(bromoacetamido) benzyl]-teta-lym-1 for radioimmunotherapy of non-hodgkin’s lymphoma,”Clinical Cancer Research, vol. 5, no. 10, p. 3330s, 1999.

[128] R. T. O’Donnell, G. L. DeNardo, D. L. Kukis et al., “A clinicaltrial of radioimmunotherapy with Cu-2IT-BAT-Lym-1 fornon-Hodgkin’s lymphoma,” Journal of Nuclear Medicine, vol.40, no. 12, pp. 2014–2020, 1999.

[129] E. Aurlien, R. H. Larsen, G. Kvalheim, and Ø. S. Bruland,“Demonstration of highly specific toxicity of the α-emit-ting radioimmunoconjugate 211At-rituximab against non-Hodgkin’s lymphoma cells,” British Journal of Cancer, vol. 83,no. 10, pp. 1375–1379, 2000.

[130] S. Knop, A. Jakob, L. Kanz, H. Hebart, R. Bares, and B.Dohmen, “186Rhenium-labeled anti-CD20 antibody radi-oimmunotherapy followed by autologous peripheral bloodstem cell transplantation in patients with relapsed or refrac-tory non-Hodgkin lymphoma,” Blood, vol. 103, no. 3, p.1175, 2004.


Review Article

Immunotherapy of Brain Cancers: The Past, the Present, andFuture Directions

Lisheng Ge,1 Neil Hoa,1 Daniela A. Bota,2 Josephine Natividad,1 Andrew Howat,1

and Martin R. Jadus1, 3, 4

1 Pathology and Laboratory Medicine Service, Department of Diagnostic and Molecular Medicine Health Care Group,VA Long Beach Healthcare System, 5901 E. 7th Street, Long Beach, CA 90822, USA

2 Department of Neurology and Department of Neurological Surgery, Chao Family Comprehensive Cancer Center,UC Irvine School of Medicine, University of California, Irvine, CA 92697, USA

3 Chao Family Comprehensive Cancer Center, UC, Irvine School of Medicine, University of California, Irvine, CA 92697, USA4 Pathology and Laboratory Medicine, University of California, Irvine, CA 92697, USA

Correspondence should be addressed to Martin R. Jadus, [email protected]

Received 30 June 2010; Accepted 30 December 2010

Academic Editor: Bernhard Fleischer

Copyright © 2010 Lisheng Ge et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Treatment of brain cancers, especially high grade gliomas (WHO stage III and IV) is slowly making progress, but not as fast asmedical researchers and the patients would like. Immunotherapy offers the opportunity to allow the patient’s own immune systema chance to help eliminate the cancer. Immunotherapy’s strength is that it efficiently treats relatively small tumors in experimentalanimal models. For some patients, immunotherapy has worked for them while not showing long-term toxicity. In this paper, wewill trace the history of immunotherapy for brain cancers. We will also highlight some of the possible directions that this field maybe taking in the immediate future for improving this therapeutic option.

1. Introduction

Immunotherapy for cancer has made progress and is nowbecoming part of the treatment options that are morefrequently discussed with oncology patients. Previously, thistype of treatment was given to patients with advanced dis-ease, with only a few months to live. Needless to say, the finalresults were often disappointing. While the failures told uswhat strategies did not work, it showed that immunotherapywas generally safe and did not immediately kill the patient.It also showed that the dreaded autoimmunity was notbeing induced. These results also spurred the developmentof different approaches, after better understandings of cancerimmunology were unexpectedly discovered. This illustratesour need to learn more about basic in vivo cancer immunol-ogy before clinical therapies can be fully predicted. Theproper timing and use of the right antibodies or cells hasalso allowed this progress to occur. The herceptin antibodytargeting the her2/neu proto-oncogene has benefited thosewomen with breast and ovarian cancers that overexpressed

this receptor. This discovery showed that targeting a cell-surface receptor controlling a key biological function, asopposed to any available tumor surface antigen, was the keyto generating useful clinical responses. Recently, PROVENGEmarketed by Denderon Corp, was given FDA approvalin the USA to treat refractory prostate cancer in men.This prostate tumor-antigen (prostatic acid phosphatase)-granulocyte macrophage-colony stimulating factor fusionprotein does stimulate dendritic cells in vitro. When these exvivo activated dendritic cells are reintroduced back into thepatient, the host’s antitumor T cells are restimulated, whichsubsequently attacks the cancer. This immune response doestranslate into an additional four months of life. These twosuccess stories demonstrate that progress towards cancer isslowly advancing and we eagerly await more successes as theoverall field continues to advance and mature.

Glioblastoma multiforme (GBM, WHO stage IV) andanaplastic astrocytomas (WHO stage III) are aggressiveand lethal cancers. These cancers are almost always fatalwithin five years (2010 Central Brain Tumor Registry). These


tumors are very invasive; this contributes to their resistanceto be cured by traditional surgical resection and directedradiation therapy. Hence the need to develop better therapiesstill exists. The advantage of generating an immune responsetowards a cancer is that the immune effectors (cells orantibodies) can now seek out and destroy the tumor cellsthat are located in inaccessible sites that traditional surgery,radiation, or chemotherapeutic drugs cannot reach.

Due to the relative isolation from the systemic circula-tion, because of the blood brain barrier, the initiation of pro-ductive immune responses in the brain is more limited thanother types of cancers [1]. Local microglial cells can processand present tumor-associated antigens to T lymphocytes [2–5]. However few naıve T cells normally transit into the brain.Normal brain cells also express Fas Ligand and express TGF-β [6, 7], making immune responses harder to be sustained.Hence lymphoid cells must be recruited from the peripheryby a variety of cytokines and chemokines. Once effectorlymphocytes infiltrate the tumor, they can mediate antibraintumor immunity. Despite these obstacles, progress is slowlybeing made in neuro-onco-immunotherapy. Unless someextraordinary discovery is made, immune-based therapiesmust be combined with other modalities that target othercritical aspects of cancer biology. This paper will focus on thenatural progressions that are leading us towards successfulimmunotherapy for brain cancers.

2. Types of Immunotherapy

Immunological-based treatments have been used in severalways to treat cancer. These include (1) nonspecific methodsusing adjuvants, lymphokine activated killer cells, or gene-modified tumor cells; (2) specific immunotherapy includeusing monoclonal antibodies, tumor infiltrating lympho-cytes, allogeneic reactive T cells, chimeric antigen-redirectedT cells, purified and cloned tumor antigens used eitheralone or in combination with in vitro cultured dendritic cells(DCs).

2.1. Nonspecific Approaches

2.1.1. Adjuvants. Nonspecific approaches include using nat-ural adjuvants such as bacillus Calmette-Guerin (BCG,Mycobacteria bovis), muramyl dipeptide (MDP), Detox(lipopolysaccharide with lipid A removed). Janeway [8] oncewrote: “adjuvants are the immunologist’s dirty little secret,”in that these molecules are needed to provoke the immuneresponse. Adjuvants work by stimulating local antigen pre-senting cells such as dendritic cells, macrophages, and Bcells via toll-like receptors and pathogen-associated recog-nition molecules. TLR receptor stimulants as imiquimod,polyinosinic-polycytidylic acid stabilized with polylysine andcarboxymethyl-cellulose (Poly ICLC), CpG containing DNA,and other synthetic molecules are used to stimulate dendriticcells, which activate either cell-mediated or humoral immu-nity provided the antigens of choice are also present.

In some parts of the world, BCG is used as a prophylacticvaccine to stimulate immunity towards Mycobacteria tuber-

culosis, due to its very strong immunogenic properties, aswell as common antigenic determinants. A purified proteinderived (PPD) from M. tuberculosis is the difference betweenFreund’s complete and incomplete adjuvant commonly usedin antibody production in animals. BCG is also used in atherapeutic vaccination setting to actively treat human blad-der cancer [9]. Here the initial nonspecific inflammation inresponse to BCG injected directly into the tumor leads to aninnate immune response that causes tumor cell death. Afterthe tumor dies it is followed by lasting cellular immunitytowards the bladder cancer. Wikstrand and Bigner [10] usedBCG and human glioma cells to generate good antibodyresponses towards the glioma cells without any signs ofautoimmunity. Albright et al. [11] used BCG to treat GBMpatients in 1976. Here 107 BCG organisms were used as anintradermal stimulus to induce delayed type hypersensitivity(DTH) reactions. Their patients were subsequently injectedwith autologous glioma cells with the purified proteinderivative (PPD). The theory here was that when the hostmounted the recall response towards the PPD, the gliomaantigens would be incorporated in this DTH response andmount a primary immune response towards the glioma cells.Once these antibodies and immunized lymphocytes wereelicited, they would then home into the brain tumors andmediate their antitumor effects. Unfortunately, this therapyfailed to achieve much inflammation within the relapsingglioma along with no improved patient survival. The use ofadjuvants was largely abandoned, until its ability to stimulatedendritic cell maturation via different receptors was recentlydiscovered (see dendritic cells).

2.1.2. Natural Adjuvants: Heat Shock Proteins (HSP). Heatshock proteins are induced by a variety of stressful con-ditions: heat, radiation, chemotherapy, nutrient starvation,hypoxia, and so forth. These molecules are responsiblefor assisting in the synthesis and correct folding of newlysynthesized proteins, thereby replacing the stress-damagedproteins.

The HSPs were described as “natural adjuvants,” sincethey provoke immunity [12]. Several of these HSPs: HSP70and gp96 (also known as GRP94) were identified as tumor-specific antigens [13]. In other cases, HSP70, HSP90, andgp96 increased the immunogenicity of tumors by improvingT-cells immune responses [14, 15]. Proteins synthesized bytumor cells include potential antigens. As these antigensare degraded by the proteasome, it was reported that HSPsacted as chaperones [14, 15]; these molecular chaperonesthen shuttle these antigenic peptides to the endoplasmicreticulum, where they can be eventually loaded onto theMHC. Tumor cells can present HSPs on their surfaces(our unpublished data), release these peptide/HSP com-plexes, presumably via exosomes [16–18]. Here the host’santigen presenting cells (APCs), both dendritic cells andmacrophages, can take up these antigens/HSPs via the CD91receptor [19]. Peptides complexed to HSPs stimulate betterimmune responses than when the antigenic peptide is notcomplexed to the HSP [13]. HSP70 and gp96 also havebeen reported to enhance dendritic cell maturation [20, 21].HSP70 also acts as a cytokine, which stimulates tumor


necrosis factor, IL-1 and IL-6 production from CD14+monocytes [22]. These CD14+ cells also include the imma-ture dendritic cells. Thus, HSPs might be able to stimulateimmune responses via its natural adjuvant activity, whilesimultaneously delivering the antigenic peptides provokingspecific immune responses.

On the clinical level, Antigenics, Inc. (Lexington, MA)has developed the technology to produce a clinical productcalled Oncophage. Here the surgically resected tumor is takenand the gp96 component (HSPPC-96) from the tumor isisolated. This purified gp96 (presumably binding antigenicpeptides) is then formulated as a custom-made vaccine usingthe GS-21 adjuvant for each patient. This approach has beengiven fast track and orphan drug designations by the US FDAand the EMEA for a couple of cancers. At the University ofCalifornia, San Francisco, Oncophage is being used in Phase2 trials to treat GBM. At the recent International Conferenceon Brain Tumor in Travemunde, Germany (May 2010), itwas reported that of the 32 evaluable patients with recurrentGBM given Oncophage, 41% survived up to a year or longer[23]. The authors also saw a robust immune response withinincreased Th1 cytokine production by the immunized T cells.

2.1.3. Lymphokine Activated Killer Cells (LAK). LAK cells areNK and NK-T like (CD8+) cells that when stimulated withlymphokines (cytokines) like interleukin-2 (IL-2) becomenonspecific tumoricidal cells. LAK cells kill most, if not all,tumor cells quite well in vitro in a non-MHC restrictedmanner. When IL-2 or interferon-γ (IFN-γ) transducedfibroblasts were coinjected with the murine GL261 glioma,the glioma cells were rejected by the recruited NK and LAKcells. The LAK cells were activated in situ by the cytokines[24]. However in a rat glioma model using the F98 glioma cellline, the recruited rat LAK cells were not as successful as theprevious mouse model [25]. The clinical application of LAKcells has been effective only towards some melanoma andrenal cancers [26]. Occasionally a response towards a humanglioma is seen [27, 28]. Hoag Hospital in Newport Beach(California) is currently using LAK cells that are implantedinto their patient’s brain tumor cavity after surgery [29, 30].The main disadvantage of LAK cells, is that they releasemultiple cytokines (IFN-γ, tumor necrosis factor-α (TNF-α)), which cause many of the unwanted pharmacologicaltoxicities associated with this clinical therapy.

2.1.4. γδ T Cells. Normally the T cells that we think about,are those T cells with the classic αβ T-cell receptor (TCR)rearrangements. These cells normally circulate through theblood and reside in the lymph nodes and spleen. These cellsreside in many tumors as the tumor-infiltrating lymphocytes(TILs). But another cell type also goes through the samethymus-education pathway, except that these cells utilizetheir rearranged γδ T-cell receptors to recognize theirantigens. These γδ T-cell receptors are more restricted intheir TCR diversity and are not MHC restricted, althoughthey may recognize nonclassic HLA-E and HLA-G molecules.These lymphocytes were initially discovered to be cytotoxictowards leukemia cells, but Fujimiya and colleagues [31] dis-

covered that these cells also had the ability to recognize andkill glioma cells in vitro. Several of the ligands that γδ T cellscan recognize tumor cells (MICA, MICB, and UL-16 bindingproteins) are also found on gliomas [32]. In the UnitedStates, Lamb and colleagues [33, 34] confirmed the previousstudy. Their human γδ T cells failed to kill normal astrocytes.They also discovered that the γδ T cells can be expanded inthe presence of low doses of IL-2 and zoledronic acid, so thatsufficient number of cells could be generated for infusionback into patients. Human γδ T cells when implanted intonude mice showed immunological efficacy against U251xenografts [35]. This non-MHC restricted killing by γδ Tcells opens up the possibility that allogeneic donors couldbe used for therapeutic purposes in gliomas without riskingthe possibility of graft-versus-host reactions or autoimmunediseases. Clinical trials using this approach against braincancers are expected to begin in the summer/fall of 2011 atthe University of Alabama, Birmingham.

2.1.5. Gene Therapy. Gene therapy using various cytokinesand costimulatory molecules was used in experimentalglioma models to induce stronger immune responses. IL-2 and IFN-γ transduced rat RG2 (also known as D74)glioma cells, when injected into the brains of naıve rats,resulted in premature death of the rats due to changes in thevasculature of the brain [36]. Peripheral vaccination usingN32 rat glioma cells transduced with IFN-γ and interleukin7 induced intracranial rejection of the parental N32 glioma[37]. The membrane form of macrophage colony stimulatingfactor (mM-CSF) as opposed to the soluble form of M-CSF,when transduced into T9 (also known as 9L) glioma cellscaused the transduced cells to be immediately rejected [38–40]. After glioma rejection occurred, tumor immunity wasconcurrently induced. These rejected mM-CSF positive cells,not only lead to excellent prophylactic vaccination, but couldalso be successfully combined with antiangiogenic therapyto therapeutically treat seven-day established intracranialgliomas [41]. Granulocyte-macrophage colony stimulatingfactor (GM-CSF) and interleukin-4 (IL-4) transduced 9Lgliomas also led to tumor immunization under similarconditions [42, 43]. Since GBM patients relapse so fast, itwas considered unlikely, that one could establish the patient’sprimary glioma cell line and then transduce them withimmunostimulatory molecules or cytokines fast enough,before the glioma relapse occurs. Thus this genetic approachusing autologous gliomas has not been used for neuro-oncology.

One limitation of using GM-CSF transduced gliomacells as a tumor vaccine is that some human gliomas makeand use GM-CSF as a potential autocrine growth factor[44, 45]. So this cytokine must be carefully used, so as notto enhance the growth of the primary glioma. One way toavoid any possible GM-CSF-dependent autocrine pathwaysby gliomas themselves is to make use of the versatility ofthe APC such as dendritic cells (DCs) in an ex vivo settingusing the recombinant cytokine (see below). This way thisrecombinant cytokine does not directly interact with theglioma, while still mediating its therapeutic beneficial effects.


2.2. Antigenic Specific Pathways

2.2.1. Antibody Approaches towards Glioma’s Vasculature.Angiogenesis is crucial for tumor growth greater than 1-2 mm3. Multiple growth factors and proteolytic enzymesplay different roles in angiogenesis, but which pathway isthe most critical one for any given tumor is still activelydebated. Mostly likely, several antiangiogenic agents areneeded to target multiple sites, simultaneously to shut downthis entire process. Since gliomas are highly vascularized,these antiangiogenic approaches have a potential to work,assuming the right glioma and its angiogenic pathway canbe selected. Based upon microarray data analysis [46, 47],gliomas are classified into at least three subtypes: classical,proneural, or mesenchymal. Each form has its own uniquecharacteristics and survival rates. This data may allow forbetter targeting of these types of glioma, once these angio-genic pathways are identified for that individual glioma.We will discuss only the antibodies that are currently beingtried against brain cancer angiogenesis. A number of small,cell-permeable, receptor tyrosine kinase inhibitors are beingused clinically against brain tumor angiogenesis, but we willnot discuss them here, since they are not immunologicallybased. For further references see the current reviews by[48–50]. There is evidence that some antiangiogenic drugscan be successfully combined with a tumor vaccine to treatan experimental one-week established intracranial glioma[41].

Antivascular Endothelial Growth Factor (VEGF) Pathways.VEGF was the first cytokine to be associated with tumorangiogenesis. There are four forms of VEGF: VEGF-A,VEGF-B, VEGF-C, and VEGF-D. These growth factors bindto two specific VEGF receptors, types 1 and 2. The Avastin(bevacizumab) antibody binds and neutralizes the VEGF andprevents the cytokine from properly stimulating the VEGFreceptors. Theoretical interfering with this pathway shouldprevent the endothelial cell precursors from being recruitedinto the growing tumor, thereby blocking early angiogenesis.There is some association with VEGF-driven pathwayswith the mesenchymal type of GBM. Avastin along withdifferent chemotherapeutics did give high response ratesin recurrent gliomas ranging from 43–63% [51–53]. Thiscombined approach using an antiangiogenic antibody withchemotherapeutics could improve the efficacy of treatment,even if Avastin alone has some antiglioma effect [54]. Asa consequence, Avastin has been registered by the FDA inMay 2009 for treatment of relapsing GBM after standardtreatment.

Another antibody is ramucirumab; this antibody targetsthe VEGF Receptor 2 [55]. This antibody (IMC-1121B) isbeing developed by Imclone Systems. In Phase I studies usinga variety of solid tumors (no glioma patients were tested),systemic serum VEGF-A levels did rise during the therapy.Tumor perfusion and vascularity were diminished in mostof the patients that received the ramucirumab as predicted[55]. For GBM therapy, it would seem that both antibodiestowards VEGF and VEGFR2 could be used together tocompletely inhibit this VEGF-mediated pathway.

Antihepatocyte Growth Factor/Scatter Factor (HGF/SF). Asits name implies, this cytokine/growth factor was initiallydiscovered in liver cancers. But gliomas make this proteinand use it as an autocrine cytokine by binding to its receptorcalled c-Met [56]. Upon binding its receptor, HGF/SF isthought to stimulate the invasive behavior of the gliomas.Amgen has developed the AMG102 antibody, which hasin vivo efficacy against human U87 gliomas growing inimmunodeficient mice [57]. We are not aware of any currentclinical trials being performed for neuro-oncology with thisantibody.

2.2.2. Antibodies Directed towards the Glioma

Antiepidermal Growth Factor Receptor (EGFR) Antibodies.EGFR is a predominant pathway that helps characterize theclassical/proliferative type of gliomas. These receptors bindeither to EGF or transforming growth factor-α (TGF-α).These receptors can either be mutated or overexpressed dueto genetic amplifications. The most common mutation ofthe EGFR is the EGFRvIII mutation, caused by a deletion of268 amino acids in the extracellular region that constitutivelyactivates this receptor. At least 2 different antibodies towardsthis receptor are currently used for clinical studies: Cetux-imab/Erbitux [58] and Nimotuzumab [59]. A preclinicalmodel showed promise in immunodeficient mice [59], butthis success was not observed in clinical trials [60].

Antiplatelet Derived Growth Factor Receptor α (PDGFRα)Antibody. The PDGFRα-mediated pathway is representativeof the proneural subclass of GBM. This receptor can bindeither to PDGF-AA, PDGF-AB, PDGF-BB, or PDGF-CC[61]. Again, PDGFR is amplified and overexpressed in someGBM. The IMC3G3 antibody is being explored as a clinicaltherapy for this receptor. Cytomegalovirus (CMV) has beenlinked with human GBM [62] (see viral antigens). CMVis reported to use the PDGFRα as an attachment factor[63]. This antibody may be clinically significant, because itinhibits growth factor stimulation of the glioma along withinterfering with CMV infections. One other advantage oftargeting the PDGF receptor is that it can also target thepericytes/fibroblasts which give structural support to theendothelial cells [64]. This antibody may also interfere withthe antiangiogenic pathway, too.

Tenascin C. Tenascin C is a glycoprotein specifically madeby gliomas. It is laid down as an extracellular matrix.Monoclonal antibodies towards tenascin-C (clone 81C6)[65] by Duke University or the BC-2 and BC-4 clonesused at Bufalini Hospital (Cesena, Italy) [66, 67] have beendeveloped and are capable of localizing to various GBM inpatients. Attempts have been made to use radioisotope (I131,Y90 or At211)-conjugated antibodies to treat gliomas. Whenthese antibodies were injected into their respective patients,these antibodies localized to the glioma. In theory, theradiation released from the isotope-labeled antibody shoulddamage and kill the adjacent cancer. To date some successes(stabilized disease) are seen in their respective American and


Italian cohorts using these radiolabeled antibodies [65, 66].These antibodies will probably be quite useful for findingresidual pockets of the tumors. Since the tenascin c is notstrictly a glioma membrane protein, this may be a limitingfactor for the direct glioma treatment, and is probably not theoptimal way to treat glioma cells. Another possible problemwith this overall approach is that GBM stem cells are resistantto the effects of radiation (see GBM stem cells).

Bispecific Antibody. In the 1980’s the concept of using bispe-cific antibodies came into vogue. Here two different mono-clonal antibodies are used, the first antibody binds specifi-cally to the cancer while the second antibody binds the T orNK cells (via CD3). The antibodies are selectively reduced,so that a heavy and light chains still remain together as aheterodimer, maintaining the antibody binding specificity.Then the single heavy/light chain from the first antibody ismixed together with an identically prepared second antibody,which binds to the cancer cell’s surface. Afterwards, the twoheterodimers are allowed to reform their disulfide bonds,producing a stable antibody now with two different antigen-binding specificities. The bispecific hybrid antibody isselected, which simultaneously binds to both cells/antigens.Thus using this bispecific antibody, the effector lymphocytenow physically binds to the cancer cells and helps initiatescytolytic function by the lymphocyte against the cancer.

Nitta et al. [68] used a bispecific antibody towards CD3and a tumor antigen, originally developed against lungcancer (NE150, 69). But NE150 also had cross-reactiveproperties towards human gliomas. When they used thisbispecific antibody with LAK cells, this group achieved betterclinical responses against gliomas (four out of ten patientsshowed tumor regression within 10–18 months), while alleight patients treated with LAK cells alone showed tumorrecurrence. This study using bispecific antibody seemedsuccessful, but it was very labor intensive in the biochemicalpreparation. It was deemed impractical to generate sufficientantibody to treat multiple patients in order to demonstratestatistical significant improvement in a larger study. But theproof of concept here was established with this study.

Single-Chain Variable Fragmented Antibodies. Antibodieshave a molecular weight of 150,000 kd, so their ability to pen-etrate deeply into tumors or tissues is thereby limited. Withthe advent of genetic engineering, one can take hybridomacells and isolate the mRNA for the antibody. The variablebinding regions of the N terminals (first domains) of theheavy and light chains can be genetically cloned and ligatedtogether to maintain their ability to bind to the antigen. Sothe term, single-chain variable antibody fragment (scFv), wascoined. These recombinant molecules are now only about25 kd in size, which in theory should be able to penetratebetween cells better than normal antibodies. Early studieswere used against the EGFRvIII protein [69, 70]. Improvedtumor penetrance by this scFv was noted. Recently, this tech-nique has been used with a phage-display technology to pro-duce unlimited amounts of this recombinant protein. Lastyear, Kuan et al. [71] used this technique to make scFv that

target the multidrug resistance protein-3 (MRP3) gene foundon gliomas. These recombinant proteins had very good bind-ing affinities for gliomas and could be able to be conjugatedwith either drugs or radioisotopes. Since this MRP3 specificscFV targets a key biological response (reverse chemothera-peutic drug transporter), this scFV should be combined withchemotherapy to generate synergistic clinical effects.

Antibodies on the Horizon. Two antibodies (ipilimumab anddaclizumab) are on the horizon, which could have potentialimpact on glioma immunotherapy. Ipilimumab is the anti-body that targets an immunomodulatory molecule, calledCTLA-4. When naıve T cells become activated, a late antigencalled CTLA-4 is expressed. CTLA-4 then binds and inhibitsthe CD28 costimulatory pathway. Thus, this molecule natu-rally represses T cells. By preventing this CTLA-4-mediateddownregulation, an enhanced immune response can besustained and can probably enhance antitumor immuneresponses. Recently, this antibody has been successfully usedfor the treatment of melanoma [72]. Here an additional, fourmonths of survival were noted in these patients.

