Current Drug Targets, 2011, 12, 1957-1973 1957 and Other Immunological Approaches for Immunoprevention

Pier-Luigi Lollini*,1, Giordano Nicoletti2, Lorena Landuzzi2, Federica Cavallo3, Guido Forni3, Carla De Giovanni4 and Patrizia Nanni4

1Department of Hematology and Oncological Sciences, University of Bologna; 2Rizzoli Orthopedic Institute, Bologna; 3Molecular Biotechnology Center, Department of Clinical and Biological Sciences, University of Turin; 4Department of Experimental Pathology, University of Bologna, Italy

Abstract: The immune system effectively prevents cancer, whereas severe immunodepression increases its incidence. is a strategy based on the concept that enhancement of tumor immunity in healthy individuals reduces cancer risk. It can be viewed as a kind of chemoprevention. For cancer immunoprevention, the cancer universe can be neatly divided between tumors caused - directly or indirectly - by infectious agents and all other tumors. Immunoprevention of tumors caused by infectious agents is already implemented at the population level for hepatitis B virus (HBV)-related and for tumors caused by human papillomaviruses (HPV), like cervical carcinoma. Now the challenge is to develop immunological strategies to prevent the bulk (>80%) of human tumor burden, unrelated to infections. Both vaccines against tumor and immune modulators can prevent tumor onset in cancer- prone mice. These studies outlined the target antigens and the molecular and cellular mechanisms of cancer immunoprevention: a) the best target antigens are surface molecules controlling tumor growth and progression (); b) combinations of potent vaccines and nonspecific stimuli (adjuvants) yield the strongest protection; c) immunoprevention must start early in the natural history of tumors, before key progression events like the onset of carcinoma in situ; d) lifetime protection requires repeated boosts, to maintain a strong and steady immune response; e) and helper, rather than cytotoxic, T cells mediate long-term protection from tumor onset; f) immunoprevention can be combined with chemoprevention. The development of agents like tamoxifen, which went from cancer therapy to chemoprevention, could be a model for the translation of cancer immunoprevention from mice to humans. Keywords: Antibodies, cancer vaccines, ErbB2, immunoprevention, mammary carcinoma, oncoantigens, viral tumors.

FROM IMMUNE SURVEILLANCE TO IMMUNO- a) The nude mutation primarily afflicts epithelial com- PREVENTION ponents, including the thymus, and the lack of an appro- priate thymic environment hampers the normal matura- Prevention of tumor onset and progression is a concern of tion of T cells; however, an extrathymic maturation all multicellular organisms. The diverse array of biological leads to the appearance of a few mature (though fre- strategies deployed to fend off exogenous and endogenous quently autoreactive) T cells; carcinogenic processes ranges from DNA repair and tumor suppressor genes to xenobiotic metabolism and skin pigmen- b) Athymic nude mice compensate their scarcitywith tation. One issue is whether the immune system is part of a potentiation of natural killer (NK) cell activity. NK these strategies. In other words, does it play a significant role cells actively prevent tumor onset and diffusion. In sum- in the natural prevention of carcinogenesis? This question mary, and especially for long-term studies of sponta- was set forth in the 1960s by the theory of cancer immune neous or carcinogen-induced tumor onset, nude mice are surveillance, which defined the ability to identify and des- not bonafide immune deficient hosts; troy nascent tumors as a fundamental property of the c) Many immune mechanisms that affect cancer onset and immune system [1]. progression are only marginally disturbed by the In the 1970s, however, the theory received an apparently absence of a normal T cell population. deadly blow by the lack of an increase in tumor incidence in Indeed the new millennium saw a vindication of the athymic nude mice [2]. Being the maturation of T cells immune surveillance theory, thanks to the advent of gene- inhibited, athymic nude mice lacking T cell-mediated activi- tically-modified mice, which allowed the generation of ties should be more prone to cancer than immunocompetent defined immune defects. Mice with an intrinsic, stable and wild-type ones. We now know that athymic mice are not an complete T and deficit were found to be more prone to appropriate animal model to study immune surveillance, for spontaneous or chemical carcinogenesis than wild-type mice, three main reasons [1]: and additional gene defects affecting natural immune responses exposed mice to the risk of more aggressive and precocious tumors [3]. These observations were repeatedly

confirmed by distinct laboratories studying gene knockout *Address correspondence to this author at the Sezione di Cancerologia, mice of different kind. At present, diverse experimental Viale Filopanti 22, 40126 Bologna, Italy; Tel: +39051-209-4786. Fax: evidence endorses the basic tenet of the immune surveillance +39051-242-169, E-mail: [email protected] theory: the immune system protects the host from tumor