The second antibody is Daclizumab, which targets thehigh-affinity interleukin-2 receptor-α on T cells. This isanother potential monoclonal antibody that can improvepatient survival by preventing the actions of T-regulatorycells. T-regulatory cells (Treg) are IL-2Rα+ (CD25) cellsand thus more sensitive to the antibody compared to thecytotoxic T cells (see below). By eliminating Treg cells, amore sustained antitumor immune response can also bemaintained. Here the idea is to eliminate the Treg beforethey inhibit the an optimal antiglioma immune response. Inexperimental models, eliminating Tregs improves therapeu-tic efficacy of immunotherapy [73]. It will not be long beforeeither of these two antibodies will be combined with someclinical trial to improve glioma therapy.

IgE? An unexpected discovery was initially reported byWrensch and colleagues [74, 75]. Here atopic patients whofrequently suffer from immediate hypersensitivity reactions:hay fever, asthma have a lowered risk of contracting gliomas.Those patients who have high serum levels of IgE and whodo develop glioma, statistically survive somewhat longerthan those patients with low IgE levels. These studieshave been reproduced in a larger meta-analysis and seemhighly credible [76]. IgE is the antibody that mediatesimmediate hyper sensitivities. Nothing is known about howthe degranulating basophils and mast cells responding toIgE-mediated cross-linking affect glioma cells or the glioma’svasculature. So this phenomenon could prove to have majorrepercussions for future glioma therapy. Genetic engineeringwith some of the antibodies described above could beconstructed using the IgE framework. Most IgG antibodieswork therapeutically when applied in the milli to microgramrange, while immunopharmacological effects of IgE occursin the nano to picogram range. Thus, these redirectedantibodies might have unique properties in achieving clinicaleffects at lower doses than the IgG-based antibodies. Ofcourse, these proposed studies are very highly speculative


and require stringent animal safety tests to assure thatanimals and then patients do not immediately go in acuteanaphylactic shock upon contact with gliomas. Nevertheless,this is a very intriguing concept.

2.2.3. Cellular Approaches. After the nonspecific LAK cellexperience in the mid to late 1980’s, the next progression ofcellular immunotherapy was to use those tumor infiltratinglymphocytes (TIL) and the effector T cells that are foundin the local lymph nodes draining the tumor. These T cellswere already primed in vivo towards the patient’s own tumorcells. This methodology was developed in the days prior toour current understanding and routine use of dendritic cells.Here TILs were selectively expanded from either the tumoror draining lymph node cells, by using IL-2, supplementedwith LAK-conditioned supernatant as a source of other T-cell immunostimulatory cytokines [77]. By routinely restim-ulating these cells with irradiated or killed tumor cells, thishelped maintain T-cell specificity. When reinfused back intothe patients, the CD8+ CTLs had the inherent advantageover the LAK cells, in that these CTLs were capable ofkilling multiple target cells. CTLs only release their cytokineswhen properly stimulated by the self-MHC and peptide.These T cells reduced much of the clinical toxicity previouslyseen with LAK cells. CD8+ CTL were often the cell ofchoice to examine since their effector function (cytolysis) waseasily measured by radioisotope release assays. This TIL/CTLapproach again proved to be somewhat better at generatingclinical responses to melanoma and renal cancer [77].

In rodent models, the use of TILs and draining lymph-node-derived T cells expanded ex vivo did prove to be effica-cious for the treatment of rodent gliomas [78–80]. Dunn etal. [81] have recently reviewed the history of clinical glioma-based T cells in better details; readers are encouraged to readthis article. GBM TILs were successfully used by Quattrocchiet al. [82] to treat their patients, where they took TILs derivedfrom recurrent malignant gliomas and expanded the CD3+T cell in vitro with IL-2. Both CD4+ and CD8+ cells wereexpanded and then reinfused into the surgical cavity viaan Ommaya reservoir. After the infusion of the T cells, thepatient was given IL-2 maintenance therapy, three times aweek for a month. There was one complete responder (45months) out of 6 patients treated. Plautz et al. [83] alsoshowed some clinical successes using immunized T cellsobtained from inguinal lymph nodes and expanding themwith a mitogen [83]. After a short-term ex vivo expansionthese cells were then reinfused back into the patient.

Allogeneic Mixed Lymphocyte Reactive T Cells. Anotherlymphocyte approach towards brain cancers was pioneeredby Kruse et al. [84] and reviewed in Yang et al. [85].Here lymphocytes derived from histoincompatible allogeneichuman blood donors are combined with the patient’s irra-diated lymphocytes. A mixed lymphocyte reaction sensitizesthe allogeneic donor’s peripheral blood mononuclear cellstowards the patient’s MHC. These alloactivated T cells arethen expanded in low doses of IL-2 for another 2-3 weeks.These alloactivated lymphocytes are cytotoxic towards the

patient’s lymphoblasts. When these effector cells are thenimplanted into the resection cavity, these CTLs can eliminatethe remaining glioma cells. This procedure was repeated upto 5 times for each patient. Some long-term (>15 yr) sur-vivors were documented against stage III gliomas [86]. Earlyconcerns that these allospecific CTLs would indiscriminatelykill nontumorous host brain cells and induce autoimmunityhave been proven to be unfounded. Thus, a larger dose-escalation study using this technique for stage III astrocy-tomas is open for accrual for 15 patients in the southern Cal-ifornia area in collaboration with Dr. Linda Liau (UCLA) toexpand and confirm the validity of this therapeutic modality.

T-Helper Cells. CD4+ T cells also have important antitumorimmune effector functions. CD4+ cells recognize MHC classII restricted peptides. Some CD4+ T cells can kill tumor cellseither via Fas Ligand-dependent [87] or perforin-dependentpathways [88]. But most tumors do not express MHC class IIantigens, so how antitumor effects are directly mediated bythese T cells is not really known. The CD4+ cells’ probablemechanism of action involves the release of cytokines andother mediators, which either targets the tumor directly orthe tumor’s vasculature. The best possibility is that as a resultof stimulation by the DC, these CD4+ cells release cytokines(IL-2, IL-6, IFN-γ, TNF, lymphotoxin (LT)) that assist in theexpansion of the CD8+ CTLs. Some cytokines released fromtype 2 helper T cells (Th2) can assist in B-cell activation andmaturation into making specific IgG antibodies. In the rat9L (also known as T9) glioma model, effective immunitywas seen by the adoptive transfer of immunized CD4+T cells [39, 89]. Furthermore, Okada et al. [90] showedthat rats immunized with IL-4-transduced 9L gliomas didmake antibodies against at least three rat glioma-associatedproteins, not previously known to be glioma-associated anti-gens. In a humanized SCID mouse model, CD4+ T cells wereisolated from a mouse that was actively rejecting a membraneisoform of macrophage colony stimulating factor (mM-CSF)transduced U251 glioma [91]. But no exact mechanismswere provided in these last studies, explaining how the directbeneficial role of CD4+ T cells occurred in these gliomamodels other than by the “classic” T-helper cell function.

T-Regulatory Cells. Gliomas frequently contain T-regulatory(Treg) cells [92–94] which are CD4+, CD25+ (IL-2Rα+),and FoxP3+ cells. These cells are probably induced in aneffort to maintain immune homeostasis. In the gut, thesetypes of cells also can be induced into becoming follicularhelper T cells that assist B cells into making IgA [95].Because of microenvironmental conditions in the gliomasuch TGF-β and PGE2, these T cells are forced to becomethese suppressor types of cells. These Treg cells inhibit T-cell effector functions; this might account for the failure ofGBM-derived TILs to successfully eliminate the glioma inclinical trials. Tregs work in several ways [92–94] to inhibitthe necessary antitumor effector mechanism. Methods toeliminate Treg function will likely improve clinical resultsin future trials. Currently antibodies against the IL-2Rα(Daclizumab) are used to eliminate these Treg cells. The


earnest investigation of Treg in cancer has blossomed in thelast five to seven years. So it is unknown what percentageof Tregs was previously expanded as TIL populations andinadvertedly used in previous clinical TIL studies, whichmost likely failed to treat these patients.

T-reg cells and another type of T-helper cell, called Th17cells, seem to share a common early stage pathway [96, 97].Naıve T cells upon exposure to antigen and TGF-β can giverise to either Treg or Th17 cells. To get Th17 cells the presenceof IL-6 is required. Both cytokines are produced by gliomas.Recently, Th17 cells were described in murine and humangliomas [98], but their beneficial or inhibitory actions wasnot elucidated. In a mouse model of melanoma, Th17 cellscould be used to eliminate very large established tumors [99].

Redirected T Cells Using Chimeric Antigen Receptors (CARs).Generating T-cell clones responding towards tumor-specificantigens either by TILs or by DC restimulated T-cells clonesis quite labor-intensive and naturally quite costly. Thepotential for microbial contaminations and incubator/powerfailures increases with time. This logistical problem lead tothe concept of redirecting T cells or NK cells by geneticallymanipulating these effector lymphocytes by using man-made chimeric antigen receptors (CARs). The same basictechnology described for scFv antibodies can now bemarried to lymphoid effector cells. Here one splices the scFvregion via a spacer region to the transmembrane spanningregions of the CD28 molecule. The intracellular region ofTCRζ chains is also ligated into this construct. This artificialreceptor, when activated upon proper T-cell surface-bindingto the tumor, initiates cytolytic T cell function (release ofperforin/granzymes or cytokines). These kinds of geneticallyengineered receptors have also been called zetakines orT-bodies.

These redirected T cells can recognize a tumor’s cell-surface molecule. Other advantages of using CAR constructsare (1) independent of HLA expression where HLA isfrequently down regulated or eliminated on the gliomas; (2)can react better to modestly expressed tumor targets; (3)CARs have uniformity and high-degree of expression; and(4) reliably generate T cells in a relatively short time forclinical usage (10–15 days as opposed to 10–12 weeks neededfor CTLs).

Chimeric antigen redirected (CAR) T cells have beendeveloped, so that they can bind to the IL13Rα2 [100], her2receptor [101], EGFRvIII [102] or to the ganglioside GD3[103]. Figure 1 shows the current forms of CAR currentlyused for possible therapy of human gliomas. After the CARconstruct is engineered, then the gene is packaged withineither adenoviral or retroviral vectors. This allows oneto quickly transfect as many patient’s T lymphocytes aspossible. When these transduced peripheral blood T cellsare reintroduced back into the patient, preferably near or inthe cancer inside, the CAR-redirected T cells can attack thecancer. IL-13Rα2 based CAR/zetakine transduced T cells killseveral human glioma cells in vitro and appear effective inintracranial gliomas in immunodeficient mice. With her2specific CAR lymphocytes, therapeutic efficacy was seen in a

xenogeneic model with the Daoy (her2+) medulloblastoma,when these human CAR T cells were adoptively transferredinto these mice [104]. Her2 CAR constructed human T cellskilled both CD133+ and CD133- GBM cells. Her2-redirectedCAR T cells showed some efficacy against the human her2+gliomas growing in SCID mice [101]. The advantage of usingCAR-T cells is that they are also applicable towards othercancers like her2+ breast and ovarian cancers; EGFRvIII+engineered CAR T cells can also target lung cancers;while ganglioside GD3 CAR-T cells can also interact withmelanoma cells. Currently CAR T cells are being usedclinically at the City of Hope (Duarte, Ca) and the clinicaltrial using the CAR-her2 cells is expected to start recruitingpatients in the mid-late Fall of 2010 at the Center for Cell andGene Therapy of Baylor College of Medicine (Houston, TX).

Dendritic Cell-Based Vaccines. Dendritic cells (DC) are cur-rently the favorite therapeutic modality now used in cancerimmunotherapy. Monocytoid dendritic cells are readilyavailable from the peripheral blood monocytes and can bequickly activated ex vivo using cytokines as granulocyte-macrophage colony stimulating factor (GM-CSF) and inter-leukin 4 (IL-4). This technique is quite versatile in thatdifferent sources of antigens can be added to the DC. Killedtumor cells, tumor cell extracts, purified tumor mRNA, orpurified tumor antigens can be given to the DC and thesecells can then properly process the tumor antigens, so thatthe peptides are presented in the MHC. These DC are capableof immunizing naıve animals [105–107]. Plasmacytoid DCcells are beginning to be used for cancer immunotherapy[108], but so far, they have not been developed for braincancer therapy. Once ex vivo activated and antigen-pulsedDC are generated, some protocols allow the non-maturedDC or matured DC to be injected back into their patients.The non-matured DC are thought to become mature afterreintroduction back into the patient, especially after theinjection of the TLR antagonists, which act as an adjuvantto cause DC maturation.

Worldwide there are multiple centers [109–114] generat-ing DC used for brain cancers. Clinically positive responsesfor usually reported for a subset of glioma patients. Usuallythis means that the mean time to progression for thesetreated patients with high grade gliomas increased in theseresponder populations. Kim and Liau [115] reported thattheir vaccine responders survived 642 ± 61 days whencompared to the non-responders (430 ± 50 days). Diseasefree progression was also improved by 4.5 months. Somepatients are reported to survive more than five years. Onekey finding that has been repeatedly reported is that this DC-based immunotherapy is safe with few serious side effects.

A variant of the dendritic cell-based vaccine occurs byfusing the DC with the glioma cells to form an immunos-timulatory cellular hybridoma. Here the DCs are fused withthe autologous glioma cell line. This strategy is analogousto the classical hybridoma used for antibody production,except the end function of this hybrid is to stimulate animmune response. The glioma parent cell supplies the correcttumor antigens, which should provoke the proper host


specific immune response. The DC parental cell providesthe machinery to take the glioma antigens, and process thepeptides onto the patient’s MHC (HLA-A, B, C, and D loci).The DC parent also supplies the costimulatory molecules andcytokines to make this hybrid an immunostimulatory cell.When these x-irradiated hybrid cells are injected back intothe patient, they now provoke an immune response, so thatmultiple T clones responding to multiple glioma antigens areelicited. In clinical studies, Kikuchi and colleagues used thisapproach to treat 6 patients [116]. This first study showedthis hybridoma vaccine was safe, but it failed to achieve anyclinical effect. In the follow-up study, the DC hybridomawas combined with an injection with recombinant IL-12, 3,and 7 days after the hybrids were injected. This combinationachieved some disease stabilization and tumor shrinkage infour out of 15 patients that were treated [117]. After one year,two of the patients still survived.

Finally, one aspect of DC that is not now fully appreciatedis that IL-4/GM-CSF activated human DC can kill gliomacells [118]. Both human and rodent DC can directly killhuman gliomas either by direct contact or by the release ofnitric oxide. Others have previously reported this cytolyticphenomenon towards other human cancers [119, 120] bythe release of type I or type II interferon and membranecytolytic-inducing molecules (i.e., TRAIL, NKG2D, FasLigand). No one is currently using activated DC as theactual cytolytic effector cells against glioma. Most researchersconsider the best use of DC is to stimulate T-effector cells bya systemic vaccination route.

3. Tumor Antigens

Identification of clinically relevant tumor antigens is activelyresearched. New tumor antigens seem to be reportedmonthly. Tumor antigens are identified by either antibodiesor by T cells. The latter are recognized by T cells in the contextof the TCR with either MHC class I which are recognizedwith the help of CD8+ molecules or MHC class II thatare recognized by CD4+ molecules. Cheever and colleagues[121] have recently prioritized a set of 75 human tumor anti-gens, which they determined should be further developedfor cancer immunotherapy. This prioritization was basedon a number of key factors, such as possible therapeuticfunction, the immunogenicity of these molecules, their rolesin oncogenesis, its specificity, and its frequency in a numberof cancers. At least 18 of these listed tumor antigens arepertinent to human brain cancers.

Tumor antigens can be defined as either being tumor-specific or tumor associated. Tumor-specific antigens areactually rare, while tumor-associated antigens are expressedon normal tissue and are simply overexpressed by the tumor.Table 1 lists the antigens that can be considered glioma-associated antigens. Surprisingly, there are really no trulyglioma-specific antigens currently known. All these antigenscan be found within a variety of other tumor types. Butfrom experimental evidence, we know there is tumor specificimmunity. So there are undoubtedly many glioma-specificand glioma-associated antigens still to be found.

3.1. Tumor-Specific Antigens. Tumor-specific antigens in-clude p53, EGFRvIII and ras mutations. These antigensare quite common in many types of cancers. Gliomasrarely have ras mutations, but frequently possess p53 pointmutations, which inactivate its normal function. Mutatedp53 can be recognized by murine CTLs by wild type-p53peptides that bind to MHC class I alleles [123]. HumanCTLs responses towards p53 can also be developed in anidentical fashion [124, 125]. Many cancer patients possessdiscernable antibody responses to p53 [126], so someTh2-mediated responses generated towards MHC class II-restricted antigens are needed to help produce these higheraffinity IgG antibodies. Many glioma antigens are called“antigen recognized by T cells” (ART), for example, ART-1, ART-4, or “squamous antigen recognized by T cells”(SART); that is, SART-1, -2, and -3. These last 5 antigens wereidentified within either glioma cell lines or within adult orpediatric brain tumors [127].

3.2. Overexpressed Antigens. These antigens are found innormal cells and tissue, such as B-cyclin and CD133. Theseantigens appear to be overexpressed on their cancerous coun-terparts. Some antigens are found only within the testes andcancers. Hence the term “cancer-testes antigen” is frequentlygiven, to describe them. Some of these antigens includeMage-1, Gage-1, SSX2, and NY-Eso-I. These antigens arefound in terminally differentiated melanocytes and in theirtransformed progeny (melanomas) and in gliomas. Somedifferentiation antigens are not found in the testes, but arefound on normal cells like melanocytes and in melanomasand in gliomas, with Trp-1, and Trp-2 being representativeof this group. Since melanomas and glioma cells share a com-mon embryonic neuroectoderm precursor, it is not that sur-prising that these two cancers share many common antigens.

3.3. Viral Antigens. Many viruses are thought to play acausative role in some human cancers: HTLVI, hepatitis Band C virus, EBV, and papilloma virus. Recently, Cobbsand his colleagues [62] linked cytomegalovirus (CMV)with human gliomas. CMV is frequently detected withinchronically immunosuppressed patients with either trans-plant patients or in late stage HIV infections sufferingfrom Acquired Immune Deficiency Syndrome (AIDS). Itis thought that 70–90% of the population are previouslyexposed to CMV and might be chronically infected withthis virus. Our immune systems keep this virus under tightcontrol. Glioma patients are frequently considered immuno-suppressed by a number of mechanisms [128]. So whenthe immune system is impaired as in GBM, the CMV cannow reappear. Whether CMV directly causes glioma is acontroversial topic. The possibility that CMV attaches itselfto glioma via the PDGFRα allows some interesting therapiesto be explored. Viruses are usually good targets for theimmune system. One CMV antigen, pp65, induced humanHLA-A2 immune responses in a GBM patient [129]. Freshlyisolated glioma samples seem to express this antigen to ahigh degree [130], but cell lines lose expression of CMV. Ifa high number of GBM cells do attract and harbor CMV in


her2 EGFRvIII Ganglioside GD3

TCRζ

Spacer

Antibody

T-cell

Interleukin 13Rα2

CD28

TCRζ

Spacer

CD28

TCRζ

Spacer

CD28

TCRζ

Spacer

CD28

scFv(single-chain

variable fragment)

CCDDD22888CCCDD22888CCCCCDDD288888CCCCCCCDD222288888

Figure 1: Chimeric Antigen Receptors (CAR) used for potential human therapies of brain cancers. The monoclonal antibody towards eitherher2/neu, EGFRvIII, ganglioside GD3, or IL13Rα2 is the initial source of the genetic material. The first domains of the heavy and light chainsare ligated together with a short spacer region to create the single chain variable fragment (scFv), to preserve the recombinant proteins’antigen binding region. Another spacer region is ligated from the scFV region to the transmembrane CD28 molecule, followed by the TCRζchain. After the T cells are transfected with the adenoviral construct, these T cells are then allowed to interact with the tumors. Upon contactwith the antigen on the tumor, the CAR is activated and the TCRζ chain is now activated which then stimulates antitumor mediator effectorfunction, that is, cytolysin or cytokine release.

Table 1: List of tumor-associated antigens known within human brain cancers.

Aim-2 Art-1 Art-4 B-cyclin CD133 EGFRvIII

Epha2 Ezh2 Fosl1 (fra-1) Gage-1 Galt-3 Ganglioside GD3

Gp100 GnT-V Her2 HNRPL IL-13Rα2 Livin

Mage-A1 Mart-1 MELK MRP-3 NY-Eso-1 Prame

PTH-rP Sart-1 Sart-2 Sart-3 Sox 2 Sox10

Sox 11 SSX-2 Survivin Tert TRP-1 TRP-2

Tyrosinase Ube2V Whsc2 WT-1 YKL-40

SLC01C1∗ BCAN∗ CHI3LI∗ CLIP2∗ FABP7∗ NR2E1∗

NLGN4X∗ NES∗ NRCAM∗ PDPN∗

Asterisk denotes potential tumor antigens described in [122].

vivo, then this opens up the possibility of developing CMVpeptides to vaccinate against the virus and therefore theglioma is targeted indirectly. Currently, Duke University isactively using DC-based vaccines targeting CMV antigens totreat their patients who are CMV positive. Clinicians at PennState University are using allogeneic CMV-specific CTLs totreat their glioma patients [131, 132]. The Center for Celland Gene Therapy of Baylor College of Medicine is alsodeveloping a CTL approach against CMV as the way to treatGBM, these clinical trials will shortly begin in the next fewmonths. Finally, another possibility is to use the CMV virusas a vector to deliver therapeutic agents or genes specificallyinto these gliomas to make them more vulnerable to assortedtherapies. Thus, CMV may be a very useful immunologicaltool to attack GBM.

4. Source of Antigenic Materials

The choice of a source of the tumor antigen is probablythe most important decision to be made when it comes tovaccinating cancer patients. There are multiple sources ofantigenic material: cell lines (whole cells or lysates), freshsurgical tissue (cell lysate, mRNA, or primary cultures oftumor “stem cells” grown as neurospheres), and peptides(acid eluted or synthetic). Each choice has its own pros andcons for their clinical usefulness. The knowledge gained fromthese clinical studies using cell lines, neurospheres, surgicalspecimens, and peptides will undoubtedly advance the field,once we determine what the best source of tumor antigens is.Unfortunately, only trial and error will tell us the best sourceof tumor antigens for clinical responses.


4.1. Cell Lines. Traditionally, cell lines were used for exam-ining immunization properties either in animals or in earlyclinical trials. These cells generally represent a stable andcontinuous source of tumor cells and they can reproduciblyform cancers in experimental animals. Additionally, thesecells can be modified with cytokines or costimulatorymolecules to improve their immunogenicity. Cell linescreated from the 1960 to 1980’s are still widely used:Uppsala Sweden-derived cells (U87, U251, U373, etc.), DukeUniversity-derived (D54, D68, etc.), Lucerne Switzerland-derived (LN18, LN 229, etc.), Surgical Neurology Branch(Bethesda, MD)-derived cells (SNB19) are still quite usefulfor studying various aspects of glioma biology both in vitroand in vivo within immunodeficient mice. These glioma cellsreadily form either intracranial or subcutaneous tumors inthese immunocompromised mice. SCID/NOD mice can be“humanized” by prior transplantation of human thymus andbone marrow. One can examine human immune responsesin vivo without doing expensive clinical trials, while quicklyexploring the feasibility of generating these human-specificresponses.

Established cell lines can either be used as whole-cells oras a lysate to be the immunogen. Cell lysates can be combinedwith adjuvants or it could be used as an irradiated wholecell. Cell lines can be genetically modified with cytokines orcostimulatory molecules to improve the immunogenicity ofthe cells. Parney and colleagues [133, 134] have shown thispossibility of genetically engineered glioma vaccines withinimmunodeficient mice.

Zhang and colleagues [135], examined 20 human GBMcell lines, some well-established cells such as U87, U251,D54, LN18, SNB19, and so forth, along with some recentlyderived cells (NovaRx: NR203, 206, etc.). These glioma celllines were characterized for their tumor antigen expressionof 20 tumor-associated antigens by quantitative reverse-transcriptase real-time polymerase chain reactions (qRT-PCR) techniques. The translated proteins were confirmed byimmunoflourescent antibody staining and intracellular flowcytometry for 16 tumor antigens, since those antibodies wereavailable. With the exception of 3-4 antigens, the cell lineswere all quite homogenous and had high tumor-associatedantigen mRNA expression. Hence a cell line or combinationof cell lines can easily be used as a universal source ofantigenic material for any potential vaccine. Besides thecurrently known tumor-associated antigens (Table 1), theother advantage is that cell lines can also be a source ofcurrently undiscovered antigens. Zhang et al. [127] showedthat surgical specimens derived from adults with GBM werehighly antigenic (29 antigens were routinely expressed),while the surgical specimens derived from pediatric GBM,ependymoma, pilocytic astrocytomas were more restricted intheir tumor antigen profile (9–16 antigens).

Cell lines grow quite easy and can be produced in bulkin an economic manner. To use them as a clinical product,one needs to do extensive clinical testing for all types ofpathogens (mycoplasma, bacteria, fungi and viruses). Somesafety testing takes the cells/cell lysates and exposes themto other human/animal cells, injected into suckling mice,guinea pigs, and chicken eggs to assure no cryptic pathogens

are present when grown in a more permissive environment.As expected this procedure is quite expensive ranging fromUS $50–80 K/cell line. This procedure is required, if thevaccine material from one person is being injected into adifferent patient. But if the autologous cells are custom-madefor that given patient, then this added safety test usually is notrequired for the exotic viral pathogens. Only the commonmicrobial contaminants need to be tested. This financialrestraint helps explain why most current protocols are usingautologous gliomas. A cell lysate is made directly from theresected surgical specimen, so the risk is minimized forpossible microbial contamination. Currently two companies,NovaRx (San Diego, CA, USA) and Epitopoietic ResearchCorporation (ERC, Gembloux Isnes, Belgium) are develop-ing allogeneic-based vaccines. NovaRx is developing wholecell line vaccines by knocking down the glioma’s ability tomake TGF-β. Thus, shutting down the ability of these cellsthat lead to immunosuppression, and this leads to improvedin vivo responses [136]. This irradiated whole-cell-basedvaccine is known as Glionix. The Glionix has been safely usedto vaccinate six patients and the results were encouraging, sothat further Phase II/III studies are currently being initiated.ERC is developing a cell lysate approach by combining tumorlysates from a number of GBM specimens. This company iscurrently accumulating glioma tissue to be used for their celllysate-based vaccine.

An early success using cell lines as a source of tumorvaccines was reported in the early 1980’s when the DukeUniversity group used D54 and U251 cell lines to vaccinatehuman glioma patients [137]. Some vaccinated patientsbecame long-term survivors after being vaccinated with theseirradiated whole cells. The use of the killed U251 (HLA-A2+)cells was reported to be the most effective, when comparedto the D54 (HLA-A3+) vaccine. Unfortunately, this studyproved to be premature and illustrated the “growing pains”of the field. It is now felt that patient selection, such as usingpatients with lower grade gliomas, better prognosis status, oryounger patients with a better survival prognosis (proneuralGBM subtypes) were used with the U251-based vaccine.

The human glioma cells used for that clinical vaccinationwere cultured in fetal calf serum (FCS), which is routinelyused for tissue culture. When the vaccinated patient’ssera were tested, most patients had antibodies against thebovine proteins found in FCS [138]. Antibodies against theimmunizing glioma’s HLA antigens were additionally found.One patient had an undefined antibody specific for theU251 cells. This work clearly showed that humoral immunitywas induced within these patients by using an irradiatedwhole cell vaccine. Whether the FCS acted as a xenogeneicadjuvant or as a mechanism of “epitope spreading” [139] isan intriguing possibility.

4.2. Stem Cell Lines. The discovery of proper cell culturetechniques for growing “stem cells” or “cancer initiatingcells” as neurospheres has helped advance the field of gliomabiology. Here surgically removed tumor cells are dissociatedand grown in serum-free media containing epidermalgrowth factor and basic fibroblast growth factor [140].


The concept of “glioma stem cell” has emerged as a hottopic. Many of the features seen in clinical glioma arebetter reproduced when the CD133+ stem cells, ratherthan those coming from established cell lines are injectedinto immunodeficient mice. Only 10–100 of these cells areneeded to form tumors in these mice within 30–60 days.The resulting gliomas derived from CD133+ cells displayan invasive phenotype, as opposed to well-circumscribedborders that form when established (differentiated) celllines are used. GBM stem cells can also explain why gliomasresist certain drugs and radiation [141, 142]. CD133 wasinitially described as a marker which resisted the uptakeof fluorescent markers, which coincidentally resemblechemotherapeutic drugs [143]. However some GBM stemcells are also reported as being CD133-negative [144]. So thewhole concept of GBM stem cells is still being refined.

Glioma stem cell lines have been used as a vaccine inrodent models and appear safe and effective to vaccinateagainst some rodent gliomas [145]. Some human gliomastem cell lines have been characterized for their tumorantigen profile [146]. With the exception of CD133 for somestem cells, GBM stem cells possess few truly tumor-specificantigens. So it is unlikely we will be able to specificallyonly target stem cells by immunotherapy. Tumor associatedantigens such as TRP2, GP100, EGFR, AIM2, and Sox2 arepresent on these human glioma “stem” cells. In contrast, IL-13Rα2 and her2 seems to be diminished in these human stemcells.

4.3. Surgical Resected Lysates. Tumor lysates are most fre-quently used, clinically. After the patient recovers 6–8 weeksfrom debulking surgery, the patient is leukopheresed toacquire sufficient DC precursors. The debulked tumor wascollected, analyzed and aliquots are saved for vaccinatingprotocols with the DCs. This process is relatively simpleand straightforward with minimal risk of contamination, asopposed to long-term cell culturing.

When glioma surgical samples (WR-GBM) were injectedwith Freund’s adjuvant (complete or incomplete) intomonkeys and guinea pigs, an autoimmunity resemblingexperimental allergic encephalitis (EAE) was produced[147]. However, when the glioma cell line that was derivedfrom WR-GBM (D-68) was used as the vaccine withoutFreund’s adjuvant, no autoimmunity was seen. As a resultof these potential autoimmune complications, vaccinationusing glioma surgical samples was largely abandoned andfuture attempts were usually discouraged by citing thiscase. The results using DC pulsed with surgically resectedtissue have been shown to induce tumor immunity withoutany signs of EAE [109–115]. This empirical evidence alsoovercame the prejudice that was initially elicited from thatprior EAE induction paper [147].

So what are the best sources of antigenic materialcoming from cellular sources? Since cell lines and stemcell neurosphere cultures are pure tumor cells, it would beexpected that these sources would contain the most tumor-associated antigens, while the surgical specimens are a mixof tumor cells and normal hosts cells like: endothelial cells,neurons, microglial, and other hematopoietic cells that were

present within the tumor. These normal cells will thereforedilute out the tumor antigens coming from the tumor cells.This is potentially significant, because when one uses thismaterial to pulse dendritic cells, irrelevant normal hostantigens may be loaded onto the binding grooves of the MHCmolecules, hence stimulating the immune system to a lowerextent. Since there is a finite number of MHC moleculesper dendritic cell, loading irrelevant peptides may make theimmunostimulatory process to T cells less efficient.

The concept of vaccinating against the tumor’s vascu-lature was reported [148]. The late Judah Folkman [149]used to argue that the endothelial cells of tumors are normalcells. These normal endothelial cells would therefore notbe subject to the same mutational rates as cancer cells. Sothese cells would make the best target for cancer therapy,rather than directly targeting the cancer cells. Virrey et al.[150] have found that endothelial cells derived from gliomasare different than those endothelial cells derived from thenormal noncancerous brain, these cells have morphologicalchanges and grow slower than expected. Two recent studiessuggest that GBM “stem” cells can differentiate into gliomaendothelial cells [151, 152]. So there could be legitimatereasons for using the whole tumor lysate as a vaccine in orderto target these abnormal endothelial cells. May be some of thesuccess seen, when the tumor cell lysate was fed to the DCwas due to immune responses directed towards the glioma’svasculature. This open question is an important issue thatwill be needed to be answered before the next major advancetowards cancer immunotherapy is made by using cell-basedmaterials.

4.4. Peptides

4.4.1. Peptides Vaccines. A more refined approach is to usepossible antigenic peptides as the starting vaccine. Oneknows exactly how much antigen is given to the DC, asopposed to cells or tumor lysates. Liau and her coworkers[153, 154] and Yu and his colleagues [155] used this strategyto pulse their dendritic cells. Here tumors or tumor cell lineswere acid eluted and the extracted peptides were loaded ontothe patient’s DC. Some T cells stimulated with this methoddid generate T cells that infiltrated the recurrent glioma.Liau et al. [153] did use peptides derived from allogeneicglioma cell lines and then pulsed the patient’s DC with thesepeptides. Little evidence of beneficial antitumor immunitywas seen. This might explain why these DC researchersquickly switched to using the tumor cell lysates obtainedfrom surgical specimens as the starting vaccine material.

The Duke group is using synthetic peptides-based strate-gies to vaccinate EGFRvIII mutated GBM [156, 157]. TheUniversity of Pittsburgh (Pittsburgh, PA) is using gliomaassociated peptides (Survivin 96–104(2M), WT1 126–134(1Y), EphA2 883–891, and IL-13Rα2 345–353(1A9V))along with poly ICLC to vaccinate their patients [158]. Oneof the advantages of using synthetic peptides, is that onecan design peptides that have a higher binding affinities forthe MHC molecules than the actual tumor-derived peptides.This is evidenced by the use of redesigned survivin, WT-1and IL-13Rα2 peptides that the Pittsburgh group is using.


Currently, the NYU Pediatric Neuro-Oncology Group incollaboration with their Melanoma-DC program is treatingpatients with the autologous pediatric DC with severalpeptides (gp100, TRP2, EphA2 and her-2) for various pedi-atric brain cancers (low grade gliomas, GBM, ependymoma,and medulloblastoma). To date, they have vaccinated fourchildren without any signs of toxicity (Dr. Sharon Gardner,personal communication). Since children survive longerwith brain cancers that their adult counterparts (CBTRUS,2010), it will take some more time before the actuarialdata is collected before we know whether this therapyworked.

Of course, one restriction using specific peptides is thatthese patients have to be HLA-matched to assure that thepeptides will properly bind to the patient’s own MHC.Roughly, half of population in the USA have the HLA-A2allele, while Japan and China have higher proportions ofHLA-A24 allele, so most studies will need to be based ontheir correct HLA-alleles. It remains to be seen whetherbetter clinical responses will be generated with these peptide-based vaccines when compared to the cell lysate pulsed DC.One possibility is to use the entire tumor-associated antigenprecursor protein as opposed to small peptides to pulse theDC. Here the DC will process the proteins so that the correctpeptides will bind to its own unique MHC molecules. Sothe patient’s own DC will customize their proper antigenicpeptides to fit with their own immunogenetics.

4.5. Nongenetic Manipulated Vaccines. Cells can die by atleast three distinct pathways: apoptosis, autophagy, necro-sis/paraptosis. Bredesen et al. [159] and Hotchkiss et al.[160] have reviewed the biological processes involved in thesedifferent pathways possess. Each pathway has its own uniqueability to interact with the immune system. These types ofvaccines could be used either with cell lines or the stem cells.

4.5.1. Apoptosis. Radiation and most chemotherapeuticagents kill tumor cells by initiation of apoptosis [161, 162].In response to these treatments, many in situ tumors initiallyshrink and regress, only to reoccur with a more malignantphenotype sometime later. Despite a massive release oftumor-specific material, including tumor antigens, no lastingimmunity or tolerance occurs [163, 164]. Apoptosis hasbeen called the “silent death” and does not usually provokeimmunological responses. Most apoptotic cell remnants aretaken up by adjacent cells. Apoptosis is the driving pathwaythat induces immune tolerance towards many self-antigenseither in the thymus for T cells or in the bone marrowfor B cells. The only way to override this immunotolerizingproperty of apoptotic cells is use some immunostimulatorycytokines like either IL-4 or GM-CSF [165, 166] or bycostimulatory molecules which help stimulate the APCfunction [134].

4.5.2. Autophagy. This process occurs when cells are stressedor deprived of key nutrients. These affected cells undergoa process whereby they begin to self-digest themselves. Thekey morphological change in autophagic cells is having

double membraned vesicles within the cell. Like apoptoticcells, autophagic cells do not induce in vivo inflammatoryresponses. Autophagy may be a way that cells infected withintracellular pathogens commit suicide, thereby limiting theintracellular infection. Breast cancer cells treated with 4-hydroxytamoxifen can be stimulated to undergo autophagy.So autophagy may be useful for certain clinical therapies.However some breast cells survive this therapy and then showdrug resistance [167]. Under in vitro conditions, autophagictumor cells can be fed to dendritic cells and produce T-effector cells [168, 169].

4.5.3. Necrosis/Paraptosis. The mechanism of necrosis-induction is the least well-defined pathways of the threeforms of programmed cell death. It is thought that para-ptosis is the programmed pathway that leads to necrosis.Paraptotic cells are characterized by a swelling and vacuoliza-tion process that starts with the physical enlargement ofthe endoplasmic reticulum (ER) and mitochondria [170].Swollen cells suggested that ionic disregulation is accom-panied by water influx and retention. The disruption ofintracellular ion homeostasis ultimately causes these cellsto osmotically lyse releasing intracellular contents, such ashigh gel mobility binding protein-1 (HMGB1) [171], heatshock proteins [172], and various proteases. These proteinsact as “Danger Signals” promoting massive inflammationand cellular immunity [173]. Hence, the best way toimprove the natural immunogenicity of the tumor cell isby using necrotic/paraptotic cells. In contrast, vaccinationwith necrotic tumor cells produces superior T-cell immuneresponses in comparison to those responses elicited byimmunization with apoptotic cells [163, 164, 174]. Theadvantages of using the tumor’s own natural death pathwayallows for the autologous glioma cells to be used as a vaccinewithout having to do any genetic manipulations, which coulddelay the time that a vaccine could be given to the patientwith a relapsing glioma.

Over the last decade our lab pioneered the use ofgenetically engineered tumor cells with membrane M-CSF asa tumor vaccine (mM-CSF, 38–41). Activated macrophageskilled these mM-CSF transduced tumor cells quite easily inat least four different tumor models (rat glioma, humanglioma, mouse hepatoma [175], and rat breast cancer [176]).In our glioma models (rat T9 and human U251), the mM-CSF+ tumor cells are killed upon binding by the respondingmonocytes followed by the release of reactive oxygen species(ROS). The interaction with ROS resulted in paraptosis[49, 177–179].

The limitation of using mM-CSF-based glioma vaccineswas that this tumor vaccine required living cells in order toproduce the immunogenic stimuli needed for the vaccine.If the mM-CSF transduced T9 cells were either x-irradiated,mitomycin-C or freeze-thawed prior to subcutaneous injec-tion, the vaccinating effects of mM-CSF+ glioma cells wouldbe lost [40]. Hence no IRB or study section would permita living tumor cell vaccine to be used in human patients.So the mechanism by which these paraptotic cells wereinduced, was investigated. This would then reproduce the


same immunogenicity of mM-CSF+ cells but now withkilled glioma cells. By forcing open BK (big potassium ionchannels) with phloretin or pimaric acid, paraptosis wasinduced within the glioma cells along with the increasedproduction of heat shock proteins (HSP60, HSP70, HSP90,and Grp94) and the peripheral migration of HMGB1 to thecell surface. When these cells now osmotically lyse, all these“danger signals” are now released and available to stimulatethe local APCs. All vaccinating effects of living mM-CSF+tumor cells can now be reproduced by using our killedBK channel activated/killed glioma cells. All gliomas (rator human) that we tested have been successfully killed byprolonged exposure to BK channel activators [175, 176]. Sothis method can be developed for possible clinical trials.

5. Summary

Progress towards brain cancers by using immunotherapyis slowly moving forward. Initial attempts used nonspe-cific approaches like adjuvants and LAK cells. Nonspe-cific cellular approaches were only effective for a smallminority of fortunate people. The general focus now isdirected towards specific methods. These specific humoralmethods include using monoclonal antibodies and scFVfragmented antibodies. Specific cellular approaches includeusing TILs/CTLs, alloreactive CTL stimulated by MLRs,all appear to have generated some clinical success. Activeimmunization with autologous DCs that have been loadedwith tumor-antigens also appear to generate long-termsurvivors. Glioma cells seem to possess numerous tumor-associated antigens. Identification of other strategies that canbe combined with immunotherapy approaches will certainlyimprove our success against these lethal brain cancers. Welook forward towards the next chapter of this story as thefield continues to mature.

Acknowledgments

The authors thank Dr. Carole Kruse for proofreading ourpaper. They would like to thank Drs. Habib Fakrai and HelenLin from NovaRx for their discussion on safety testing oftheir vaccines. They also thank Drs. Lawrence Lamb, JohnOhlfest, and Nabil Ahmed for their insights into variousaspects of glioma immunotherapy as well as Drs. SharonGardner and Hideho Okada for updating them on theirongoing clinical trials.

References

[1] N. D. Doolittle, L. E. Abrey, W. A. Bleyer et al., “New frontiersin translational research in neuro-oncology and the blood-brain barrier: report of the 10th annual blood-brain barrierdisruption consortium meeting,” Clinical Cancer Research,vol. 11, no. 2, pp. 421–428, 2005.

[2] P. R. Walker, T. Calzascia, and P. Y. Dietrich, “All in thehead: obstacles for immune rejection of brain tumours,”Immunology, vol. 107, no. 1, pp. 28–38, 2002.

[3] M. J. Carson, J. G. Sutcliffe, and I. L. Campbell, “Microgliastimulate naive T-cell differentiation without stimulating T-cell proliferation,” Journal of Neuroscience Research, vol. 55,no. 1, pp. 127–134, 1999.

[4] F. Aloisi, F. Ria, G. Penna, and L. Adorini, “Microglia aremore efficient than astrocytes in antigen processing and inTh1 but not Th2 cell activation,” Journal of Immunology, vol.160, no. 10, pp. 4671–4680, 1998.

[5] J. Karman, C. Ling, M. Sandor, and Z. Fabry, “Initiation ofimmune responses in brain is promoted by local dendriticcells,” Journal of Immunology, vol. 173, no. 4, pp. 2353–2361,2004.

[6] C. Choi, J. Y. Park, J. Lee et al., “Fas ligand and Fas areexpressed constitutively in human astrocytes and the ex-pression increases with IL-1, IL-6, TNF-α, or IFN-γ,” Journalof Immunology, vol. 162, no. 4, pp. 1889–1895, 1999.

[7] R. W. Pelton, B. Saxena, M. Jones, H. L. Moses, and L. I. Gold,“Immunohistochemical localization of TGFβ1, TGFβ2,and TGFβ3 in the mouse embryo: expression patternssuggest multiple roles during embryonic development,”Journal of Cell Biology, vol. 115, no. 4, pp. 1091–1105,1991.

[8] C. A. Janeway, “Approaching the asymptote? Evolution andrevolution in immunology,” Cold Spring Harbor Symposia onQuantitative Biology, vol. 54, no. 1, pp. 1–13, 1989.

[9] M. S. Cookson, H. W. Herr, Z. F. Zhang, S. Soloway, P. C.Sogani, and W. R. Fair, “The treated natural history of highrisk superficial bladder cancer: 15-year outcome,” Journal ofUrology, vol. 158, no. 1, pp. 62–67, 1997.

[10] C. J. Wikstrand and D. D. Bigner, “Hyperimmunization ofnon-human primates with BCG-CW and cultured humanglioma-derived cells. Production of reactive antisera andabsence of EAE induction,” Journal of Neuroimmunology,vol. 1, no. 3, pp. 249–260, 1981.

[11] L. Albright, J. A. Seab, and A. K. Ommaya, “Intracerebraldelayed hypersensitivity reactions in glioblastoma multi-forme patients,” Cancer, vol. 39, no. 3, pp. 1331–1336,1977.

[12] S. Gallucci, M. Lolkema, and P. Matzinger, “Naturaladjuvants: endogenous activators of dendritic cells,” NatureMedicine, vol. 5, no. 11, pp. 1249–1255, 1999.

[13] P. Srivastava, “Interaction of heat shock proteins withpeptides and antigen presenting cells: chaperoning of theinnate and adaptive immune responses,” Annual Review ofImmunology, vol. 20, pp. 395–425, 2002.

[14] R. Suto and P. K. Srivastava, “A mechanism for thespecific immunogenicity of heat shock protein-chaperonedpeptides,” Science, vol. 269, no. 5230, pp. 1585–1588, 1995.

[15] R. J. Binder, N. E. Blachere, and P. K. Srivastava, “Heatshock protein-chaperoned peptides but not free peptidesintroduced into the cytosol are presented efficiently bymajor histocompatibility complex I molecules,” Journal ofBiological Chemistry, vol. 276, no. 20, pp. 17163–17171,2001.

[16] M. W. Graner, R. I. Cumming, and D. D. Bigner, “The heatshock response and chaperones/heat shock proteins in braintumors: surface expression, release, and possible immuneconsequences,” Journal of Neuroscience, vol. 27, no. 42, pp.11214–11227, 2007.

[17] M. W. Graner, O. Alzate, A. M. Dechkovskaia et al., “Pro-teomic and immunologic analyses of brain tumor exosomes,”FASEB Journal, vol. 23, no. 5, pp. 1541–1557, 2009.


[18] M. W. Graner, D. A. Raynes, D. D. Bigner, and V. Guerriero,“Heat shock protein 70-binding protein 1 is highly expressedin high-grade gliomas, interacts with multiple heat shockprotein 70 family members, and specifically binds braintumor cell surfaces,” Cancer Science, vol. 100, no. 10, pp.1870–1879, 2009.

[19] S. Basu, R. J. Binder, T. Ramalingam, and P. K. Srivastava,“CD91 is a common receptor for heat shock proteins gp96,hsp90, hsp70, and calreticulin,” Immunity, vol. 14, no. 3, pp.303–313, 2001.

[20] S. Basu, R. J. Binder, R. Suto, K. M. Anderson, and P. K. Sri-vastava, “Necrotic but not apoptotic cell death releases heatshock proteins, which deliver a partial maturation signal todendritic cells and activate the NF-κB pathway,” InternationalImmunology, vol. 12, no. 11, pp. 1539–1546, 2000.

[21] M. C. Kuppner, R. Gastpar, S. Gelwer et al., “The role ofheat shock protein (hsp70) in dendritic cell maturation:Hsp70 induces the maturation of immature dendritic cellsbut reduces DC differentiation from monocyte precursors,”European Journal of Immunology, vol. 31, no. 5, pp. 1602–1609, 2001.

[22] A. Asea, S. K. Kraeft, E. A. Kurt-Jones et al., “HSP70stimulates cytokine production through a CD 14-dependantpathway, demonstrating its dual role as a chaperone andcytokine,” Nature Medicine, vol. 6, no. 4, pp. 435–442, 2000.

[23] A. T. Parsa, M. Aghi, B. Ahn, J. Bruce, N. Butkowski et al.,“Autologous heat shock protein vaccine for patients withnewly diagnosed and recurrent glioblastoma,” in Proceedingsof the 18th International Conference on Brain Tumor Researchand Therapy, 2010, Abstract 9-2.

[24] R. P. Glick, T. Lichtor, T. S. Kim et al., “Fibroblasts geneticallyengineered to secrete cytokines suppress tumor growth andinduce antitumor immunity to a murine glioma in vivo,”Neurosurgery, vol. 36, no. 3, pp. 548–555, 1995.

[25] J. J. Tzeng, R. F. Barth, N. R. Clendenon, and W. A.Gordon, “Adoptive immunotherapy of a rat glioma usinglymphokine-activated killer cells and interleukin 2,” CancerResearch, vol. 50, no. 14, pp. 4338–4343, 1990.

[26] S. A. Rosenberg, M. T. Lotze, and L. M. Muul, “A progressreport on the treatment of 157 patients with advanced cancerusing lymphokine-activated killer cells and interleukin-2or high-dose interleukin-2 alone,” New England Journal ofMedicine, vol. 316, no. 15, pp. 889–897, 1987.

[27] R. L. Hayes, M. Koslow, E. M. Hiesiger et al., “Improvedlong term survival after intracavitary interleukin-2 andlymphokine-activated killer cells for adults with recurrentmalignant glioma,” Cancer, vol. 76, no. 5, pp. 840–852, 1995.

[28] Y. Huang, R. L. Hayes, S. Wertheim, E. Arbit, and R. Scheff,“Treatment of refractory recurrent malignant glioma withadoptive cellular immunotherapy: a case report,” CriticalReviews in Oncology/Hematology, vol. 39, no. 1-2, pp. 17–23,2001.

[29] R. O. Dillman, C. M. Duma, P. M. Schiltz et al., “Intracavitaryplacement of autologous lymphokine-activated killer (LAK)cells after resection of recurrent glioblastoma,” Journal ofImmunotherapy, vol. 27, no. 5, pp. 398–404, 2004.

[30] R. O. Dillman, C. M. Duma, R. A. Ellis et al., “Intralesionallymphokine-activated killer cells as adjuvant therapy forprimary glioblastoma,” Journal of Immunotherapy, vol. 32,no. 9, pp. 914–919, 2009.

[31] Y. Fujimiya, Y. Suzuki, R. Katakura et al., “In vitro interleukin12 activation of peripheral blood CD3 CD56 and CD3 CD56γδ T cells from glioblastoma patients,” Clinical CancerResearch, vol. 3, no. 4, pp. 633–643, 1997.

[32] M. A. Friese, M. Platten, S. Z. Lutz et al., “MICA/NKG2D-mediated immunogene therapy of experimental gliomas,”Cancer Research, vol. 63, no. 24, pp. 8996–9006, 2003.

[33] N. L. Bryant, G. Y. Gillespie, R. D. Lopez et al., “Preclinicalevaluation of ex vivo expanded/activated γδ T cells forimmunotherapy of glioblastoma multiforme,” Journal ofNeuro-Oncology, vol. 101, no. 2, pp. 179–188, 2010.

[34] L. S. Lamb, “γδ T cells as immune effectors against high-grade gliomas,” Immunologic Research, vol. 45, no. 1, pp.85–95, 2009.

[35] N. L. Bryant, C. Suarez-Cuervo, G. Y. Gillespie et al.,“Characterization and immunotherapeutic potential of γδT-cells in patients with glioblastoma,” Neuro-Oncology, vol.11, no. 4, pp. 357–367, 2009.

[36] J. Tjuvajev, B. Gansbacher, R. Desai et al., “RG-2 gliomagrowth attenuation and severe brain edema caused bylocal production of interleukin-2 and interferon-γ,” CancerResearch, vol. 55, no. 9, pp. 1902–1910, 1995.

[37] E. Visse, P. Siesjo, B. Widegren, and H. O. Sjogren,“Regression of intracerebral rat glioma isografts bytherapeutic subcutaneous immunization with interferon-γ,interleukin-7, or B7-1-transfected tumor cells,” Cancer GeneTherapy, vol. 6, no. 1, pp. 37–44, 1999.

[38] M. R. Graf, M. R. Jadus, J. C. Hiserodt, H. T. Wepsic,and G. A. Granger, “Development of systemic immunityto glioblastoma multiforme using tumor cells geneticallyengineered to express the membrane-associated isoformof macrophage colony-stimulating factor,” Journal ofImmunology, vol. 163, no. 10, pp. 5544–5551, 1999.

[39] R. Sanchez, C. Williams, J. L. Daza et al., “T9 glioma cellsexpressing membrane-macrophage colony stimulating factorproduce CD4+ T cell-associated protective immunity againstT9 intracranial gliomas and systemic immunity againstdifferent syngeneic gliomas,” Cellular Immunology, vol. 215,no. 1, pp. 1–11, 2002.

[40] Y. Chen, T. Douglass, E. W.B. Jeffes et al., “Living T9glioma cells expressing membrane macrophage colony-stimulating factor produce immediate tumor destructionby polymorphonuclear leukocytes and macrophages viaa ”paraptosis”-induced pathway that promotes systemicimmunity against intracranial T9 gliomas,” Blood, vol. 100,no. 4, pp. 1373–1380, 2002.

[41] E. W. B. Jeffes, J. G. Zhang, N. Hoa et al., “Antiangiogenicdrugs synergize with a membrane macrophage colony-stimulating factor-based tumor vaccine to therapeuticallytreat rats with an established malignant intracranial glioma,”Journal of Immunology, vol. 174, no. 5, pp. 2533–2543,2005.

[42] H. Okada, K. M. Giezeman-Smits, H. Tahara et al., “Effectivecytokine gene therapy against an intracranial glioma usinga retrovirally transduced IL-4 plus HSVtk tumor vaccine,”Gene Therapy, vol. 6, no. 2, pp. 219–226, 1999.

[43] U. Herrlinger, C. M. Kramm, K. M. Johnston et al.,“Vaccination for experimental gliomas using GM-CSF-transduced glioma cells,” Cancer Gene Therapy, vol. 4, no. 6,pp. 345–352, 1997.

[44] M. M. Mueller, C. C. Herold-Mende, D. Riede, M. Lange, H.H. Steiner, and N. E. Fusenig, “Autocrine growth regulationby granulocyte colony-stimulating factor and granulocytemacrophage colony-stimulating factor in human gliomaswith tumor progression,” American Journal of Pathology, vol.155, no. 5, pp. 1557–1567, 1999.


[45] T. Nitta, K. Sato, M. Allegretta et al., “Expression of gran-ulocyte colony stimulating factor and granulocyte-macro-phage colony stimulating factor genes in human astrocytomacell lines and glioma specimens,” Brain Research, vol. 571,no. 1, pp. 19–25, 1992.

[46] Y. Liang, M. Diehn, N. Watson et al., “Gene expressionprofiling reveals molecularly and clinically distinct subtypesof glioblastoma multiforme,” Proceedings of the NationalAcademy of Sciences of the United States of America, vol. 102,no. 16, pp. 5814–5819, 2005.

[47] H. S. Phillips, S. Kharbanda, R. Chen et al., “Molecularsubclasses of high-grade glioma predict prognosis, delineatea pattern of disease progression, and resemble stages inneurogenesis,” Cancer Cell, vol. 9, no. 3, pp. 157–173, 2006.

[48] A. S. Chi and P. Y. Wen, “Inhibiting kinases in malignantgliomas,” Expert Opinion on Therapeutic Targets, vol. 11, no.4, pp. 473–496, 2007.

[49] R. Rahman, S. Smith, C. Rahman, and R. Grundy,“Antiangiogenic therapy and mechanisms of tumorresistance in malignant glioma,” Journal of Oncology,vol. 2010, Article ID 251231, 16 pages, 2010.

[50] S. Takano, T. Yamashita, and O. Ohneda, “Moleculartherapeutic targets for glioma angiogenesis,” Journal ofOncology, vol. 2010, Article ID 351908, 11 pages, 2010.

[51] V. Stark-Vance, “Bevacizumab and CPT-11 in the treatmentof relapsed malignant glioma,” Neuro-Oncology, vol. 7, no. 3,p. 369, 2005.

[52] W. B. Pope, A. Lai, P. Nghiemphu, P. Mischel, and T. F.Cloughesy, “MRI in patients with high-grade gliomas treatedwith bevacizumab and chemotherapy,” Neurology, vol. 66,no. 8, pp. 1258–1260, 2006.

[53] J. J. Vredenburgh, A. Desjardins, J. E. Herndon et al.,“Bevacizumab plus irinotecan in recurrent glioblastomamultiforme,” Journal of Clinical Oncology, vol. 25, no. 30, pp.4722–4729, 2007.

[54] H. S. Friedman, M. D. Prados, P. Y. Wen et al., “Bevacizumabalone and in combination with irinotecan in recurrentglioblastoma,” Journal of Clinical Oncology, vol. 27, no. 28,pp. 4733–4740, 2009.

[55] J. L. Spratlin, R. B. Cohen, M. Eadens et al., “Phase I phar-macologic and biologic study of ramucirumab (imc-1121b),a fully human immunoglobulin G monoclonal antibodytargeting the vascular endothelial growth factor receptor-2,”Journal of Clinical Oncology, vol. 28, no. 5, pp. 780–787, 2010.

[56] K. Lamszus, J. Laterra, M. Westphal, and E. M. Rosen, “Scat-ter factor/hepatocyte growth factor (SF/HGF) content andfunction in human gliomas,” International Journal of Devel-opmental Neuroscience, vol. 17, no. 5-6, pp. 517–530, 1999.

[57] H. T. Jun, J. Sun, K. Rex et al., “AMG 102, a fully humananti-hepatocyte growth factor/scatter factor neutralizingantibody, enhances the efficacy of temozolomide or docetaxelin U-87 MG cells and xenografts,” Clinical Cancer Research,vol. 13, no. 22, pp. 6735–6742, 2007.

[58] S. E. Combs, S. Heeger, R. Haselmann, L. Edler, J. Debus,and D. Schulz-Ertner, “Treatment of primary glioblastomamultiforme with cetuximab, radiotherapy and temozolomide(GERT)—phase I/II trial: study protocol,” BMC Cancer, vol.6, article 133, 2006.

[59] M. A. Diaz, J. Rolff, M. Lemm, I. Fichtner, R. Perez, andE. Montero, “Radiosensitisation of U87MG brain tumoursby anti-epidermal growth factor receptor monoclonalantibodies,” British Journal of Cancer, vol. 100, no. 6, pp.950–958, 2009.

[60] U. Bode, S. Buchen, M. Warmuth-Metz, T. Pietsch, F. Bach,and G. Fleischhack, “Final report of a phase II trial ofnimotuzumab in the treatment of refractory and relapsedhigh-grade gliomas in children and adolescents,” Journal ofClinical Oncology, vol. 25, supplement 18, 2007, Proceedingsof the ASCO Annual Meeting (ASCO ’07).

[61] L. Fredriksson, H. Li, and U. Eriksson, “The PDGF family:four gene products form five dimeric isoforms,” Cytokineand Growth Factor Reviews, vol. 15, no. 4, pp. 197–204,2004.

[62] C. S. Cobbs, L. Harkins, M. Samanta et al., “Human cytome-galovirus infection and expression in human malignant gli-oma,” Cancer Research, vol. 62, no. 12, pp. 3347–3350, 2002.

[63] L. Soroceanu, A. Akhavan, and C. S. Cobbs, “Platelet-derivedgrowth factor-α receptor activation is required for humancytomegalovirus infection,” Nature, vol. 455, no. 7211, pp.391–395, 2008.

[64] S. Song, A. J. Ewald, W. Stallcup, Z. Werb, and G. Bergers,“PDGFRβ+ perivascular progenitor cells in tumours regulatepericyte differentiation and vascular survival,” Nature CellBiology, vol. 7, no. 9, pp. 870–879, 2005.

[65] D. A. Reardon, G. Akabani, R. E. Coleman et al., “PhaseII trial of murine 131I-labeled antitenascin monoclonalantibody 81C6 administered into surgically created resectioncavities of patients with newly diagnosed malignant gliomas,”Journal of Clinical Oncology, vol. 20, no. 5, pp. 1389–1397,2002.

[66] P. Riva, A. Arista, V. Tison et al., “Intralesional radioim-munotherapy of malignant gliomas: an effective treatmentin recurrent tumors,” Cancer, vol. 73, pp. 1076–1082,1994.

[67] P. Riva, G. Franceschi, M. Frattarelli et al., “Loco-regionalradioimmunotherapy of high-grade malignant gliomasusing specific monoclonal antibodies labeled with 90Y: aphase I study,” Clinical Cancer Research, vol. 5, no. 10, pp.3275s–3280s, 1999.

[68] T. Nitta, K. Sato, H. Yagita, K. Okumura, and S. Ishii, “Pre-liminary trial of specific targeting therapy against malignantglioma,” Lancet, vol. 335, no. 8686, pp. 368–371, 1990.

[69] C. T. Kuan, C. J. Wikstrand, G. Archer et al., “Increasedbinding affinity enhances targeting of glioma xenografts byEGFRVIII-specific scFV,” International Journal of Cancer,vol. 88, no. 6, pp. 962–969, 2000.

[70] C. T. Kuan, C. J. Reist, C. F. Foulon et al., “125I-labeled anti-epidermal growth factor receptor-vIII single-chain Fvexhibits specific and high-level targeting of glioma xeno-grafts,” Clinical Cancer Research, vol. 5, no. 6, pp. 1539–1549,1999.

[71] C. T. Kuan, N. Srivastava, R. E. McLendon, W. A. Marasco,M. R. Zalutsky, and D. D. Bigner, “Recombinant single-chainvariable fragment antibodies against extracellular epitopesof human multidrug resistance protein MRP3 for targetingmalignant gliomas,” International Journal of Cancer, vol. 127,no. 3, pp. 598–611, 2010.

[72] F. S. Hodi, S. J. O’Day, D. F. McDermott et al., “Improvedsurvival with ipilimumab in patients with metastaticmelanoma,” New England Journal of Medicine, vol. 363, no.8, pp. 711–723, 2010.

[73] P. E. Fecci, A. E. Sweeney, P. M. Grossi et al., “Systemicanti-CD25 monoclonal antibody administration safelyenhances immunity in murine glioma without eliminatingregulatory T cells,” Clinical Cancer Research, vol. 12, no. 14,pp. 4294–4305, 2006.


[74] M. Wrensch, J. K. Wiencke, J. Wiemels et al., “Serum IgE,tumor epidermal growth factor receptor expression, andinherited polymorphisms associated with glioma survival,”Cancer Research, vol. 66, no. 8, pp. 4531–4541, 2006.

[75] J. L. Wiemels, J. K. Wiencke, J. Patoka et al., “Reducedimmunoglobulin E and allergy among adults with gliomacompared with controls,” Cancer Research, vol. 64, no. 22,pp. 8468–8473, 2004.

[76] E. Linos, T. Raine, A. Alonso, and D. Michaud, “Atopy andrisk of brain tumors: a meta-analysis,” Journal of the NationalCancer Institute, vol. 99, no. 20, pp. 1544–1550, 2007.

[77] S. A. Rosenberg, P. Spiess, and R. Lafreniere, “A newapproach to the adoptive immunotherapy of cancer withtumor-infiltrating lymphocytes,” Science, vol. 233, no. 4770,pp. 1318–1321, 1986.

[78] F. P. Holladay, G. Lopez, M. De, R. A. Morantz, and G. W.Wood, “Generation of cytotoxic immune responses againsta rat glioma by in vivo priming and secondary in vitrostimulation with tumor cells,” Neurosurgery, vol. 30, no. 4,pp. 499–505, 1992.

[79] F. P. Holladay, T. Heitz, Y. L. Chen, M. Chiga, G. W. Wood,and J. J. Olson, “Successful treatment of a malignant ratglioma with cytotoxic T lymphocytes,” Neurosurgery, vol. 31,no. 3, pp. 528–533, 1992.

[80] G. E. Plautz, J. E. Touhalisky, and S. Shu, “Treatment ofmurine gliomas by adoptive transfer of ex vivo activatedtumor-draining lymph node cells,” Cellular Immunology, vol.178, no. 2, pp. 101–107, 1997.

[81] G. P. Dunn, I. F. Dunn, and W. T. Curry, “Focus on TILs:prognostic significance of tumor infiltrating lymphocytes inhuman glioma,” Cancer Immunity, vol. 7, p. 12, 2007.

[82] K. B. Quattrocchi, C. H. Miller, S. Cush et al., “Pilot studyof local autologous tumor infiltrating lymphocytes forthe treatment of recurrent malignant gliomas,” Journal ofNeuro-Oncology, vol. 45, no. 2, pp. 141–157, 1999.

[83] G. E. Plautz, D. W. Miller, G. H. Barnett et al., “T celladoptive immunotherapy of newly diagnosed gliomas,”Clinical Cancer Research, vol. 6, no. 6, pp. 2209–2218, 2000.

[84] C. A. Kruse, L. Cepeda, B. Owens, S. D. Johnson, J. Stears,and K. O. Lillehei, “Treatment of recurrent glioma withintracavitary alloreactive cytotoxic T lymphocytes andinterleukin-2,” Cancer Immunology Immunotherapy, vol. 45,no. 2, pp. 77–87, 1997.

[85] L. Yang, K. Y. Ng, and K. O. Lillehei, “Cell-mediatedimmunotherapy: a new approach to the treatment ofmalignant glioma,” Cancer Control, vol. 10, no. 2, pp.138–147, 2003.

[86] C. A. Kruse and D. Rubinstein, “Cytotoxic T lymphocytesreactive to patient major histocompatibility proteins fortherapy of recurrent primary brain tumors,” in Brain TumorImmunotherapy, L. M. Liau, D. P. Becker, T. F. Cloughsey,and D. Bigner, Eds., pp. 149–170, Humana Press, New York,NY, USA, 2001.

[87] M. T. Esser, R. D. Dinglasan, B. Krishnamurthy, C. A.Gullo, M. B. Graham, and V. L. Braciale, “IL-2 induces Fasligand/Fas (CD95L/CD95) cytotoxicity in CD8 and CD4 Tlymphocyte clones,” Journal of Immunology, vol. 158, no. 12,pp. 5612–5618, 1997.

[88] N. S. Williams and V. H. Engelhard, “Identification of apopulation of CD4+ CTL that utilizes a perforin- ratherthan a Fas ligand-dependent cytotoxic mechanism,” Journalof Immunology, vol. 156, no. 1, pp. 153–159, 1996.

[89] K. M. Giezeman-Smits, H. Okada, C. S. Brissette-Storkus etal., “Cytokine gene therapy of gliomas: induction of reactiveCD4+ T cells by interleukin-4-transfected 9L gliosarcoma isessential for protective immunity,” Cancer Research, vol. 60,no. 9, pp. 2449–2457, 2000.

[90] H. Okada, J. Attanucci, K. M. Giezeman-Smits et al., “Im-munization with an antigen identified by cytokine tumorvaccine-assisted SEREX (CAS) suppressed growth of therat 9L glioma in Vivo,” Cancer Research, vol. 61, no. 6, pp.2625–2631, 2001.

[91] C. Delgado, N. Hoa, L. L. Callahan et al., “Generationof human innate immune responses towards membranemacrophage colony stimulating factor (mM-CSF) express-ing U251 glioma cells within immunodeficient (NIH-nu/beige/xid) mice,” Cytokine, vol. 38, no. 3, pp. 165–176, 2007.

[92] A. El Andaloussi and M. S. Lesniak, “An increase in CD4+CD25+FOXP3+ regulatory T cells in tumor-infiltratinglymphocytes of human glioblastoma multiforme,” Neuro-Oncology, vol. 8, no. 3, pp. 234–243, 2006.

[93] P. E. Fecci, D. A. Mitchell, J. F. Whitesides et al., “Increasedregulatory T-cell fraction amidst a diminished CD4compartment explains cellular immune defects in patientswith malignant glioma,” Cancer Research, vol. 66, no. 6, pp.3294–3302, 2006.

[94] A. B. Heimberger, M. Abou-Ghazal, C. Reina-Ortiz et al.,“Incidence and prognostic impact of FoxP3+ regulatory Tcells in human gliomas,” Clinical Cancer Research, vol. 14,no. 16, pp. 5166–5172, 2008.

[95] M. Tsuji, N. Komatsu, S. Kawamoto et al., “Preferentialgeneration of follicular B helper T cells from Foxp3+ Tcells in gut Peyer’s patches,” Science, vol. 323, no. 5920, pp.1488–1492, 2009.

[96] F. Annunziato, L. Cosmi, F. Liotta, E. Maggi, and S.Romagnani, “The phenotype of human Th17 cells and theirprecursors, the cytokines that mediate their differentiationand the role of Th17 cells in inflammation,” InternationalImmunology, vol. 20, no. 11, pp. 1361–1368, 2008.

[97] Y. Y. Wan and R. A. Flavell, “How diverse—CD4 effector Tcells and their functions,” Journal of Molecular Cell Biology,vol. 1, no. 1, pp. 20–36, 2009.

[98] D. A. Wainwright, S. Sengupta, Y. Han, I. V. Ulasov, and M.S. Lesniak, “The presence of IL-17A and T helper 17 cells inexperimental mouse brain tumors and human gliom,” PLoSONE, vol. 5, no. 10, article e15390, 2010.

[99] P. Muranski, A. Boni, P. A. Antony et al., “Tumor-specificTh17-polarized cells eradicate large established melanoma,”Blood, vol. 112, no. 2, pp. 362–373, 2008.

[100] K. S. Kahlon, C. Brown, L. J. N. Cooper, A. Raubitschek, S. J.Forman, and M. C. Jensen, “Specific recognition and killingof glioblastoma multiforme by interleukin 13-zetakineredirected cytolytic T cells,” Cancer Research, vol. 64, no. 24,pp. 9160–9166, 2004.

[101] N. Ahmed, V. S. Salsman, Y. Kew et al., “HER2-specificT cells target primary glioblastoma stem cells and induceregression of autologous experimental tumors,” ClinicalCancer Research, vol. 16, no. 2, pp. 474–485, 2010.

[102] S. S. Bullain, A. Sahin, O. Szentirmai et al., “Genetically engi-neered T cells to target EGFRvIII expressing glioblastoma,”Journal of Neuro-Oncology, vol. 94, no. 3, pp. 373–382, 2009.

[103] A. S. Y. Lo, Q. Ma, D. L. Liu, and R. P. Junghans, “Anti-GD3 chimeric sFv-CD28/T-cell receptor ζ designer Tcells for treatment of metastatic melanoma and otherneuroectodermal tumors,” Clinical Cancer Research, vol. 16,no. 10, pp. 2769–2780, 2010.


[104] N. Ahmed, M. Ratnayake, B. Savoldo et al., “Regressionof experimental medulloblastoma following transfer ofHER2-specific T cells,” Cancer Research, vol. 67, no. 12, pp.5957–5964, 2007.

[105] H. Okada, H. Tahara, M. R. Shurin et al., “Bone marrow-derived dendritic cells pulsed with a tumor-specific peptideelicit effective anti-tumor immunity against intracranialneoplasms,” International Journal of Cancer, vol. 78, no. 2,pp. 196–201, 1998.

[106] L. M. Liau, K. L. Black, R. M. Prins et al., “Treatment ofintracranial gliomas with bone marrow-derived dendriticcells pulsed with tumor antigens,” Journal of Neurosurgery,vol. 90, no. 6, pp. 1115–1124, 1999.

[107] A. B. Heimberger, L. E. Crotty, G. E. Archer et al.,“Bone marrow-derived dendritic cells pulsed with tumorhomogenate induce immunity against syngeneic intracere-bral glioma,” Journal of Neuroimmunology, vol. 103, no. 1,pp. 16–25, 2000.

[108] J. Schettini and P. Mukherjee, “Physiological role ofplasmacytoid dendritic cells and their potential use in cancerimmunity,” Clinical and Developmental Immunology, vol.2008, Article ID 106321, 10 pages, 2008.

[109] H. Ardon, S. De Vleeschouwer, F. Van Calenbergh et al.,“Adjuvant dendritic cell-based tumour vaccination forchildren with malignant brain tumours,” Pediatric Blood andCancer, vol. 54, no. 4, pp. 519–525, 2010.

[110] S. De Vleeschouwer, S. Fieuws, S. Rutkowski et al., “Post-operative adjuvant dendritic cell-based immunotherapy inpatients with relapsed glioblastoma multiforme,” ClinicalCancer Research, vol. 14, no. 10, pp. 3098–3104, 2008.

[111] L. M. Liau, R. M. Prins, S. M. Kiertscher et al., “Dendritic cellvaccination in glioblastoma patients induces systemic andintracranial T-cell responses modulated by the local centralnervous system tumor microenvironment,” Clinical CancerResearch, vol. 11, no. 15, pp. 5515–5525, 2005.

[112] L. M. Liau, R. M. Prins, S. M. Kiertscher et al., “Immuno-therapy for patients with malignant glioma: from theoreticalprinciples to clinical applications,” Expert Review ofNeurotherapeutics, vol. 6, no. 10, pp. 1481–1494, 2006.

[113] J. S. Yu, C. J. Wheeler, P. M. Zeltzer et al., “Vaccination ofmalignant glioma patients with peptide-pulsed dendriticcells elicits systemic cytotoxicity and intracranial T-cellinfiltration,” Cancer Research, vol. 61, no. 3, pp. 842–847,2001.

[114] J. S. Yu, G. Liu, H. Ying, W. H. Yong, K. L. Black, and C. J.Wheeler, “Vaccination with tumor lysate-pulsed dendriticcells elicits antigen-specific, cytotoxic T-cells in patients withmalignant glioma,” Cancer Research, vol. 64, no. 14, pp.4973–4979, 2004.

[115] W. Kim and L. M. Liau, “Dendritic cell vaccines for braintumors,” Neurosurgery Clinics of North America, vol. 21, no.1, pp. 139–157, 2010.

[116] T. Kikuchi, Y. Akasaki, M. Irie, S. Homma, T. Abe, and T.Ohno, “Results of a phase I clinical trial of vaccination ofglioma patients with fusions of dendritic and glioma cells,”Cancer Immunology, Immunotherapy, vol. 50, no. 7, pp.337–344, 2001.

[117] T. Kikuchi, Y. Akasaki, T. Abe et al., “Vaccination of gliomapatients with fusions of dendritic and glioma cells andrecombinant human interleukin 12,” Journal of Immuno-therapy, vol. 27, no. 6, pp. 452–459, 2004.

[118] P. M. Schiltz, G. J. Lee, J. G. Zhang et al., “Human allogeneicand murine xenogeneic dendritic cells are cytotoxic tohuman tumor cells via two distinct pathways,” CancerBiotherapy and Radiopharmaceuticals, vol. 22, no. 5, pp.672–683, 2007.

[119] M. Schmitz, S. Zhao, Y. Deuse et al., “Tumoricidal potentialof native blood dendritic cells: direct tumor cell killingand activation of NK cell-mediated cytotoxicity,” Journal ofImmunology, vol. 174, no. 7, pp. 4127–4134, 2005.

[120] C. W. Chan, E. Crafton, H.-N. Fan et al., “Interferon-producing killer dendritic cells provide a link between innateand adaptive immunity,” Nature Medicine, vol. 12, no. 2, pp.207–213, 2006.

[121] M. A. Cheever, J. P. Allison, A. S. Ferris et al., “Theprioritization of cancer antigens: a National Cancer Institutepilot project for the acceleration of translational research,”Clinical Cancer Research, vol. 15, no. 17, pp. 5323–5337, 2009.

[122] T. Weinschenk, O. Schoor, C. Trautwein, N. Hilf, S. Walter,and H. Singh, “Novel immunotherapy against several tumorsincluding neuronal and brain tumors,” United States Patentno. 2010/0158931 A1, 2010.

[123] P. Parajuli, V. Pisarev, J. Sublet et al., “Immunization withwild-type p53 gene sequences coadministered with Flt3ligand induces an antigen-specific type 1 T-cell response,”Cancer Research, vol. 61, no. 22, pp. 8227–8234, 2001.

[124] J. G. A. Houbiers, H. W. Nijman, S. H. Van Der Burg et al., “Invitro induction of human cytotoxic T Iymphocyte responsesagainst peptides of mutant and wild-type p53,” EuropeanJournal of Immunology, vol. 23, no. 9, pp. 2072–2077, 1993.

[125] S. Gnjatic, Z. Cai, M. Viguier, S. Chouaib, J. G. Guillet,and J. Choppin, “Accumulation of the p53 protein allowsrecognition by human CTL of a wild-type p53 epitopepresented by breast carcinomas and melanomas,” Journal ofImmunology, vol. 160, no. 1, pp. 328–333, 1998.

[126] T. Soussi, “p53 antibodies in the sera of patients with varioustypes of cancer: a review,” Cancer Research, vol. 60, no. 7, pp.1777–1788, 2000.

[127] J. G. Zhang, C. A. Kruse, L. Driggers et al., “Tumor antigenprecursor protein profiles of adult and pediatric braintumors identify potential targets for immunotherapy,”Journal of Neuro-Oncology, vol. 88, no. 1, pp. 65–76, 2008.

[128] G. G. Gomez and C. A. Kruse, “Mechanisms of malignantglioma immune resistance and sources of immuno-suppression,” Gene Therapy and Molecular Biology, vol. 10,no. 1, pp. 133–146, 2006.

[129] R. M. Prins, T. F. Cloughesy, and L. M. Liau, “Cyto-megalovirus immunity after vaccination with autologousglioblastoma lysate,” New England Journal of Medicine, vol.359, no. 5, pp. 539–541, 2008.

[130] D. A. Mitchell, W. Xie, R. Schmittling et al., “Sensitivedetection of human cytomegalovirus in tumors andperipheral blood of patients diagnosed with glioblastoma,”Neuro-Oncology, vol. 10, no. 1, pp. 10–18, 2008.

[131] L. Bao, K. Dunham, M. Stamer, K. M. Mulieri, and K. G.Lucas, “Expansion of cytomegalovirus pp65 and IE-1 specificcytotoxic T lymphocytes for cytomegalovirus-specificimmunotherapy following allogeneic stem cell transplanta-tion,” Biology of Blood and Marrow Transplantation, vol. 14,no. 10, pp. 1156–1162, 2008.

[132] L. Bao, Q. Sun, and K. G. Lucas, “Rapid generation ofCMV pp65-specific T cells for immunotherapy,” Journal ofImmunotherapy, vol. 30, no. 5, pp. 557–561, 2007.


[133] I. F. Parney, M. A. Farr-Jones, L. J. Chang, and K. C. Petruk,“Human glioma immunobiology in vitro: implications forimmunogene therapy,” Neurosurgery, vol. 46, no. 5, pp.1169–1178, 2000.

[134] I. F. Parney, K. C. Petruk, C. Zhang, M. Farr-Jones, D. B.Sykes, and L. J. Chang, “Granulocyte-macrophage colony-stimulating factor and B7-2 combination immunogenetherapy in an allogeneic Hu-PBL-SCID/beige mouse-humanglioblastoma multiforme model,” Human Gene Therapy, vol.8, no. 9, pp. 1073–1085, 1997.

[135] J. G. Zhang, J. Eguchi, C. A. Kruse et al., “Antigenic profilesof glioma cells to generate allogeneic vaccines or DC-basedtherapeutics,” Clinical Cancer Research, vol. 13, no. 2, pp.566–575, 2007.

[136] H. Fakhrai, O. Dorigo, D. L. Shawler et al., “Eradication ofestablished intracranial rat gliomas by transforming growthfactor β antisense gene therapy,” Proceedings of the NationalAcademy of Sciences of the United States of America, vol. 93,no. 7, pp. 2909–2914, 1996.

[137] M. S. Mahaley, D. D. Bigner, and L. F. Dudka, “Immuno-biology of primary intracranial tumors—part 7: activeimmunization of patients with anaplastic human gliomacells: a pilot study,” Journal of Neurosurgery, vol. 59, no. 2,pp. 201–207, 1983.

[138] M. S. Mahaley Jr., G. Y. Gillespie, and R. P. Gillespie,“Immunobiology of primary intracranial tumors—part 8:serological responses to active immunization of patients withanaplastic gliomas,” Journal of Neurosurgery, vol. 59, no. 2,pp. 208–216, 1983.

[139] K. El-Shami, B. Tirosh, E. Bar-Haım et al., “MHC class I-restricted epitope spreading in the context of tumor rejectionfollowing vaccination with a single immunodominant CTLepitope,” European Journal of Immunology, vol. 29, no. 10,pp. 3295–3301, 1999.

[140] S. K. Singh, C. Hawkins, I. D. Clarke et al., “Identification ofhuman brain tumour initiating cells,” Nature, vol. 432, no.7015, pp. 396–401, 2004.

[141] S. Bao, Q. Wu, R. E. McLendon et al., “Glioma stem cellspromote radioresistance by preferential activation of theDNA damage response,” Nature, vol. 444, no. 7120, pp.756–760, 2006.

[142] G. Liu, X. Yuan, Z. Zeng et al., “Analysis of gene expressionand chemoresistance of CD133+ cancer stem cells inglioblastoma,” Molecular Cancer, vol. 5, article 67,2006.

[143] E. M. Hurt and W. L. Farrar, “Purification and characteriza-tion of cancer stem cells,” in Cancer Stem Cells, W. L. Farrar,Ed., Chapter 1, Cambridge University Press, 2009.

[144] L. Prestegarden, A. Svendsen, J. Wang et al., “Glioma cellpopulations grouped by different cell type markers drivebrain tumor growth,” Cancer Research, vol. 70, no. 11, pp.4274–4279, 2010.

[145] S. Pellegatta, P. L. Poliani, D. Corno et al., “Neurospheresenriched in cancer stem-like cells are highly effective ineliciting a dendritic cell-mediated immune response againstmalignant gliomas,” Cancer Research, vol. 66, no. 21, pp.10247–10252, 2006.

[146] Q. Xu, G. Liu, X. Yuan et al., “Antigen-specific T-cellresponse from dendritic cell vaccination using cancer stem-like cell-associated antigens,” Stem Cells, vol. 27, no. 8, pp.1734–1740, 2009.

[147] D. D. Bigner, O. M. Pitts, and C. J. Wikstrand, “Induction oflethal experimental allergic encephalomyelitis in nonhumanprimates and guinea pigs with human glioblastomamultiforme tissue,” Journal of Neurosurgery, vol. 55, no. 1,pp. 32–42, 1981.

[148] Y. Q. Wei, Q. R. Wang, X. Zhao et al., “Immunotherapyof tumors with xenogeneic endothelial cells as a vaccine,”Nature Medicine, vol. 6, no. 10, pp. 1160–1166, 2000.

[149] J. Folkman, “Angiogenesis in cancer, vascular, rheumatoidand other disease,” Nature Medicine, vol. 1, no. 1, pp. 27–31,1995.

[150] J. J. Virrey, E. B. Golden, W. Sivakumar et al., “Glioma-associated endothelial cells are chemoresistant totemozolomide,” Journal of Neuro-Oncology, vol. 95, no.1, pp. 13–22, 2009.

[151] R. Wang, K. Chadalavada, J. Wilshire et al., “Glioblastomastem-like cells give rise to tumour endothelium,” Nature, vol.468, no. 7325, pp. 829–835, 2010.

[152] L. Ricci-Vitiani, R. Pallini, M. Biffoni et al., “Tumour vas-cularization via endothelial differentiation of glioblastomastem-like cells,” Nature, vol. 468, no. 7325, pp. 824–830, 2010.

[153] L. M. Liau, K. L. Black, N. A. Martin et al., “Treatmentof a patient by vaccination with autologous dendritic cellspulsed with allogeneic major histocompatibility complexclass I-matched tumor peptides. Case Report,” NeurosurgicalFocus, vol. 9, no. 6, article e8, 2000.

[154] L. M. Liau, M. L. Salgaller, S. M. Kiertscher et al., “Dendriticcell vaccine presenting eluted tumor antigens for treatmentof patients with glioblastoma multiforme interim phaseI trial results of DCVax[tm]-Brain,” in Proceedings of theAmerican Association for Cancer Research, San Francisco,Calif, USA, 2002.

[155] J. S. Yu, C. J. Wheeler, P. M. Zeltzer et al., “Vaccinationof malignant glioma patients with peptide-pulsed dendriticcells elicits systemic cytotoxicity and intracranial T-cell infil-tration,” Cancer Research, vol. 61, no. 3, pp. 842–847, 2001.

[156] A. B. Heimberger and J. H. Sampson, “The PEPvIII-KLH(CDX-110) vaccine in glioblastoma multiforme patients,”Expert Opinion on Biological Therapy, vol. 9, no. 8, pp.1087–1098, 2009.

[157] A. B. Heimberger, W. Sun, S. F. Hussain et al.,“Immunological responses in a patient with glioblastomamultiforme treated with sequential courses of temozolomideand immunotherapy: case study,” Neuro-Oncology, vol. 10,no. 1, pp. 98–103, 2008.

[158] N. Butowski, K. R. Lamborn, B. L. Lee et al., “A NorthAmerican brain tumor consortium phase II study of poly-ICLC for adult patients with recurrent anaplastic gliomas,”Journal of Neuro-Oncology, vol. 91, no. 2, pp. 183–189, 2009.

[159] D. E. Bredesen, R. V. Rao, and P. Mehlen, “Cell death in thenervous system,” Nature, vol. 443, no. 7113, pp. 796–802,2006.

[160] R. S. Hotchkiss, A. Strasser, J. E. McDunn, and P. E. Swanson,“Cell death,” New England Journal of Medicine, vol. 361, no.16, pp. 1570–1583, 2009.

[161] A. H. Wyllie, J. F. R. Kerr, and A. R. Currie, “Cell death: thesignificance of apoptosis,” International Review of Cytology,vol. 68, pp. 251–306, 1980.

[162] J. J. Cohen, “Apoptosis,” Immunology Today, vol. 14, no. 3,pp. 126–130, 1993.

[163] K. L. Rock and L. Shen, “Cross-presentation: underlyingmechanisms and role in immune surveillance,” Immuno-logical Reviews, vol. 207, pp. 166–183, 2005.


[164] H. Kono and K. L. Rock, “How dying cells alert the immunesystem to danger,” Nature Reviews Immunology, vol. 8, no. 4,pp. 279–289, 2008.

[165] G. Dranoff, E. Jaffee, A. Lazenby et al., “Vaccinationwith irradiated tumor cells engineered to secrete murinegranulocyte-macrophage colony-stimulating factor stimu-lates potent, specific, and long-lasting anti-tumor immunity,”Proceedings of the National Academy of Sciences of the UnitedStates of America, vol. 90, no. 8, pp. 3539–3543, 1993.

[166] T. F. Witham, M. L. Erff, H. Okada et al., “7-Hydroxystauro-sporine-induced apoptosis in 9L glioma cells provides aneffective antigen source for dendritic cells and yields apotent vaccine strategy in an intracranial glioma model,”Neurosurgery, vol. 50, no. 6, pp. 1327–1335, 2002.

[167] J. S. Samaddar, V. T. Gaddy, J. Duplantier et al., “A role formacroautophagy in protection against 4-hydroxytamoxifen-induced cell death and the development of antiestrogenresistance,” Molecular Cancer Therapeutics, vol. 7, no. 9, pp.2977–2987, 2008.

[168] Y. Li, L. X. Wang, G. Yang, F. Hao, W. J. Urba, and H. M.Hu, “Efficient cross-presentation depends on autophagy intumor cells,” Cancer Research, vol. 68, no. 17, pp. 6889–6895,2008.

[169] V. L. Crotzer and J. S. Blum, “Autophagy and its role in MHC-mediated antigen presentation,” Journal of Immunology, vol.182, no. 6, pp. 3335–3341, 2009.

[170] S. Sperandio, I. de Belle, and D. E. Bredesen, “An alternative,non-apoptotic form of programmed cell death,” Proceedingsof the National Academy of Sciences of the Unites States ofAmerica, vol. 97, no. 26, pp. 14376–14381, 2000.

[171] H. Yang, H. Wang, C. J. Czura, and K. J. Tracey, “Thecytokine activity of HMGB1,” Journal of Leukocyte Biology,vol. 78, no. 1, pp. 1–8, 2005.

[172] A. Melcher, S. Todryk, N. Hardwick, M. Ford, M. Jacobson,and R. G. Vile, “Tumor immunogenicity is determined by themechanism of cell death via induction of heat shock proteinexpression,” Nature Medicine, vol. 4, no. 5, pp. 581–587,1998.

[173] S. Gallucci and P. Matzinger, “Danger signals: SOS to theimmune system,” Current Opinion in Immunology, vol. 13,no. 1, pp. 114–119, 2001.

[174] K. L. Rock and H. Kono, “The inflammatory response tocell death,” Annual Review of Pathology, vol. 3, pp. 99–126,2008.

[175] Q. Dan, R. Sanchez, C. Delgado et al., “Non-immunogenicmurine hepatocellular carcinoma Hepa 1–6 cells expressingthe membrane form of macrophage colony stimulatingfactor are rejected in vivo and lead to CD8+ T-cell immunityagainst the parental tumor,” Molecular Therapy, vol. 4, no. 5,pp. 427–437, 2001.

[176] C. C. Williams, H. Trinh, T. V. Tran et al., “Membraneexpression of macrophage colony stimulating factor onMADB106 breast cancer cells is not associated with in vitrocytotoxicity by macrophages but is associated with the induc-tion lasting tumor immunity against syngeneic breast cancercells in vivo,” Molecular Therapy, vol. 3, no. 2, pp. 216–224,2001.

[177] N. Hoa, J. G. Zhang, C. Delgado et al., “BK channelactivation results in paraptosis: a possible mechanism toexplain monocyte-mediated killing of tumor cells expressingmembrane macrophage colony stimulating factor,”Laboratory Investigation, vol. 87, no. 2, pp. 115–129, 2007.

[178] N. Hoa, M. P. Myers, T. G. Douglass et al., “Molecularmechanisms of paraptosis induction: implications for anon-genetically modified tumor vaccine,” PLoS ONE, vol. 4,no. 2, article e4631, 2009.

[179] M. R. Jadus, Y. Chen, M. T. Boldaji et al., “Human U251MGglioma cells expressing the membrane form of macrophagecolony-stimulating factor (mM-CSF) are killed by humanmonocytes in vitro and are rejected within immunodeficientmice via paraptosis that is associated with increasedexpression of three different heat shock proteins,” CancerGene Therapy, vol. 10, no. 5, pp. 411–420, 2003.


Review Article

Immune Suppression in Head and Neck Cancers: A Review

Anaelle Duray,1 Stephanie Demoulin,2 Pascale Hubert,2

Philippe Delvenne,2, 3 and Sven Saussez1, 4

1 Laboratory of Anatomy, Faculty of Medicine and Pharmacy, University of Mons, 7000 Mons, Belgium2 Department of Pathology, CHU Sart-Tilman, University of Liege, 4000 Liege, Belgium3 Belgian National Fund for Scientific Research (FNRS), 1000 Brussels, Belgium4 Department of Oto-Rhino-Laryngology, CHU Saint-Pierre, Universite Libre de Bruxelles, 1000 Brussels, Belgium

Correspondence should be addressed to Sven Saussez, [email protected]

Received 30 June 2010; Revised 20 December 2010; Accepted 27 December 2010

Academic Editor: Enrico Maggi

Copyright © 2010 Anaelle Duray et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Head and neck squamous cell carcinomas (HNSCCs) are the sixth most common cancer in the world. Despite significantadvances in the treatment modalities involving surgery, radiotherapy, and concomitant chemoradiotherapy, the 5-year survivalrate remained below 50% for the past 30 years. The worse prognosis of these cancers must certainly be link to the fact thatHNSCCs strongly influence the host immune system. We present a critical review of our understanding of the HNSCC escapeto the antitumor immune response such as a downregulation of HLA class I and/or components of APM. Antitumor responsesof HNSCC patients are compromised in the presence of functional defects or apoptosis of T-cells, both circulating and tumor-infiltrating. Langerhans cells are increased in the first steps of the carcinogenesis but decreased in invasive carcinomas. Theaccumulation of macrophages in the peritumoral areas seems to play a protumoral role by secreting VEGF and stimulating theneoangiogenesis.

1. Epidemiology, Treatment, and Prognosis

Head and neck squamous cell carcinomas (HNSCCs) remaina significant cause of morbidity worldwide, with approx-imately 650,000 new cases diagnosed each year [1, 2].HNSCCs constitute a collection of diseases that, althoughunited by location and histology, can become very differenttypes of tumors that differ in pathogenesis, biology, sublo-cation and treatment and that can affect quality of life,including survival [1, 2]. HNSCC patients associated withlow clinical stages (stages I and II) have similar survival rates,with a 5-year survival between 70% and 90%, independentof the sublocation [3]. In contrast, HNSCC patients withadvanced clinical stages (stages III and IV) display completelydifferent survival rates depending on the histological typeof the tumor and its sublocation [3, 4]. The treatment ofHNSCC patients with advanced stages of disease combinessurgery, radiation oncology, medical oncology, medicalimaging, and clinical pathology [1–4]. This type of collabo-rative medical approach was initiated as early as 1970, when

Fletcher and Evers reported the first convincing evidence ofthe benefits of combining radiotherapy with surgery [5]. Inthis context, cisplatin was investigated in the treatment ofHNSCC in the early 1970s, and from the late 1970s to theearly 1990s, promising results were obtained with the use ofvarious combinations of postoperative chemotherapy withradiotherapy in randomized [6] and nonrandomized studies[7]. In the early 2000s, the Radiation Therapy OncologyGroup [4] and the European Organization for Researchand Treatment of Cancer (EORTC) [8] conducted tworandomized studies to test the relative efficacy of concurrentpostoperative cisplatin administration and radiotherapy inthe treatment of HNSCC. These two studies demonstratedthat local control of the disease was significantly higherin the combined therapy group than in the group thatreceived radiotherapy alone [4, 8]. Unfortunately, thesecombined treatments were frequently associated with adverseside effects. Although significant progress has been observedafter combined treatments, a number of statements currentlyremain valid concerning HNSCCs: (i) almost two-thirds of


HNSCC patients have advanced forms (stages III and IV)of the disease at diagnosis, (ii) 50% of the patients die ofHNSCC within the two years following initial diagnosis, and(iii) every year, 5% of the patients develop additional primarytumors. Therefore, novel approaches seem to be requiredto provide head and neck oncologists with a more effectivearmamentarium against this challenging disease [9, 10].

2. Immune System and Cancers

In the 1950s, Burnet and Thomas proposed the concept ofimmune surveillance of cancer. This physiological functionwould have the ability to recognize tumor cells as abnormalcells and to destroy them before they develop into dangerous,detectable tumors [11]. Tumor growth, invasion, and metas-tasis are important aspects of the tumor immune escape.The different mechanisms that are developed by tumor cellsare a defect of expression of antigens on the tumor cellsurface; a loss or a reduction of the expression of MHC(major histocompatibility complex) class 1 molecules, a lossof expression of costimulatory molecules, the production ofimmunosuppressive molecules such as transforming growthfactor (TGF)-β, prostaglandin (PG) E2 and adenosine, or ofcytokines such as interleukin (IL)-6 and IL-10, the resistanceto apoptosis, and/or the expression of Fas ligand (FasL),which leads to the death of tumor-infiltrating lymphocytes(TILs) [12–15] (Figure 1).

Moreover, tumor cells recruit macrophages called tumorassociated macrophages (TAMs) by secreting the colonystimulating factor (CSF-1), the chemokine ligand 2, 3, 4, 5,and 8 (CCL2, 3, 4, 5, and 8) and the vascular endothelialgrowth factor (VEGF) [16–18]. TAMs constitute the majorinflammatory component of tumor microenvironment [19–22]. Their functions within the tumor site are various andsometimes paradoxical. Indeed, according to the environ-mental stimuli, macrophages present two different pheno-types. Macrophages of the M1 phenotype kill pathogensand promote the activation of cytotoxic CD8+ T cells andthe differentiation of naıve CD4+ T cells into Th1 effectorcells and Th17 cells [17, 18, 23]. M2 macrophages stimulateCD4+ Th2 cells and regulatory T cell differentiation andcan promote angiogenesis and tissue remodeling [17, 18,23] (Figure 1). Multiple studies have shown a correlationbetween a large number of macrophages in the tumormicroenvironment and a worse prognosis. TAMs, therefore,exercise different protumor functions associated with the M2phenotype [22, 23].

During tumor initiation, TAMs create a favorable envi-ronment for tumor growth by secreting epidermal growthfactor (EGF), platelet-derived growth factor (PDGF), TGF-β, IL-6, IL-1, and tumor necrosis factor (TNF)-α. In hypoxicareas, TAMs stimulate angiogenesis (by secreting severalfactors, such as TGF-β, VEGF, granulocyte macrophage(GM)-CSF, TNF-α, IL-1, IL-6, and IL-8), promote tumorcell migration and invasion (via matrix metalloproteinases(MMPs), TNF-α, and IL-1) and induce immunosuppression(via TGF-β, PGE2, and IL-10). A subpopulation of TAMs,which are associated with factors such as EGF, is able topromote metastasis by guiding tumor cells in the stroma

toward blood vessels, where they then escape into thecirculation [16–18, 24] (Figure 1). On the other hand, otherstudies have shown that TAMs could also be correlatedwith a good prognosis. TAMs, therefore, exercise antitumorfunctions linked to the M1 phenotype [25–29].

In a similar way, CD4+ T cells can also contribute totumor destruction or facilitate its development. Among thefour subpopulations of naıve CD4+ T cells, type 1 CD4+

T cells (Th1) facilitate tumor rejection by assisting in thefunction of cytotoxic CD8+ T cells whereas type 2 CD4+

T cells (Th2) promote antibody production by B cells bysecreting cytokines [30] (Figure 1). CD4+ Th17 cells, byproducing IL-17, stimulate the production of cytokines andchemokines, promoting inflammation [31] (Figure 1). Sev-eral studies have shown that CD4+ T regulatory cells (Tregs)promote tumor progression by inhibiting the functions of Tcells and natural killer (NK) cells [32, 33] (Figure 1) and thattheir accumulation is associated with a worse prognosis [34].In contrast, Salama et al. have shown that the presence ofTregs is associated with a better survival rate [35].

Myeloid-derived suppressor cells (MDSCs), which areinduced by VEGF, GM-CSF, TGF-β, IL-6, PGE2, andcyclooxygenase (COX)-2, are also implicated in tumorprogression by inhibiting the actions of CD4+ and CD8+

T cells (by the production of arginase and reactive oxygenspecies (ROS)) [30], by inducing Tregs (through IL-10 andINF-γ-dependent process) [36]. They also interact withmacrophages inducing a shift of the immunity towards atype 2 phenotype by increasing the secretion of IL-10 anddecreasing the secretion of IL-12 [37] (Figure 1).

3. Immune System and Head and Neck Cancers

It appears that the origin of head and neck cancer is linkedto environmental carcinogens (tobacco, alcohol) whereastumor progression could be linked to a failure of the immunesystem to fight against cancer. In addition to escaping theimmune system, some head and neck cancers can alsocorrupt the antitumor response via several mechanisms [38].Strategies employed by head and neck cancers are variedand can target the antigen-processing machinery (APM)via the downregulation or a loss of expression of humanleukocyte antigen (HLA) class I molecules and/or of othercomponents of the APM [39, 40]. Although effective anti-tumor immune responses likely involve many componentsof the immune system, T-cells continue to be considered asthe critical immune cells involved in antitumor immunity.The development of HNSCCs is strongly influenced bythe host immune system [38, 41–45]. Recent evidencesuggests that the antitumor responses of HNSCC patientsare compromised in the presence of functional defects orapoptosis of T-cells, both circulating and tumor-infiltrating[41–45]. Tumor-derived factors or factors produced bynormal cells in a local microenvironment favor tumors anddisable TIL. In fact, TILs look like activated T-cells but arefunctionally compromised [38]. Functional assays with TILsisolated from the tumor bed have identified a number ofdefects, including (i) absent (or low) expression of the CD3zeta chain (CD3ζ), which is the key signaling molecule in


VEGF, CSF-1

CCL2, 3, 4, 5, 8

IL-1

0,IN

F-γ

TGF-β, VEGF, GM-CSF

Costimulatorymolecules

MHC

Tumor cell

Treg

EGF, PDGF, TGF-β, IL-6, IL-1, TNF-α

Tumor growth

Invasion

Metastasis

TGF-β, VEGF, GM-CSF, TGF-α, IL-1, IL-6, IL-8 MMPs, TNF-α, IL-1

Immunosuppression

Angiogenesis

TGF-β, PGE2, IL-10MDSC

TGF-β, PGE2, IL-6, IL-10, adenosine

Immunosuppression

Apoptosis

Apoptoticcell

NK cell

InflammationIL-17 Cytokine

CytokineAntibody productionby B cells

IL-12 IL-10

Arginase/

ROS

Tumor cell

TAM M2

phenotypeTAM M2

phenotype

MDSC Treg

NK cell

Promote

Inhibit

Interact

CD8+ T cell

CD4+ Th2 T cell

CD4+ Th17 T cell

CD8+ T cell

ApoptoticCD8+ T cell

CD4+ T cell

CD4+ T cell

T cellCD4+ Th17

CD8+ T

T cellCD4+ Th2

Chemokine

IL-6, PGE2, COX-2

Figure 1: Immunosuppressive mechanisms in the tumor microenvironment: several mechanisms are developed by cancerous cells to escapeto the immune system such as a loss or a reduction of the expression of MHC class 1 molecules and costimulatory molecules, the expressionof FasL to induce apoptosis of tumor-infiltrating lymphocytes and the production of immunosuppressive molecules such as TGF-β, PGE2,IL-6, IL-10, and adenosine. Among the subpopulations of naıve CD4+ T cells, CD4+ Th17 T cells promote inflammation by secreting IL-17whereas CD4+ Th2 T cells promote antibody production by B cells. Tregs promote tumor progression by inhibiting the functions of CD4+

and CD8+ T cells and NK cells. TAMs M2 phenotype induce the expression of CD4+ Th2 T cell and Tregs. Moreover, M2 phenotype promotegrowth tumor (EGF, PDGF, TGF-β, IL-6, IL-1, and TNF-α), angiogenesis (TGF-β, VEGF, GM-CSF, TGF-α, IL-1, IL-6, and IL-8), invasion(MMPs, TNF-α, IL-1), immunosuppression (TGF-β, PGE2, and IL-10) and metastasis. MDSCs induce Treg, secrete IL-10, and inhibit CD4+

and CD8+ T cells.

the T-cell receptor pathway [38], (ii) decreased proliferationin response to mitogens or IL-2 [38], (iii) the inability tokill tumor cell targets [44, 45], (iv) an imbalance in thecytokine profile, with the striking absence of IL-2 and/orIFN-γ production [46], and (v) evidence of pronouncedapoptotic features in a considerable proportion of TILs [38,47]. Moreover, immune cell dysfunction in HNSCC patientsappears to extend far beyond the tumor microenvironmentbecause both functional defects and massive lymphocytedeath have also been observed in the peripheral circulationof patients with advanced HNSCC [48]. In addition, HNSCCcells that produce proinflammatory cytokines autonomouslyare endowed with an advantage with respect to survival andgrowth [49]. HNSCC cells also produce high quantities of

TGF-β1, which reduces the expression of NK cell receptorNKG2D and CD16 and inhibits the biological functions ofNK cells [50]. The induction of T-cell immunity followingthe vaccination of an orthotopic murine HNSCC model witha recombinant vaccinia virus expressing IL-2 induces tumor-specific CD8+ cytotoxic T cell (CTL) and CD4+ Th1-typehelper T cells [51], which are targets of the cytocidal effectsof galectin-1 secreted by cancer cells [52].

Another mechanism employed by the tumor to escapeantitumor immunity is the immunosuppressive action ofTregs. Various studies have demonstrated an increasedabundance of Tregs in the TILs and of peripheral bloodmononuclear cells in head and neck cancer patients [53](Figure 2). Head and neck cancers can also directly inhibit


the immune response by producing soluble mediators suchas VEGF, PGE2, TGF-β, IL-6, and IL-10 [40, 54]. Finally,the number of TAMs seems to be correlated to the patientprognosis, suggesting possible protumoral functions of thesecells in head and neck cancers [55] (Figure 2).

4. Disruption of the Antigen-PresentingMachinery in Head and Neck Cancers

HLAs are proteins of the MHC in humans and are presentat the surface of antigen-presenting cells (APCs). T lympho-cytes recognize antigens that are linked to these molecules.The APM is composed of β subunits of the proteolyticdelta and MB1, inducible proteasome β-type subunits LMP2,LMP7, and LMP10, peptide transporters TAP1 and TAP2,which are essential for introducing peptides into the endo-plasmic reticulum from the cytosol, and the endoplasmicreticulum chaperones calnexin, calreticulin, ERp57, andtapasin. All of these components play an important rolein the generation of antigenic peptides, their translocationinto the endoplasmic reticulum and loading of the β2-microglobulin-associated MHC class I H chain with pep-tides. These interactions induce the trafficking of MHCclass I molecules to the cell surface and the presentation ofpeptides to CD8+ T lymphocytes [56, 57].

As mentioned previously, downregulation or loss of theexpression of HLA class I molecules and/or of componentsof the APM is one of the strategies used by tumor cells toescape the immune system. Using immunohistochemistry,Ogino and co-authors observed a downregulation of HLAclass I antigen and of most APM components in a clinicalseries of 63 primary laryngeal squamous cell carcinomas.Moreover, the downregulation of HLA class I antigen and ofLMP2 (a component of the APM) associated with low CD8+

T cell infiltration were significantly associated with lowersurvival rates in these patients [58]. These observations wereconfirmed by Grandis et al., who described the loss of HLAclass I protein expression in 50% of HNSCCs. This findingwas also correlated with the presence of regional lymph nodemetastases [59]. Oral squamous cell carcinoma (OSCC)-derived gangliosides induce the downregulation of severalMHC class I APM components, suggesting that this is oneof the mechanisms used by the tumor to induce alterationsin APM components [60].

5. Dendritic Cells and Head and Neck Cancers

5.1. Dendritic Cells Functions. Dendritic cells (DCs) are afamily of specialized APCs and are essential mediators ofimmunity and tolerance [61, 62]. DCs are derived fromthe bone marrow and may have a myeloid origin (myeloiddendritic cells, MDCs) or a lymphoid origin (plasmacytoiddendritic cells, PDCs). MDCs are divided into two groups:(i) the Langerhans cells present in the epidermis and inthe mucosae of the upper aerodigestive tract and (ii) thedermal/interstitial MDCs located in the dermis [63]. PDCsare found in the blood and in the T centers of lymphoidorgans (thymus, tonsils, spleen, lymph nodes, etc.) [64]. In

nonlymphoid tissues (peripheral tissues such as skin), DCsare immature and characterized by a high capacity for anti-gen capture and processing. The presence of inflammatorymediators (IL-1, TNF-α, and IL-12) and microbial productspromotes the maturation of DCs that have lost the abilityto capture antigens and have acquired an increased capacityto present antigens and to stimulate T cells. Moreover,mature DCs upregulate costimulatory molecules such asCD40, CD80, and CD86 and cytokines such as IL-1, IL-12,and TNF-α. Mature DCs then migrate out of nonlymphoidtissues into the blood and into secondary lymphoid organs,where they present antigens captured in peripheral tissuesto T lymphocytes and stimulate T cell differentiation ineffector cells (such as cytotoxic CD8+ T cells that are ableto kill tumor cells). For these reasons, DCs can be viewedas the sentinels of the immune system [61, 65]. In contrast,immunosuppressive agents such as IL-10 and TGF-β convertimmature DCs into tolerogenic DCs that can induce antigen-specific T-cell tolerance via several mechanisms, such as acti-vation of Tregs, silencing of differentiated antigen-specific Tcell tolerance, and differentiation of naıve CD4+ T cells intoTregs [66–68]. Three main immunohistochemical markersare used to detect DCs: CD1a and S-100 for immature DCsand CD83 for mature DCs.

5.2. Langerhans Cells and Head and Neck Cancers. Langer-hans cells (LCs) are dendritic APCs located within thestratified squamous epithelium of the skin and mucosa ofthe upper aerodigestive tract. LCs are found in the suprabasallayers and constitute 2–8% of the intraepithelial cell content(Figure 2). Although observed in these epithelia, it is nowclear that LCs are a dynamic population that migrates fromthe bone marrow to the stratified squamous epithelium.Regarding their roles, LCs intercept and bind new antigensdetected in the squamous epithelium. Subsequently, theymigrate back to the regional lymph nodes and assumethe features of interdigitating dendritic cells, where theyinitiate a primary immune response by stimulating naıveT-lymphocytes. Later, when LCs meet recall antigens, theycan present antigens to memory T-lymphocytes circulatingthrough the extranodal skin and mucosa-associated lym-phoid tissue and stimulate a secondary immune responsewithin the mucosa [69]. Several molecules are sufficientlyspecific for use as LC immunohistochemical markers, suchas CD1a, S100 protein and CD207.

Tobacco and alcohol consumption, which are well-established risk factors for abnormal oral mucosal changes(metaplasia and dysplasia) and oral squamous cell carci-noma, seem to be capable of stimulating mucosal LCs. Inter-estingly, these exposures are associated with an increasednumber of oral mucosal LCs (OMLCs) [69] (Figure 2).Indeed, a greater number of CD1a+ OMLCs has beenobserved in smokers at sites that are often affected by squa-mous cell carcinoma, such as the lips and the lateral border ofthe tongue [70]. Similarly, an increase in HLA-DR+ OMLCsin the lip has been observed [71] whereas smokeless tobacco(chewing tobacco and preparations that are absorbed by theoral or nasal mucosae (snuff)) has the opposite effect [72].LC numbers were reportedly not associated with alcohol


Apo

ptos

is

Normal epithelium Preneoplastic lesion Invasive carcinomaTobacco

VEGF

VEGF

VEGF

Langerhans cell

TAM

maturedendritic cell

VEGFTreg

CD8+ T cell

CD4+ T cell

Apoptotic CD8+ T cell

Figure 2: Description of immunosuppressive mechanisms during the head and neck tumor progression: in the normal epithelia of theupper aerodigestive tracts, LCs are present in the suprabasal layers. When mucosae of these areas are exposed to tobacco, the number ofLCs increases whereas these cells decrease in invasive carcinomas. The mature DCs are prominent in the peritumoral area and correlatedpositively with the expression of VEGF. DCs are also more abundant in patients with metastasis. A higher level of TAM is observed inHNSCCs, and these cells constitute a source of VEGF which play a crucial role in angiogenesis. HNSCCs can induce the apoptosis of CD8+

T cells using the mitochondrial and/or Fas/FasL pathways. Tregs can induce apoptosis of CD8+ T cells and inhibition of the proliferation ofCD4+ T cells.

consumption, age, or sex, but alcohol consumption may actsynergistically with tobacco use [71]. Recently, Boyle and co-authors estimated the effect of tobacco on the human oralmucosal transcriptome and demonstrated an increase of LCsin the oral mucosa of smokers [73].

The presence of S100+ LC in normal mucosa, prema-lignant and malignant lesions of the oral mucosa has beeninvestigated by Girod et al. [74]. Their results showed agreater number of S100+ LCs in benign lesions than innormal mucosa. A higher LC population was also observed inthe epithelium of vocal cord polyps in comparison with thenormal vocal cord mucosa [75]. On the other hand, neoplas-tic lesions exhibited fewer S100+ cells than did benign lesions

[74] (Figure 2). In a series of oral squamous cell carcinoma,adecrease of S100+ cells was shown in high-grade compared tolow-grade tumors [76] (Figure 2). In laryngeal carcinomas,a strong infiltration of LCs was significantly associatedwith less cervical lymph node metastasis, longer disease-free survival, less locoregional recurrence and less clinicalN-positivity [77]. Other studies dedicated to nasopharynxand larynx carcinomas have shown that a greater infiltrationof LCs is correlated with a better prognosis [78, 79].Moreover, the number of S100+ LCs decreased with the lossof tumoral differentiation [74]. These observations show thatLC infiltration is prognostically important in head and neckcancers, confirming that these cells may act as important


immune factors that function as APCs in the defense againstHNSCCs.

5.3. Myeloid Dendritic Cells and Head and Neck Cancers.HNSCCs seem have a significant impact on dendritic cells(DCs). In this context, Li et al. have observed a larger numberof DCs in nonmetastatic lymph nodes than in metastaticlymph nodes in a series of hypopharyngeal and laryngealcarcinomas. The immature DC marker CD1a was especiallypresent in the cancer “nest” whereas the mature DC markerCD83 was prominent in the peritumor area [80] (Figure 2).

The relationship between the expression of VEGF, anangiogenic factor released by tumor cells, and DC infil-tration, which plays an important role in immune defenseagainst tumors, remains unclear. Therefore, several studieshave analyzed the expression of VEGF isoforms in tumors.VEGF-A and VEGF-C were increased in the tumor tissuein comparison with the normal epithelium, and VEGF-Dwas decreased in the presence of cervical nodal metastasis.VEGF-A expression correlated with microvessel density,disease progression, a reduced number of mature DCs andan increased number of immature DCs. VEGF-A is involvedin angiogenesis, tumor progression and immunosuppression[81]. Another study showed the strong expression of VEGFin oral squamous cell carcinomas from patients with regionallymph node metastasis, but in that case, the expression ofVEGF was correlated inversely with the number of CD1a+

immature DCs and positively with the number of CD83+

mature DCs (Figure 2). The authors suggested that VEGFcould inhibit the differentiation of CD1a+ immature DCsfrom progenitor cells and increase the levels of dysfunctionalCD83+ mature DCs [82]. Moreover, in oral SCCs, Kikuchiand co-authors observed a greater number of S100+ andCD1a+ immature DCs in adjacent tissue and regional lymphnodes in patients without metastasis; in contrast, CD83+

mature DCs were more abundant in patients with metastasis[83] (Figure 2).

In lip SCCs, a higher peritumoral DC density (detectedusing anti-S100 antibody) was associated with a low rateof metastasis whereas a lower peritumoral DC densitycorrelated positively with TILs. In contrast, the intratumoralDC density did not correlate with metastasis [84].

Tumor cells can modulate the expression of TLRs presenton the surface of immune cells [85]. Frenzel et al. analyzedthe influence of HNSCC on the TLR expression of MDCsoriginating from the peripheral blood. MDCs expressed allTLRs except TLR4, -9, and -7 demonstrated the strongestexpression. This finding confirms that the alteration ofTLR expression is an important tumor-promoting event inHNSCC progression [86].

HNSCCs can also influence the circulating MDC andPDC populations. So, the proportion of circulating PDCs(LIN-DR+123+) did not differ considerably in patientssuffering from HNSCC compared with the healthy sub-jects. However, the number of circulating MDCs (LIN-DR+CD11c+) was significantly lower in patients withHNSCC. In a significant number of patients, the circulatingMDC population increased after removal of the tumor,which highlights that this reduction was due to the presence

of tumor and was also reversible. This deficiency in circu-lating MDCs could contribute to tumor immune escape inHNSCC patients [87].

5.4. Plasmacytoid Dendritic Cells and Head and Neck Cancers.PDCs produce large amounts of interferon (IFN)-α inresponse to viruses, and it seems that their antigen capturepotential is less developed compared to other APC [88].Hartmann et al. studied the presence and function of PDCsin HNSCC and showed that PDCs infiltrated the tumortissue. They used oligonucleotides containing CpG motifsknown as microbial stimuli for PDCs (recognized via Toll-like receptor (TLR) 9) to study the functional capacity ofPDCs to produce IFN-α. They noticed that HNSCC PDCsdecreased IFN-α production in response to CpG motifs. Theauthors hypothesized that this decreased IFN-α productionmay be due to a tumor-induced downregulation of TLR9expression. To test this hypothesis, they determined the levelsof TLRs 1–10 in PDCs from peripheral blood in the presenceor absence of the supernatant from the HNSCC cell line PCI-1. In the absence of the PCI-1 supernatant, PDCs expressedhigh levels of TLR1, -7, and -9 and low levels of TLR6 and-10, whereas the other TLRs were at the detection limit.However, in the presence of the PCI-1 supernatant, all ofthese TLRs showed decreased expression levels. Therefore,the downregulation of TLR9 induced by HNSCC cells islikely one mechanism that contributes to the impaired PDCfunction [89].

PGE2 and TGF-β are two immunosuppressive factorsfound in tumor tissue. A recent study showed that TGF-βsynergized with PGE2 inhibited IFN-α and tumor necrosisfactor (TNF) production of TLR7- and TLR9-stimulatedPDCs [90].

6. Macrophages and Head and Neck Cancers

Macrophages migrate from the bone marrow as imma-ture monocytes, circulate in the bloodstream and finallymigrate into tissues by extravasation to undergo differ-entiation into resident macrophages, including osteoclastsin the bone, alveolar macrophages in the lung, histiocytesin the connective tissue, microglia in the neural tissue,mesangial cells in the kidney, and Kupffer cells in theliver. Macrophages participate in the innate and adaptiveimmune systems and are critical mediators of inflammatoryprocesses. They have several functions, including antigenpresentation, target cell cytotoxicity, removal of debris andtissue remodeling, regulation of inflammation, induction ofimmunity, thrombosis, and various forms of endocytosis [18,23, 91]. The main marker used in immunohistochemistry todetect macrophages of both the M1 and M2 phenotypes isCD68.

Several studies have suggested the involvement of tumor-associated macrophages (TAMs) in angiogenesis and tumorprogression of HNSCCs. In a clinical series of oral carci-nomas, the number of TAMs (detected by immunohisto-chemistry using CD68) is higher in carcinomas. A significantassociation between the expression of TAMs and stages ofinvasion, intratumoral microvessel density, and angiogenic


factors such as VEGF was also observed (Figure 2) [92].The hypothesis of the involvement of TAMs in tumoralprogression was also issued during an analysis of theexpression of cell cycle (cyclin E and p53) and proliferationmarkers (Ki67) as well as macrophage infiltration in a seriesof HNSCCs. In general, weak expression of Ki67, cyclin E,and p53 is associated with a better prognosis. Additionally, adirect correlation between the macrophage infiltration andthe tumor proliferation index was noted, which suggestedthat the number of TAMs is functionally linked to tumorprogression [93].

Extravascular fibrin deposits are frequently observed inthe tumoral and peritumoral tissue and are involved intumoral growth. In laryngeal and hypopharyngeal carci-nomas, the accumulation of macrophages (detected usinga Ki-M7 monoclonal antibody) was observed in areas offibrin deposition, which suggests that these macrophagesparticipate in the stabilization of intratumoral fibrin andfacilitate tumor matrix generation and angiogenesis [94].It is currently well accepted that the growth and spreadof solid/malignant tumors require angiogenesis, which isdescribed as the formation of new blood vessels in thetumor microenvironment. VEGF is a secreted endothelialcell-specific growth factor and is one of the most importantfactors in angiogenesis [95, 96]. Several studies have shownthat apart from tumor cells, macrophages constitute a sourceof VEGF in carcinomas, which supports the hypothesis thatmacrophages play a role in tumoral formation by contribut-ing to neovascularization [95–97] (Figure 2). Moreover, aparacrine angiogenic loop was also discovered betweenHNSCCs and macrophages. In fact, HNSCCs could attractmacrophages by secreting MCP-1 and TGF-β1. Followingactivation, macrophages secrete VEGF and IL-8, but they alsosecrete TNF-α and IL-1, which in turn stimulate tumor cellsto secrete increased levels of VEGF and IL-8 [98].

In oral SCCs, a significant correlation was observedbetween the presence of TAMs and the lymph node involve-ment and the tumor size. Hypoxia-inducible factor (HIF-1α) expression and TAMs can change cancer cell behaviorby making them more invasive and more aggressive. Thepresence of tumor cell-lined vessels, HIF-1α expressionand the high rate of TAMs could facilitate the prognosisof patients with oral squamous cell carcinoma [99]. Theimpact of TAMs on tumoral aggressiveness was previouslystudied in a series of oral cavity or oropharyngeal squamouscell carcinomas. In that study, the authors demonstrateda correlation between the aggressive behavior of HNSCCsand the level of infiltration of macrophages in the primarytumor. Indeed, the patients whose tumors showed high levelsof macrophage infiltration tended to develop lymph nodemetastasis and to present extracapsular lymph node spread[100].

7. T Cells and Head and Neck Cancer

7.1. T Cell Functions. Immature T lymphocytes derive fromstem cells of the bone marrow and mature in the thymus(primary lymphoid organ). Mature T lymphocytes leave thethymus and travel through blood and lymphoid vessels to

reach secondary lymphoid organs (lymph nodes, spleen),where they are present in a naıve state [101]. In these organs,APCs can present antigens to naıve T lymphocytes. Theactivation of T lymphocytes requires two signals: (i) thelink between MHCs from APCs and T cell receptors (TCRs)and (ii) the expression of costimulatory molecules [101].Once activated, T lymphocytes develop into effector ormemory cells. Effector cells include (i) CD4+ helper T cells,which facilitate B lymphocyte production of antibodies andphagocytes to destroy the ingested microbes and (ii) CD8+

cytotoxic T cells, which can induce cell death [101]. HelperT cells are divided into three subpopulations (Th1, Th2 andTh17), which are characterized by the secretion of variouscytokines [30]. CD4+ T cells, or Tregs, play a critical role inthe induction of tolerance to self-antigens and are dividedinto two main groups: naturally occurring regulatory Tcells (nTregs) and peripherally induced regulatory T cells(iTregs). The iTregs include Tr1 and Th3 cells [53, 102].Memory T lymphocytes are cells that are able to inducea rapid immune response in case of a second encounterwith a previous antigen. The main immunohistochemicalmarkers characterizing the various types of T lymphocytesare CD45RA for naıve T cells, CD45RO for memory T cells,CD69 for activated T cells, CD4 for helper T cells, CD8 forcytotoxic T cells, and CD25 and forkhead box p3 (Foxp3) forTregs.

7.2. Apoptosis of T Cells in Head and Neck Cancers. Severalstudies have investigated the mechanisms responsible forT cell apoptosis in patients with head and neck cancerand have demonstrated that one of these mechanismsinvolves the Fas/FasL signaling pathway. Indeed, Gastmanet al. studied the expression of FasL on the cell surface ofHNSCC cells. To demonstrate that the expression of FasLon the cell surface can lead to the T cell apoptosis, theycoincubated HNSCC cell lines with the Fas-sensitive JurkatT cell line. As a result, an apoptotic signal was inducedin lymphocytes, which suggests that the Fas/FasL pathwayis potentially immunosuppressive [103]. They also showedthat if Fas-mediated apoptosis in Jurkat cells is executed inthe presence of mitochondria-specific inhibitors or syntheticcaspase inhibitors, tumor-induced apoptosis is inhibited,suggesting that this phenomenon is significantly amplifiedby a mitochondrial loop and that tumor cells can triggercaspase-dependent apoptotic cascades in T lymphocytes[104, 105]. Once again, Hoffmann et al. showed that theFas/FasL pathway is involved in the spontaneous apoptosisof circulating Fas+ T lymphocytes [48].

In fact, other pathways are also implicated in the T cellapoptosis. Some oral squamous cell carcinoma cell linesare also able to induce Jurkat T cell apoptosis via TRAILand TNF-α [106]. Another study showed that MAGE3/6+FasL+MHC class I+ tumor-derived membranous vesiclesisolated from the serum of patients with HNSCC induceJurkat T cell apoptosis [107]. Moreover, Kim et al. observedthat FasL+ membranous vesicles induced caspase-3 cleavage,cytochrome c release, loss of mitochondrial membranepotential, and reduced TCR-ζ chain expression and thus the


mitochondrial apoptotic pathway in Jurkat and activated Tcells [108].

Some pro- and antiapoptotic proteins of the mitochon-drial pathway were analyzed in the lymphocytes of HNSCCpatients and healthy controls. A higher level of proapoptoticBax and antiapoptotic Bcl-XL was noted in CD8+ lympho-cytes, as well as a higher ratio Bax/Bcl-2 in HNSCC patientscompared with healthy controls. These results suggest theinvolvement of the mitochondrial pathway in the apoptosisof CD8+ T cells [109] (Figure 2).

Bcl-2 protein, an inhibitor of apoptosis, seems to beinvolved in the regulation of T lymphocyte apoptosis.The Bcl-2 expression in CD4+ and CD8+ T lymphocyteswas significantly higher in laryngeal cancer patients than incontrols. In carcinoma patients, Bcl-2 expression was alsohigher in CD4+ T cells than in CD8+ T cells. These resultssupport that the Bcl-2 protein could play a role in theregulation of T lymphocyte apoptosis [110].

The hypothesis that the mechanism of Treg suppressiondepends on Fas/FasL-mediated apoptosis of responder cellswas proposed by Strauss and colleagues. Using the bloodof HNSCC patients, they showed that Tregs induced Fas-mediated apoptosis in CD8+ T cells (Figure 2). In contrast,CD4+ T cells were resistant to Fas-mediated apoptosis byTregs but were able to induce Treg apoptosis in presence oflow concentrations of IL-2 [111].

CD39 and CD73 are ectonucleotidases expressed byTregs that convert ATP into immunosuppressive adenosine.The adenosinergic pathway in Treg-mediated suppressionhas also been studied in HNSCC patients. These patientsdemonstrated higher levels of CD39, CD73, and adenosinecompared with healthy controls. This overexpression couldbe involved in the observed stronger effector T cell suppres-sion [112].

Bergmann and co-authors used cell culture techniqueswith weak doses of IL-2, IL-10, and IL-15 to show thatthe tumor microenvironment generated Tr1 cells with aphenotype distinct from nTregs and that these cells abolishedautologous responders proliferation via the secretion of IL-10 and TGF-β (Figure 2). The Tr1 cell frequency and theirsuppressor functions were significantly higher in patientswith advanced HNSCC [102, 113].

7.3. T Cells and Prognosis of HNSCCs. Recently, the prog-nostic value of various tumor-infiltrating CD4+ T-cellpopulations (CD4+CD25+, CD4+CD69+, and CD4+FOXP3+

T cells) was determined in HNSCC patients [114]. Interest-ingly, a high level of CD4+CD69+ T cells was linked to abetter prognosis, and CD4+Foxp3+ T cells were positivelycorrelated with better locoregional control. In nasopharyn-geal carcinomas, the density of Foxp3+ TILs was correlatedto better overall survival and progression-free survival [115].

Moreover, a higher density of CD4+CD25+ Tregs was alsolinked to a good prognosis in HNSCCs [116]. In contrastwith previous studies, Strauss et al. showed that the pres-ence of Tregs in TILs was linked to a worse prognosisin HNSCC patients. Indeed, suppression in the tumormicroenvironment is mediated by a unique subset of

CD4+CD25highFoxp3+ Tregs that produce IL-10 and TGF-β,exerting a more suppressive effect on proliferation [117].

The tumoral infiltration of different subpopulationsof lymphocytes (CD3+, CD20+, CD43+, CD45+ RO, andCD56+) was assessed in laryngeal carcinomas. An increaseof the CD43+ subpopulation was observed in the group ofpatients presenting lymph node metastasis. In patients withadvanced carcinoma (stage IV), a correlation was establishedbetween the survival time and intensity of CD43+ and CD45+

RO lymphocyte infiltrations [118].

TCR recognizes antigens but is not able to initiate signaltransduction in T lymphocytes. To achieve this, a complexmust form between CD3 and the ζ chain linked to the TCR.The TCR-associated ζ chain functions as a transmembranesignaling molecule in lymphocytes [119]. Changes in theexpression of the ζ chain of TILs are biologically significantbecause the absence or low expression of this chain in TILs inpatients with stage III or IV HNSCC predicts a poor survivalcompared with patients expressing a normal ζ chain [120,121]. This was confirmed by other study which demonstratedthe importance of the ζ chain by showing that circulatingCD4+ and CD8+ T cells and CD3−CD56+CD16+ NK cellspresented lower expression of the ζ chain in the blood ofpatients with HNSCC in comparison to healthy controls.Additionally, the patients that presented a more aggressivetumor or that experienced a recurrence within the last 2years of the study demonstrated the lowest expression levelsof the ζ chain [122]. Reduced expression of the ζ chain wasalso noticed in laryngeal carcinomas before and after surgicaltreatment, and this reduction was not immediately restoredafter the treatment [123]. Reichert et al. studied the DCpopulation and the expression of the ζ chain in TILs in alarge series of 132 oral SCCs. A low density of DCs and absentor low expression of the ζ chain in TILs was correlated witha poor prognosis of survival and a high risk of recurrence[124].

Distel et al. tested different immunological markers usingoro- and hypopharynx carcinomas in a low-risk group of 62patients (surgery followed by radiotherapy) and in a high-risk group of 53 patients (inoperable, radiochemotherapy).The more advanced cases demonstrated higher rates of Tregsand B cells and fewer CD8+ T cells. In the low-risk group,a high concentration of CD20+ TILs was linked to a bettersurvival rate, whereas this increase was linked to a worseprognosis in the high-risk group [125].

7.4. Circulating T Cells and Head and Neck Cancers. Theperipheral blood of patients with tobacco-related oralSCChas shown significantly decreased CD3+ and CD4+ Tcells (Figure 2). Moreover, the frequency of CD3+IL-4+ andCD8+IL-4+ T cells was significantly higher and the number ofCD4+IL-2+ T cells significantly lower in these patients thanin healthy controls. In late-stage cancer, the expression of IL-2 in CD4+ and CD8+ cells was also reduced [134]. IL-18 hasalso been assessed in patients with HNSCC, and higher levelsof this cytokine seem to be produced in this type of cancer[135]. The concentrations of IL-10 were higher in patientswith nodal metastasis and in T3/T4 stage tumors compared


with patients without nodal metastasis and in T1/T2 stagetumors. These findings suggest that patients with advancedHNSCC exhibit a decreased Th1 immune response and anincreased Th2 immune response [136].

An increase of CD4+CD25+Foxp3+ Tregs has beenobserved in the peripheral blood and in the tumor siteof patients with nasopharyngeal carcinoma (Figure 2). Thisincrease is linked to an increase in the suppressive activityof these cells on the proliferation of CD4+CD25− T cells,which suggests the involvement of Tregs in the decreasedantitumor immunity of T cells [137] (Figure 2). Increasednumbers of Tregs have also been detected in HNSCCs [138].A comparison of the numbers of CD25+Foxp3+ Tregs andCD3+Foxp3+ and CD8+Foxp3+ in TILs between oral SCCsand 15 human tumor-free tonsils again revealed an increasednumber of Tregs in carcinomas whereas no significant changewas noted in the number of CD3+Foxp3+ and CD8+Foxp3+

in TILs [139]. Strauss et al. studied the expression of Tregsin the lymphocytes of the peripheral blood in HNSCCs.Interestingly, patients with no evident disease presentedmore Tregs and a stronger suppressive function than didthe patients with active disease, suggesting that oncologictherapy favors the expansion of Tregs [140].

Young et al. analyzed the immune inhibitory mediatorsreleased from cancer tissues and from the immune infiltratewithin the tumor in 219 HNSCCs and 64 metastatic lymphnodes. Tumor cells released substantial quantities of TGF-β,PGE2 and IL-10, which were associated with a decrease inCD8+ T cells within the tumor (Figure 2). GM-CSF, whichwas associated with the intratumoral presence of CD34+

cells, was also secreted. The authors suggested that HNSCCwould evade immune suppression with reduced numbers ofCD8+ T cells and reduced numbers or altered functions ofintratumoral CD4+ T cells [47].

8. Eosinophils and Head and Neck Cancers

In HNSCCs, the function of eosinophils still remains unclear.Several studies showed that eosinophils were associated witha good prognosis [126, 127]. In fact, patients with tumour-associated tissue eosinophilia (TATE) presented higher sur-vival in oral SCCs and less incidence of distant metastasisin head and neck cancer [126, 127]. On the other hand,some studies suggested that eosinophils were associated witha poor prognosis [128–131] or even no effect on tumorprogression [132, 133]. With regard to poor prognosis, ithas been shown that eosinophilic infiltration and tumorcells expressing HLA-DR antigen were correlated with anunfavorable prognosis [128]. TATE in OSCCs also reflectedstromal invasion and metastasis [130, 131].

9. Impact of Human Papillomavirus on theImmune System of Head and Neck Cancers

It has been established that tobacco consumption and alcoholabuse are significant risk factors for the development ofHNSCC but a proportion of the patients do not have theserisk factors, and therefore several studies have suggested anassociation between the development of HNSCC and viral

infection such as oncogenic (high-risk) human papillomavirus (HPV) types. The significance of hrHPV infection andits relationship with patient prognosis is still an importantmatter of debate, especially considering the contradictoryresults that are present in different studies in the literature[141, 142]. In fact, several studies have demonstrated that thepresence of HPV DNA is a favourable prognostic factor withregard to recurrence and survival [143–148]. In contrast,other studies showed that patients with hrHPV positivityhad a worse prognosis [142, 149, 150] or did not show asignificant correlation between hrHPV infection and clinicaloutcomes [151–156]. A persistent HPV infection which canlead to te development of cancer requires immune toleranceand HPV developed several mechanisms for evading thehost’s immune system such as downregulation of IFN-α andTLR9, production of TGF-β, maintenance of low viraemia,viral gene expression and viral protein synthesis are confinedto keratinocytes and the virus does not cause cell lysis andthus no inflammatory response [157]. In HNSCCs, there isan increased frequency of T cells specific for peptides derivedfrom the oncogenic HPV E7 protein in patients whosetumors expressing HPV16 in comparison with patientswhose tumors are negative for HPV or healthy volunteers.Therefore, antiviral immunity exists against E7 oncogenicprotein but these T cells are unable to eliminate the tumor.So, further studies are necessary to explain this tumor’sresistance [158, 159]. Williams et al. investigated whetherHPV-specific immune mechanisms can result in tumorclearance. For that, they examined immune-competent andimmune-incompetent mice with or without HPV. In theimmune-competent mice group, one third of the HPV+ micecleared their tumors in comparaison with none of the miceHPV−. Moreover, mice HPV+ had a significantly longersurvival than mice HPV−. In the mice group lacking B- andT-cell immunity, there was no difference in growth patternor survival between HPV+ and HPV− group. Therefore,the difference between HPV+ and HPV− mice is immunemediated. CD4+ and CD8+ T cells were found to be requiredto mount this immune response. They also showed thatlymphocytes from mice that cleared their tumor can conferprotective tumor immunity to immunoincompetent animals[160].

10. Conclusions

A better understanding of the factors that cause an immunesuppression in HNSCCs might be relevant for the devel-opment of new therapeutic or prophylactic anticancerapproaches. The worse prognosis of these cancers mustcertainly be link to the fact that HNSCCs strongly influ-ence the host immune system. Antitumor responses ofHNSCC patients are caused by the presence of functionaldefects or apoptosis of T-cells, both circulating and tumor-infiltrating. Langerhans cells are increased in benign tumorsbut decreased in invasive carcinomas. The accumulation ofmacrophages in the peritumoral areas seems to play a crucialrole in the neoangiogenesis by secreting VEGF.


Acknowledgment

A. Duray is a Ph.D. student supported by a grant from theFNRS (Bourse Televie).

References

[1] J. R. Grandis, J. A. Pietenpol, J. S. Greenberger, R. A. Pelroy,and S. Mohla, “Head and neck cancer: meeting summary andresearch opportunities,” Cancer Research, vol. 64, no. 21, pp.8126–8129, 2004.

[2] J. P. Shah and S. G. Patel, Head and Neck Surgery andOncology, Mosby, New York, NY, USA, 2003.

[3] A. A. Forastiere and A. Trotti, “Radiotherapy and concurrentchemotherapy: a strategy that improves locoregional controland survival in oropharyngeal cancer,” Journal of the NationalCancer Institute, vol. 91, no. 24, pp. 2065–2066, 1999.

[4] F. Denis, P. Garaud, E. Bardet et al., “Final results of the 94-01 French head and neck oncology and radiotherapy grouprandomized trial comparing radiotherapy alone with con-comitant radiochemotherapy in advanced-stage oropharynxcarcinoma,” Journal of Clinical Oncology, vol. 22, no. 1, pp.69–76, 2004.

[5] D. J. Adelstein, Y. Li, G. L. Adams et al., “An intergroupphase III comparison of standard radiation therapy and twoschedules of concurrent chemoradiotherapy in patients withunresectable squamous cell head and neck cancer,” Journal ofClinical Oncology, vol. 21, no. 1, pp. 92–98, 2003.

[6] M. Al-Sarraf, M. LeBlanc, P. G. S. Giri et al., “Chemora-diotherapy versus radiotherapy in patients with advancednasopharyngeal cancer: phase III randomized Intergroupstudy 0099,” Journal of Clinical Oncology, vol. 16, no. 4, pp.1310–1317, 1998.

[7] B. Jeremic, Y. Shibamoto, B. Milicic et al., “Hyperfractionatedradiation therapy with or without concurrent low-dose dailycisplatin in locally advanced squamous cell carcinoma of thehead and neck: a prospective randomized trial,” Journal ofClinical Oncology, vol. 18, no. 7, pp. 1458–1464, 2000.

[8] J. S. Cooper, T. F. Pajak, A. A. Forastiere et al., “Postoperativeconcurrent radiotherapy and chemotherapy for high-risksquamous-cell carcinoma of the head and neck,” NewEngland Journal of Medicine, vol. 350, no. 19, pp. 1937–1944,2004.

[9] J. Bernier, C. Domenge, M. Ozsahin et al., “Postoperativeirradiation with or without concomitant chemotherapy forlocally advanced head and neck advanced head and neckcancer,” New England Journal of Medicine, vol. 350, no. 19,pp. 1945–1952, 2004.

[10] K. Lang, J. Menzin, C. C. Earle, J. Jacobson, and M. A. Hsu,“The economic cost of squamous cell cancer of the head andneck: findings from linked SEER-medicare data,” Archives ofOtolaryngology—Head and Neck Surgery, vol. 130, no. 11, pp.1269–1275, 2004.

[11] J. B. Swann and M. J. Smyth, “Immune surveillance oftumors,” Journal of Clinical Investigation, vol. 117, no. 5, pp.1137–1146, 2007.

[12] G. Gorochov and T. Papo, Immunologie, Doin GroupeLiaisons, 2000.

[13] R. T. Costello, J. A. Gastaut, and D. Olive, “Tumor escapefrom mmune surveillance,” Archivum Immunologiae et Ther-apiae Experimentalis, vol. 47, no. 2, pp. 83–88, 1999.

[14] T. L. Whiteside, “Tumor-induced death of immune cells: itsmechanisms and consequences,” Seminars in Cancer Biology,vol. 12, no. 1, pp. 43–50, 2002.

[15] D. W. Hoskin, J. S. Mader, S. J. Furlong, D. M. Conrad, and J.Blay, “Inhibition of T cell and natural killer cell function byadenosine and its contribution to immune evasion by tumorcells,” International Journal of Oncology, vol. 32, no. 3, pp.527–535, 2008.

[16] C. E. Lewis and J. W. Pollard, “Distinct role of macrophagesin different tumor microenvironments,” Cancer Research, vol.66, no. 2, pp. 605–612, 2006.

[17] C. Lamagna, M. Aurrand-Lions, and B. A. Imhof, “Dual roleof macrophages in tumor growth and angiogenesis,” Journalof Leukocyte Biology, vol. 80, no. 4, pp. 705–713, 2006.

[18] S. B. Coffelt, R. Hughes, and C. E. Lewis, “Tumor-associatedmacrophages: effectors of angiogenesis and tumor progres-sion,” Biochimica et Biophysica Acta, vol. 1796, no. 1, pp. 11–18, 2009.

[19] A. Mantovani, T. Schioppa, C. Porta, P. Allavena, andA. Sica, “Role of tumor-associated macrophages in tumorprogression and invasion,” Cancer and Metastasis Reviews,vol. 25, no. 3, pp. 315–322, 2006.

[20] A. Sica, T. Schioppa, A. Mantovani, and P. Allavena,“Tumour-associated macrophages are a distinct M2 polarisedpopulation promoting tumour progression: potential targetsof anti-cancer therapy,” European Journal of Cancer, vol. 42,no. 6, pp. 717–727, 2006.

[21] J. Condeelis and J. W. Pollard, “Macrophages: obligatepartners for tumor cell migration, invasion, and metastasis,”Cell, vol. 124, no. 2, pp. 263–266, 2006.

[22] J.-Y. Shih, A. Yuan, J. J.-W. Chen, and P.-C. Yang, “Tumor-associated macrophages: its role in cancer invasion andmetastasis,” Journal of Cancer Molecules, vol. 2, no. 3, pp. 101–106, 2006.

[23] K. S. Siveen and G. Kuttan, “Role of macrophages in tumourprogression,” Immunology Letters, vol. 123, no. 2, pp. 97–102,2009.

[24] B. Z. Qian and J. W. Pollard, “Macrophage diversity enhancestumor progression and metastasis,” Cell, vol. 141, no. 1, pp.39–51, 2010.

[25] K. Tsung, J. P. Dolan, Y. L. Tsung, and J. A. Norton,“Macrophages as effector cells in interleukin 12-induced Tcell-dependent tumor rejection,” Cancer Research, vol. 62, no.17, pp. 5069–5075, 2002.

[26] C. Brigati, D. M. Noonan, A. Albini, and R. Benelli, “Tumorsand inflammatory infiltrates: friends or foes?” Clinical andExperimental Metastasis, vol. 19, no. 3, pp. 247–258, 2002.

[27] L. Bingle, N. J. Brown, and C. E. Lewis, “The role of tumour-associated macrophages in tumour progression: implicationsfor new anticancer therapies,” Journal of Pathology, vol. 196,no. 3, pp. 254–265, 2002.

[28] R. D. Leek and A. L. Harris, “Tumor-associated macrophagesin breast cancer,” Journal of Mammary Gland Biology andNeoplasia, vol. 7, no. 2, pp. 177–189, 2002.

[29] G. J. Van Der Bij, S. J. Oosterling, S. Meijer, R. H. J. Beelen,and M. Van Egmond, “The role of macrophages in tumordevelopment,” Cellular Oncology, vol. 27, no. 4, pp. 203–213,2005.

[30] S. Ostrand-Rosenberg, “Immune surveillance: a balancebetween protumor and antitumor immunity,” Current Opin-ion in Genetics and Development, vol. 18, no. 1, pp. 11–18,2008.


[31] Y. Bi, G. Liu, and R. Yang, “Th17 cell induction and immuneregulatory effects,” Journal of Cellular Physiology, vol. 211, no.2, pp. 273–278, 2007.

[32] M. Viguier, F. Lemaıtre, O. Verola et al., “Foxp3 expressingCD4+CD25high regulatory T cells are overrepresented inhuman metastatic melanoma lymph nodes and inhibit thefunction of infiltrating T cells,” Journal of Immunology, vol.173, no. 2, pp. 1444–1453, 2004.

[33] Z. Z. Yang and S. M. Ansell, “The role of T cells in the cancerimmunological response,” American Journal of Immunology,vol. 5, no. 1, pp. 17–28, 2009.

[34] T. J. Curiel, G. Coukos, L. Zou et al., “Specific recruitmentof regulatory T cells in ovarian carcinoma fosters immuneprivilege and predicts reduced survival,” Nature Medicine,vol. 10, no. 9, pp. 942–949, 2004.

[35] P. Salama, M. Phillips, F. Grieu et al., “Tumor-infiltratingFoxp3+ T regulatory cells show strong prognostic significancein colorectal cancer,” Journal of Clinical Oncology, vol. 27, no.2, pp. 186–192, 2009.

[36] B. Huang, P. Y. Pan, Q. Li et al., “Gr-1+CD115+ immaturemyeloid suppressor cells mediate the development of tumor-induced T regulatory cells and T-cell anergy in tumor-bearing host,” Cancer Research, vol. 66, no. 2, pp. 1123–1131,2006.

[37] P. Sinha, V. K. Clements, S. K. Bunt, S. M. Albelda, andS. Ostrand-Rosenberg, “Cross-talk between myeloid-derivedsuppressor cells and macrophages subverts tumor immunitytoward a type 2 response,” Journal of Immunology, vol. 179,no. 2, pp. 977–983, 2007.

[38] T. L. Whiteside, “Immunobiology of head and neck cancer,”Cancer and Metastasis Reviews, vol. 24, no. 1, pp. 95–105,2005.

[39] C. Badoual, F. Sandoval, H. Pere et al., “Better understandingtumor-host interaction in head and neck cancer to improvethe design and development of immunotherapeutic strate-gies,” Head and Neck, vol. 32, no. 7, pp. 946–958, 2010.

[40] M. R. I. Young, “Protective mechanisms of head and necksquamous cell carcinomas from immune assault,” Head andNeck, vol. 28, no. 5, pp. 462–470, 2006.

[41] R. L. Ferris, J. L. Hunt, and S. Ferrone, “Human leukocyteantigen (HLA) class I defects in head and neck cancer:molecular mechanisms and clinical significance,” Immuno-logic Research, vol. 33, no. 2, pp. 113–133, 2005.


[43] A. Lopez-Albaitero, J. V. Nayak, T. Ogino et al., “Role ofantigen-processing machinery in the in vitro resistance ofsquamous cell carcinoma of the head and neck cells torecognition by CTL,” Journal of Immunology, vol. 176, no. 6,pp. 3402–3409, 2006.

[44] B. Hathaway, D. P. Landsittel, W. Gooding et al., “Multiplexedanalysis of serum cytokines as biomarkers in squamous cellcarcinoma of the head and neck patients,” Laryngoscope, vol.115, no. 3, pp. 522–527, 2005.

[45] T. K. Hoffmann, H. Bier, and T. L. Whiteside, “Targeting theimmune system: novel therapeutic approaches in squamouscell carcinoma of the head and neck,” Cancer Immunology,Immunotherapy, vol. 53, no. 12, pp. 1055–1067, 2004.

[46] T. E. Reichert, H. Rabinowich, J. T. Johnson, and T.L. Whiteside, “Mechanisms responsible for signaling andfunctional defects,” Journal of Immunotherapy, vol. 21, no. 4,pp. 295–306, 1998.

[47] M. R. I. Young, M. A. Wright, Y. Lozano, J. P. Matthews,J. Benefield, and M. M. Prechel, “Mechanisms of immunesuppression in patients with head and neck cancer: influenceon the immune infiltrate of the cancer,” International Journalof Cancer, vol. 67, no. 3, pp. 333–338, 1996.

[48] T. K. Hoffmann, G. Dworacki, T. Tsukihiro et al., “Sponta-neous apoptosis of circulating T lymphocytes in patients withhead and neck cancer and its clinical importance,” ClinicalCancer Research, vol. 8, no. 8, pp. 2553–2562, 2002.

[49] T. Saito, I. Kuss, G. Dworacki, W. Gooding, J. T. Johnson, andT. L. Whiteside, “Spontaneous ex vivo apoptosis of peripheralblood mononuclear cells in patients with head and neckcancer,” Clinical Cancer Research, vol. 5, no. 6, pp. 1263–1273,1999.

[50] D. C. Duffey, Z. Chen, G. Dong et al., “Expression of adominant-negative mutant inhibitor-κBα of nuclear factor-κB in human head and neck squamous cell carcinomainhibits survival, proinflammatory cytokine expression, andtumor growth in vivo,” Cancer Research, vol. 59, no. 14, pp.3468–3474, 1999.

[51] S. Dasgupta, M. Bhattacharya-Chatterjee, B. W. O’Malley Jr.,and S. K. Chatterjee, “Recombinant vaccinia virus expressinginterleukin-2 invokes anti-tumor cellular immunity in anorthotopic murine model of head and neck squamous cellcarcinoma,” Molecular Therapy, vol. 13, no. 1, pp. 183–193,2006.

[52] S. Saussez, I. Camby, G. Toubeau, and R. Kiss, “Galectinsas modulators of tumor progression in head and necksquamous cell carcinomas,” Head and Neck, vol. 29, no. 9,pp. 874–884, 2007.

[53] O. Alhamarneh, S. M. P. Amarnath, N. D. Stafford, and J.Greenman, “Regulatory T cells: what role do they play inantitumor immunity in patients with head and neck cancer?”Head and Neck, vol. 30, no. 2, pp. 251–261, 2008.

[54] A. Jewett, C. Head, and N. A. Cacalano, “Emerging mech-anisms of immunosuppression in oral cancers,” Journal ofDental Research, vol. 85, no. 12, pp. 1061–1073, 2006.

[55] K. W. Kross, J. H. Heimdal, and H. J. Aarstad, “Mononuclearphagocytes in head and neck squamous cell carcinoma,”European Archives of Oto-Rhino-Laryngology, vol. 267, no. 3,pp. 335–344, 2010.

[56] A. N. Antoniou, S. J. Powis, and T. Elliott, “Assembly andexport of MHC class I peptide ligands,” Current Opinion inImmunology, vol. 15, no. 1, pp. 75–81, 2003.

[57] R. L. Ferris, T. L. Whiteside, and S. Ferrone, “Immune escapeassociated with functional defects in antigen-processingmachinery in head and neck cancer,” Clinical CancerResearch, vol. 12, no. 13, pp. 3890–3895, 2006.

[58] T. Ogino, H. Shigyo, H. Ishii et al., “HLA class I antigendown-regulation in primary laryngeal squamous cell carci-noma lesions as a poor prognostic marker,” Cancer Research,vol. 66, no. 18, pp. 9281–9289, 2006.

[59] J. R. Grandis, D. M. Falkner, M. F. Melhem, W. E. Gooding,S. D. Drenning, and P. A. Morel, “Human leukocyte antigenclass I allelic and haplotype loss in squamous cell carcinomaof the head and neck: clinical and immunogenetic conse-quences,” Clinical Cancer Research, vol. 6, no. 7, pp. 2794–2802, 2000.

[60] I. L. Tourkova, G. V. Shurin, G. S. Chatta et al., “Restorationby IL-15 of MHC class I antigen-processing machinery inhuman dendritic cells inhibited by tumor-derived ganglio-sides,” Journal of Immunology, vol. 175, no. 5, pp. 3045–3052,2005.


[61] K. Steinbrink, K. Mahnke, S. Grabbe, A. H. Enk, and H.Jonuleit, “Myeloid dendritic cell: from sentinel of immunityto key player of peripheral tolerance?” Human Immunology,vol. 70, no. 5, pp. 289–293, 2009.

[62] K. Liu and M. C. Nussenzweig, “Origin and development ofdendritic cells,” Immunological Reviews, vol. 234, no. 1, pp.45–54, 2010.

[63] J. Valladeau and S. Saeland, “Cutaneous dendritic cells,”Seminars in Immunology, vol. 17, no. 4, pp. 273–283, 2005.

[64] J. Charles, L. Chaperot, D. Salameire et al., “Plasmacytoiddendritic cells and dermatological disorders: focus on theirrole in autoimmunity and cancer,” European Journal ofDermatology, vol. 20, no. 1, pp. 16–23, 2010.

[65] M. Cella, F. Sallusto, and A. Lanzavecchia, “Origin, matu-ration and antigen presenting function of dendritic cells,”Current Opinion in Immunology, vol. 9, no. 1, pp. 10–16,1997.

[66] D. Hawiger, K. Inaba, Y. Dorsett et al., “Dendritic cellsinduce peripheral T cell unresponsiveness under steady stateconditions in vivo,” Journal of Experimental Medicine, vol.194, no. 6, pp. 769–779, 2001.

[67] S. Yamazaki, K. Inaba, K. V. Tarbell, and R. M. Steinman,“Dendritic cells expand antigen-specific Foxp3+CD25+CD4+

regulatory T cells including suppressors of alloreactivity,”Immunological Reviews, vol. 212, pp. 314–329, 2006.

[68] H. Jonuleit, E. Schmitt, G. Schuler, J. Knop, and A. H. Enk,“Induction of interleukin 10-producing, nonproliferatingCD4+ T cells with regulatory properties by repetitive stim-ulation with allogeneic immature human dendritic cells,”Journal of Experimental Medicine, vol. 192, no. 9, pp. 1213–1222, 2000.

[69] A. W. Barrett, A. T. Cruchley, and D. M. Williams, “Oralmucosal Langerhans’ cells,” Critical Reviews in Oral Biologyand Medicine, vol. 7, no. 1, pp. 36–58, 1996.

[70] A. T. Cruchley, D. M. Williams, P. M. Farthing, P. M. Speight,C. A. Lesch, and C. A. Squier, “Langerhans cell densityin normal human oral mucosa and skin: relationship toage, smoking and alcohol consumption,” Journal of OralPathology and Medicine, vol. 23, no. 2, pp. 55–59, 1994.

[71] A. W. Barett, D. M. Williams, and J. Scott, “Effect of tobaccoand alcohol consumption on the Langerhans cell populationof human lingual epithelium determined using a monoclonalantibody against HLADR,” Journal of Oral Pathology andMedicine, vol. 20, no. 2, pp. 49–52, 1991.

[72] T. E. Daniels, L. Chou, J. S. Greenspan et al., “Reductionof Langerhans cells in smokeless tobacco-associated oralmucosal lesions,” Journal of Oral Pathology and Medicine, vol.21, no. 3, pp. 100–104, 1992.

[73] J. O. Boyle, Z. H. Gumus, A. Kacker et al., “Effects of cigarettesmoke on the human oral mucosal transcriptome,” CancerPrevention Research, vol. 3, no. 3, pp. 266–278, 2010.

[74] S. C. Girod, T. Kuhnast, S. Ulrich, and G. R. F. Krueger,“Langerhans cells in epithelial tumors and benign lesions ofthe oropharynx,” In Vivo, vol. 8, no. 4, pp. 543–547, 1994.

[75] A. C. Thompson and N. R. Griffin, “Langerhans cells innormal and pathological vocal cord mucosa,” Acta Oto-Laryngologica, vol. 115, no. 6, pp. 830–832, 1995.

[76] R. L. C. Albuquerque Jr., M. C. C. Miguel, A. L. L.Costa, and L. B. Souza, “Correlation of c-erbB-2 and S-100expression with the malignancy grading and anatomical sitein oral squamous cell carcinoma,” International Journal ofExperimental Pathology, vol. 84, no. 6, pp. 259–265, 2003.

[77] T. Yilmaz, G. Gedikoglu, A. Celik, M. Onerci, and E. Turan,“Prognostic significance of Langerhans cell infiltration in

cancer of the larynx,” Otolaryngology—Head and NeckSurgery, vol. 132, no. 2, pp. 309–316, 2005.

[78] C. X. Ma, T. C. Jia, X. R. Li, Z. F. Zhand, and C. B. Yiao,“Langerhans cells in nasopharyngeal carcinoma in relationto prognosis,” In Vivo, vol. 9, no. 3, pp. 225–229, 1995.

[79] O. Gallo, G. A. Libonati, E. Gallina et al., “Langerhans cellsrelated to prognosis in patients with laryngeal carcinoma,”Archives of Otolaryngology—Head and Neck Surgery, vol. 117,no. 9, pp. 1007–1010, 1991.

[80] X. Li, Y. Takahashi, K. Sakamoto, and T. Nakashima,“Expression of dendritic cell phenotypic antigens in cervicallymph nodes of patients with hypopharyngeal and laryngealcarcinoma,” The Journal of Laryngology and Otology. Supple-ment, no. 31, pp. 5–10, 2009.

[81] L. Strauss, D. Volland, M. Kunkel, and T. E. Reichert, “Dualrole of VEGF family members in the pathogenesis of headand neck cancer (HNSCC): possible link between angiogen-esis and immune tolerance,” Medical Science Monitor, vol. 11,no. 8, pp. 280–292, 2005.

[82] K. Kikuchi, K. Kusama, M. Sano et al., “Vascular endothelialgrowth factor and dendritic cells in human squamous cellcarcinoma of the oral cavity,” Anticancer Research, vol. 26, no.3A, pp. 1833–1848, 2006.

[83] K. Kikuchi, K. Kusama, K. Taguchi et al., “Dendritic cellsin human squamous cell carcinoma of the oral cavity,”Anticancer Research, vol. 22, no. 2A, pp. 545–557, 2002.

[84] N. Wei and S. R. Tahan, “S100+ cell response to squamouscell carcinoma of the lip: inverse correlation with metastasis,”Journal of Cutaneous Pathology, vol. 25, no. 9, pp. 463–468,1998.

[85] R. Pries, S. Wulff, and B. Wollenberg, “Toll-like receptormodulation in head and neck cancer,” Critical Reviews inImmunology, vol. 28, no. 3, pp. 201–213, 2008.

[86] H. Frenzel, B. Hoffmann, C. Brocks, P. Schlenke, R. Pries, andB. Wollenberg, “Toll-like receptor interference in myeloiddendritic cells through head and neck cancer,” AnticancerResearch, vol. 26, no. 6B, pp. 4409–4413, 2006.

[87] T. K. Hoffmann, J. Muller-Berghaus, R. L. Ferris, J. T.Johnson, W. J. Storkus, and T. L. Whiteside, “Alterationsin the frequency of dendritic cell subsets in the peripheralcirculation of patients with squamous cell carcinomas of thehead and neck,” Clinical Cancer Research, vol. 8, no. 6, pp.1787–1793, 2002.

[88] M. Colonna, G. Trinchieri, and Y. J. Liu, “Plasmacytoiddendritic cells in immunity,” Nature Immunology, vol. 5, no.12, pp. 1219–1226, 2004.

[89] E. Hartmann, B. Wollenberg, S. Rothenfusser et al., “Iden-tification and functional analysis of tumor-infiltrating plas-macytoid dendritic cells in head and neck cancer,” CancerResearch, vol. 63, no. 19, pp. 6478–6487, 2003.

[90] I. Bekeredjian-Ding, M. Schafer, E. Hartmann et al.,“Tumour-derived prostaglandin E and transforming growthfactor-β synergize to inhibit plasmacytoid dendritic cell-derived interferon-α,” Immunology, vol. 128, no. 3, pp. 439–450, 2009.

[91] M. J. Auger and J. A. Ross, “The bilogy of macrophage,” inThe Macrophage, C. E. Lewis and J. O. McGee, Eds., pp. 2–74,Oxford University Press, Oxford, UK, 1992.

[92] C. Li, S. Shintani, N. Terakado, K. I. Nakashiro, and H.Hamakawa, “Infiltration of tumor-associated macrophagesin human oral squamous cell carcinoma,” Oncology Reports,vol. 9, no. 6, pp. 1219–1223, 2002.

[93] M. Ritta, M. De Andrea, M. Mondini et al., “Cell cycle andviral and immunologic profiles of head and neck squamous


cell carcinoma as predictable variables of tumor progression,”Head and Neck, vol. 31, no. 3, pp. 318–327, 2009.

[94] H. Bardos, A. Juhasz, G. Repassy, and R. Adany, “Fibrindeposition in squamous cell carcinomas of the larynx andhypopharynx,” Thrombosis and Haemostasis, vol. 80, no. 5,pp. 767–772, 1998.

[95] P. Salven, P. Heikkila, A. Anttonen, M. Kajanti, and H.Joensuu, “Vascular endothelial growth factor in squamouscell head and neck carcinoma: expression and prognosticsignificance,” Modern Pathology, vol. 10, no. 11, pp. 1128–1133, 1997.

[96] C. Neuchrist, C. Quint, A. Pammer, and M. Burian, “Vascularendothelial growth factor (VEGF) and microvessel density insquamous cell carcinomas of the larynx: an immunohisto-chemical study,” Acta Oto-Laryngologica, vol. 119, no. 6, pp.732–738, 1999.

[97] R. V. Lalla, D. S. Boisoneau, J. D. Spiro, and D. L. Kreutzer,“Expression of vascular endothelial growth factor receptorson tumor cells in head and neck squamous cell carcinoma,”Archives of Otolaryngology—Head and Neck Surgery, vol. 129,no. 8, pp. 882–888, 2003.

[98] C. Liss, M. J. Fekete, R. Hasina, C. Diemthuy Lam, and M. W.Lingen, “Paracrine angiogenic loop between head-and-necksquamous-cell carcinomas and macrophages,” InternationalJournal of Cancer, vol. 93, no. 6, pp. 781–785, 2001.

[99] S. Y. Liu, L. C. Chang, L. F. Pan, Y. J. Hung, C. H. Lee, andY. S. Shieh, “Clinicopathologic significance of tumor cell-lined vessel and microenvironment in oral squamous cellcarcinoma,” Oral Oncology, vol. 44, no. 3, pp. 277–285, 2008.

[100] B. Marcus, D. Arenberg, J. Lee et al., “Prognostic factors inoral cavity and oropharyngeal squamous cell carcinoma: theimpact of tumor-associated macrophages,” Cancer, vol. 101,no. 12, pp. 2779–2787, 2004.

[101] A. K. Abbas and A. H. Lichtman, Les Bases de L’immunologieFondamentale et Clinique, Elsevier, 2005.

[102] C. Bergmann, L. Strauss, R. Zeidler, S. Lang, and T. L.Whiteside, “Expansion and characteristics of human Tregulatory type 1 cells in co-cultures simulating tumormicroenvironment,” Cancer Immunology, Immunotherapy,vol. 56, no. 9, pp. 1429–1442, 2007.

[103] B. R. Gastman, Y. Atarashi, T. E. Reichert et al., “Fas ligandis expressed on human squamous cell carcinomas of thehead and neck, and it promotes apoptosis of T lymphocytes,”Cancer Research, vol. 59, no. 20, pp. 5356–5364, 1999.

[104] B. R. Gastman, D. E. Johnson, T. L. Whiteside, and H.Rabinowich, “Tumor-induced apoptosis of T lymphocytes:elucidation of intracellular apoptotic events,” Blood, vol. 95,no. 6, pp. 2015–2023, 2000.

[105] B. R. Gastman, X. M. Yin, D. E. Johnson et al., “Tumor-induced apoptosis of T cells: amplification by a mitochon-drial cascade,” Cancer Research, vol. 60, no. 24, pp. 6811–6817, 2000.

[106] N. Kassouf and M. H. Thornhill, “Oral cancer cell lines canuse multiple ligands, including Fas-L, TRAIL and TNF-α, toinduce apoptosis in Jurkat T cells: possible mechanisms forimmune escape by head and neck cancers,” Oral Oncology,vol. 44, no. 7, pp. 672–682, 2008.

[107] C. Bergmann, L. Strauss, E. Wieckowski et al., “Tumor-derived microvesicles in sera of patients with head and neckcancer and their role in tumor progression,” Head and Neck,vol. 31, no. 3, pp. 371–380, 2009.

[108] J. W. Kim, E. Wieckowski, D. D. Taylor, T. E. Reichert, S.Watkins, and T. L. Whiteside, “Fas ligand-positive membra-nous vesicles isolated from sera of patients with oral cancer

induce apoptosis of activated T lymphocytes,” Clinical CancerResearch, vol. 11, no. 3, pp. 1010–1020, 2005.

[109] J. W. Kim, T. Tsukishiro, J. T. Johnson, and T. L. Whiteside,“Expression of pro- and antiapoptotic proteins in circulatingCD8+ T cells of patients with squamous cell carcinoma of thehead and neck,” Clinical Cancer Research, vol. 10, no. 15, pp.5101–5110, 2004.

[110] J. Klatka, J. Rolinski, K. Kupisz, S. Klonowski, and D. Skomra,“Expression of bcl-2 protein in lymphocytes of patientswith laryngeal carcinoma,” European Archives of Oto-Rhino-Laryngology, vol. 256, no. 6, pp. 299–302, 1999.

[111] L. Strauss, C. Bergmann, and T. L. Whiteside, “Humancirculating CD4+CD25highFoxp3+ regulatory T cells killautologous CD8+ but not CD4+ responder cells by Fas-mediated apoptosis,” Journal of Immunology, vol. 182, no. 3,pp. 1469–1480, 2009.

[112] M. Mandapathil, M. J. Szczepanski, M. Szajnik et al.,“Increased ectonucleotidase expression and activity in regu-latory T cells of patients with head and neck cancer,” ClinicalCancer Research, vol. 15, no. 20, pp. 6348–6357, 2009.

[113] C. Bergmann, L. Strauss, Y. Wang et al., “T regulatorytype 1 cells in squamous cell carcinoma of the head andneck: mechanisms of suppression and expansion in advanceddisease,” Clinical Cancer Research, vol. 14, no. 12, pp. 3706–3715, 2008.

[114] C. Badoual, S. Hans, J. Rodriguez et al., “Prognostic value oftumor-infiltrating CD4+ T-cell subpopulations in head andneck cancers,” Clinical Cancer Research, vol. 12, no. 2, pp.465–472, 2006.

[115] Y. L. Zhang, J. Li, H. Y. Mo et al., “Different subsets of tumorinfiltrating lymphocytes correlate with NPC progression indifferent ways,” Molecular Cancer, vol. 9, article 4, 2010.

[116] D. Loose, A. Signore, E. Bonanno et al., “Prognostic valueof CD25 expression on lymphocytes and tumor cells insquamous-cell carcinoma of the head and neck,” CancerBiotherapy and Radiopharmaceuticals, vol. 23, no. 1, pp. 25–33, 2008.

[117] L. Strauss, C. Bergmann, M. Szczepanski, W. Gooding,J. T. Johnson, and T. L. Whiteside, “A unique subset ofCD4+CD25high Foxp3+ T cells secreting interleukin-10 andtransforming growth factor-β1 mediates suppression in thetumor microenvironment,” Clinical Cancer Research, vol. 13,no. 15, pp. 4345–4354, 2007.

[118] A. Gabriel, G. Namysłowski, A. Ziołkowski, K. Morawski, K.Steplewska-Mazur, and P. Urbaniec, “Immunohistochemicalanalysis of lymphocytic infiltration in the tumor microen-vironment in patients operated on for laryngeal cancer,”European Archives of Oto-Rhino-Laryngology, vol. 256, no. 8,pp. 384–387, 1999.

[119] T. L. Whiteside, “Down-regulation of ζ-chain expressionin T cells: a biomarker of prognosis in cancer?” CancerImmunology, Immunotherapy, vol. 53, no. 10, pp. 865–878,2004.

[120] T. E. Reichert, R. Day, E. M. Wagner, and T. L. Whiteside,“Absent or low expression of the ζ chain in T cells at thetumor site correlates with poor survival in patients with oralcarcinoma,” Cancer Research, vol. 58, no. 23, pp. 5344–5347,1998.

[121] T. L. Whiteside, “Signaling defects in T lymphocytes ofpatients with malignancy,” Cancer Immunology Immunother-apy, vol. 48, no. 7, pp. 346–352, 1999.

[122] I. Kuss, T. Saito, J. T. Johnson, and T. L. Whiteside, “Clinicalsignificance of decreased ζ chain expression in peripheral


blood lymphocytes of patients with head and neck cancer,”Clinical Cancer Research, vol. 5, no. 2, pp. 329–334, 1999.

[123] L. Pignataro, D. Pagani, B. Brando, G. Sambataro, B. Scarpati,and M. M. Corsi, “Down-regulation of ζ chain and zeta-associated protein 70 (Zap 70) expression in circulatingT lymphocytes in laryngeal squamous cell carcinoma,”Analytical and Quantitative Cytology and Histology, vol. 29,no. 1, pp. 57–62, 2007.

[124] T. E. Reichert, C. Scheuer, R. Day, W. Wagner, and T. L.Whiteside, “The number of intratumoral dendritic cells andζ-chain expression in T cells as prognostic and survivalbiomarkers in patients with oral carcinoma,” Cancer, vol. 91,no. 11, pp. 2136–2147, 2001.

[125] L. V. Distel, R. Fickenscher, K. Dietel et al., “Tumourinfiltrating lymphocytes in squamous cell carcinoma of theoro- and hypopharynx: prognostic impact may depend ontype of treatment and stage of disease,” Oral Oncology, vol.45, no. 10, pp. e167–e174, 2009.

[126] R. G. Dorta, G. Landman, L. P. Kowalski, J. R. P. Lauris,M. R. D. O. Latorre, and D. T. Oliveira, “Tumour-associatedtissue eosinophilia as a prognostic factor in oral squamouscell carcinomas,” Histopathology, vol. 41, no. 2, pp. 152–157,2002.

[127] M. M. Goldsmith, D. A. Belchis, D. H. Cresson, W. D. Merritt,and F. B. Askin, “The importance of the eosinophil in headand neck cancer,” Otolaryngology—Head and Neck Surgery,vol. 106, no. 1, pp. 27–33, 1992.

[128] K. Horiuchi, K. Mishima, M. Ohsawa, M. Sugimura, and K.Aozasa, “Prognostic factors for well-differentiated squamouscell carcinoma in the oral cavity with emphasis on immuno-histochemical evaluation,” Journal of Surgical Oncology, vol.53, no. 2, pp. 92–96, 1993.

[129] M. Said, S. Wiseman, J. Yang et al., “Tissue eosinophilia:a morphologic marker for assessing stromal invasion inlaryngeal squamous neoplasms,” BMC Clinical Pathology,vol. 5, article 1, 2005.

[130] G. Falconieri, M. A. Luna, S. Pizzolitto, G. DeMaglio, V.Angione, and M. Rocco, “Eosinophil-rich squamous carci-noma of the oral cavity: a study of 13 cases and delineationof a possible new microscopic entity,” Annals of DiagnosticPathology, vol. 12, no. 5, pp. 322–327, 2008.

[131] D. Tostes Oliveira, K. C. Tjioe, A. Assao et al., “Tissueeosinophilia and its association with tumoral invasion of oralcancer,” International Journal of Surgical Pathology, vol. 17,no. 3, pp. 244–249, 2009.

[132] A. A. Tadbir, M. J. Ashraf, and Y. Sardari, “Prognostic signif-icance of stromal eosinophilic infiltration in oral squamouscell carcinoma,” Journal of Craniofacial Surgery, vol. 20, no. 2,pp. 287–289, 2009.

[133] S. E. J. Leighton, J. G. C. Teo, S. F. Leung, A. Y. K. Cheung,J. C. K. Lee, and C. A. Van Hasselt, “Prevalence and prog-nostic significance of tumor-associated tissue eosinophilia innasopharyngeal carcinoma,” Cancer, vol. 77, no. 3, pp. 436–440, 1996.

[134] P. Manchanda, S. C. Sharma, and S. N. Das, “Differentialregulation of IL-2 and IL-4 in patients with tobacco-relatedoral squamous cell carcinoma,” Oral Diseases, vol. 12, no. 5,pp. 455–462, 2006.

[135] F. Riedel, S. Adam, P. Feick, S. Haas, K. Gotte, and K.Hormann, “Expression of IL-18 in patients with head andneck squamous cell carcinoma,” International Journal ofMolecular Medicine, vol. 13, no. 2, pp. 267–272, 2004.

[136] A. Sparano, D. M. R. Lathers, N. Achille, G. J. Petruzzelli,and M. R. I. Young, “Modulation of Th1 and Th2 cytokine

profiles and their association with advanced head and necksquamous cell carcinoma,” Otolaryngology—Head and NeckSurgery, vol. 131, no. 5, pp. 573–576, 2004.

[137] K. M. Lau, S. H. Cheng, K. W. Lo et al., “Increasein circulating Foxp3+ CD4+CD25high regulatory T cells innasopharyngeal carcinoma patients,” British Journal of Can-cer, vol. 96, no. 4, pp. 617–622, 2007.

[138] C. Schaefer, G. G. Kim, A. Albers, K. Hoermann, E. N.Myers, and T. L. Whiteside, “Characteristics of CD4+CD25+

regulatory T cells in the peripheral circulation of patientswith head and neck cancer,” British Journal of Cancer, vol. 92,no. 5, pp. 913–920, 2005.

[139] S. Schwarz, M. Butz, C. Morsczeck, T. E. Reichert, and O.Driemel, “Increased number of CD25+Foxp3+ regulatory Tcells in oral squamous cell carcinomas detected by chro-mogenic immunohistochemical double staining,” Journal ofOral Pathology and Medicine, vol. 37, no. 8, pp. 485–489,2008.

[140] L. Strauss, C. Bergmann, W. Gooding, J. T. Johnson, andT. L. Whiteside, “The frequency and suppressor function ofCD4+CD25high Foxp3+ T cells in the circulation of patientswith squamous cell carcinoma of the head and neck,” ClinicalCancer Research, vol. 13, no. 21, pp. 6301–6311, 2007.

[141] H. Dahlstrand, A. Nasman, M. Romanitan, D. Lindquist, T.Ramqvist, and T. Dalianis, “Human papillomavirus accountsboth for increased incidence and better prognosis in tonsillarcancer,” Anticancer Research, vol. 28, no. 2B, pp. 1133–1138,2008.

[142] K. Rosenquist, J. Wennerberg, K. Annertz et al., “Recurrencein patients with oral and oropharyngeal squamous cellcarcinoma: human papillomavirus and other risk factors,”Acta Oto-Laryngologica, vol. 127, no. 9, pp. 980–987, 2007.

[143] K. Lindel, K. T. Beer, J. Laissue, R. H. Greiner, and D. M.Aebersold, “Human papillomavirus positive squamous cellcarcinoma of the oropharynx: a radiosensitive subgroup ofhead and neck carcinoma,” Cancer, vol. 92, no. 4, pp. 805–813, 2001.

[144] D. Lindquist, M. Romanitan, L. Hammarstedt et al., “Humanpapillomavirus is a favourable prognostic factor in tonsillarcancer and its oncogenic role is supported by the expressionof E6 and E7,” Molecular Oncology, vol. 1, no. 3, pp. 350–355,2007.

[145] C. Fakhry, W. H. Westra, S. Li et al., “Improved survivalof patients with human papillomavirus-positive head andneck squamous cell carcinoma in a prospective clinical trial,”Journal of the National Cancer Institute, vol. 100, no. 4, pp.261–269, 2008.

[146] P. M. Weinberger, Z. Yu, B. G. Haffty et al., “Molecularclassification identifies a subset of human papillomavirus-associated oropharyngeal cancers with favorable prognosis,”Journal of Clinical Oncology, vol. 24, no. 5, pp. 736–747, 2006.

[147] P. T. Hennessey, W. H. Westra, and J. A. Califano, “Humanpapillomavirus and head and neck squamous cell carcinoma:recent evidence and clinical implications,” Journal of DentalResearch, vol. 88, no. 4, pp. 300–306, 2009.

[148] B. Kumar, K. G. Cordell, J. S. Lee et al., “EGFR, p16, HPVtiter, Bcl-xL and p53, sex, and smoking as indicators ofresponse to therapy and survival in oropharyngeal cancer,”Journal of Clinical Oncology, vol. 26, no. 19, pp. 3128–3137,2008.

[149] K. Rosenquist, “Risk factors in oral and oropharyngeal squa-mous cell carcinoma: a population-based case-control studyin southern Sweden,” Swedish Dental Journal. Supplement,vol. 179, pp. 1–66, 2005.


[150] B. G. Hansson, K. Rosenquist, A. Antonsson et al., “Strongassociation between infection with human papillomavirusand oral and oropharyngeal squamous cell carcinoma: apopulation-based case-control study in southern Sweden,”Acta Oto-Laryngologica, vol. 125, no. 12, pp. 1337–1344,2005.

[151] K. Morshed, “Association between human papillomavirusinfection and laryngeal squamous cell carcinoma,” Journal ofMedical Virology, vol. 82, no. 6, pp. 1017–1023, 2010.

[152] G. Bozdayi, Y. Kemaloglu, O. Ekinci et al., “Role ofhuman papillomavirus in the clinical and histopathologicfeatures of laryngeal and hypopharyngeal cancers,” Journal ofOtolaryngology—Head and Neck Surgery, vol. 38, no. 1, pp.119–125, 2009.

[153] S. W. Yang, Y. S. Lee, T. A. Chen, C. J. Wu, and C.N. Tsai, “Human papillomavirus in oral leukoplakia is noprognostic indicator of malignant transformation,” CancerEpidemiology, vol. 33, no. 2, pp. 118–122, 2009.

[154] K. Morshed, E. Korobowicz, M. Szymanski, D. Skomra,and W. Golabek, “Immunohistochemical demonstration ofmultiple HPV types in laryngeal squamous cell carcinoma,”European Archives of Oto-Rhino-Laryngology, vol. 262, no. 11,pp. 917–920, 2005.

[155] L. M. Aaltonen, T. Wahlstrom, H. Rihkanen, and A. Vaheri,“A novel method to culture laryngeal human papillomavirus-positive epithelial cells produces papilloma-type cytology oncollagen rafts,” European Journal of Cancer, vol. 34, no. 7, pp.1111–1116, 1998.

[156] P. Ernoux-Neufcoeur, M. Arafa, C. Decaestecker et al.,“Combined analysis of HPV DNA, p16, p21 and p53 topredict prognosis in patients with stage IV hypopharyngealcarcinoma,” Journal of Cancer Research and Clinical Oncology,vol. 137, no. 1, pp. 173–181, 2010.

[157] M. A. Stanley, “Immune responses to human papillomaviruses,” Indian Journal of Medical Research, vol. 130, no. 3,pp. 266–276, 2009.


[159] T. K. Hoffmann, C. Arsov, K. Schirlau et al., “T cellsspecific for HPV16 E7 epitopes in patients with squamouscell carcinoma of the oropharynx,” International Journal ofCancer, vol. 118, no. 8, pp. 1984–1991, 2006.

[160] R. Williams, D. W. Lee, B. D. Elzey, M. E. Anderson, B. S.Hostager, and J. H. Lee, “Preclinical models of HPV+ andHPV− HNSCC in mice: an immune clearance of HPV+HNSCC,” Head and Neck, vol. 31, no. 7, pp. 911–918, 2009.

Tumor Immunology - Hindawi Publishing...

Documents

Transcript of Tumor Immunology - Hindawi Publishing...