1 - 5 /11 $58.00+.00 © 2011 Bentham Science Publishers 873 5 92 1958 Current Drug Targets, 2011, Vol. 12, No. 13 Lollini et al. onset, and immune deficits are accompanied by an increased trate the intrinsic carcinogenic potency of inflammation. This risk of cancer [1, 4-6]. opens up novel perspectives of immunoprevention, based not only on anti-inflammatory treatments, but also on more In humans, studies were mostly confined to syndromes of immune-oriented strategies that strive to re-orient the acquired immune deficiencies, since life expectancy in children with severe primary immune deficiencies was too immune response from pro-carcinogenic, frustrated chronic inflammation to tumor-preventive immune responses [11- short to study long-term tumor incidence. Thus we have 13]. mainly information concerning partial and variable immune deficiencies, like AIDS and post-transplant immune sup- The main perspectives in cancer immunoprevention are: pression. Nonetheless, human data confirm an increase in 1. Vaccines to prevent infection, especially chronic tumor incidence in immunodeficient individuals, in a special infection, by carcinogenic viruses and other infectious way for what concerns lymphoid tumors and tumors related agents; to viral infections, such as Kaposi sarcoma (human herpes virus-8) and anogenital carcinomas (HPV) [7], whereas the 2. Strategies to diminish chronic inflammation or to incidence of prevalent tumor types (e.g. breast, prostate and convert it into a beneficial immune response; colorectal carcinomas) is mostly indistinguishable from that 3. Induction of protective immune responses against of the general population. Among the “big killers”, only lung tumors unrelated to infectious agents, which we will cancer is increased in immunodeficient individuals, however call here collectively “non-infectious tumors”. many confounding factors prevent the establishment of a straightforward relationship with immune surveillance [7]. Here we will deal mainly with the first and third, because other reviews in this issue of Current Drug Targets deal with The increase of virus-related in immune deficien- inflammation, and at present it is not clear whether the cies has become a cornerstone of tumor and viral immuno- conversion of chronic inflammation into protective immunity logy, now leading to novel etiological hypotheses through might be addressed by original strategies, or more likely will reverse reasoning: a viral or infectious etiology for a given be based on various combinations of the other approaches tumor histotype can be postulated and investigated on the described here. basis of a high incidence among immunodeficient popula- tions [8, 9]. Two major differences separate immunoprevention of infectious and non-infectious tumors: a) prevention of To summarize, it is now evident that the immune system infectious tumors is a type of primary prevention, aimed at contributes to the prevention of cancer onset. Hosts with a removing risk factors, whereas prevention of non-infectious severe and stable immune deficiency experience a genera- tumors is at the interface between primary and secondary lized enhancement of carcinogenesis, though even transient prevention, because the typical target is a preneoplastic cell impairments of immune defences expose the host to the risk that has experienced carcinogenic hits; b) even more of specific tumor types, in particular those caused by infec- importantly, immunoprevention of infectious tumors is tious agents. Moreover, immune mechanisms can hold occult already implemented in human populations, whereas only cancer at bay for periods equivalent to the natural life span of promising preclinical results are available for non-infectious the individual, while a temporary immunodepression permits tumors. the tumor to progress [6, 10]. Natural cancer prevention mechanisms, from DNA repair to immunity, share common fundamental features: all limit IMMUNOPREVENTION OF INFECTIOUS TUMORS cancer development, but none is 100% efficient. Tumors Marek’s Disease eventually arise, because preventive systems become less efficient with aging, while the risk of cancer increases. Given the advancement of human applications, we will Importantly, most mechanisms are modulable, suggesting not deal with the abundance of preclinical studies in animal that exogenous treatments can yield a further decrease in models. It is worth mentioning, however, that the first appli- cancer risk [5]. cation of cancer immunoprevention outside the laboratory was the against Marek’s disease, a chicken T cell lymphoma caused by the eponymous alphaherpesvirus STRATEGIES OF IMMUNOPREVENTION MDV. The considerable economic damage caused to inten- sive poultry farms by outbreaks of Marek’s disease led in the Cancer immunoprevention strives to prevent tumor onset late 1960s to the development of a first effective vaccine and tumor progression by means of immunological inter- based on a non-virulent turkey strain of MDV [14, 15]. Now ventions administered to healthy individuals. the number of chickens receiving MDV vaccination is larger A further contribution to the field comes from the than the entire human population, more than 20 billions are renewed interest in the role of chronic inflammation in vaccinated in ovo each year worldwide. carcinogenesis. A continuing activation of leukocyte popula- Interestingly, MDV vaccines effectively prevent high tions, possibly triggered by unrelenting infections and/or level viremia and lymphomagenesis, but do not block low tissue damage, contributes to neoplastic transformation with level viremia and virus transmission. Under this respect, multiple mechanisms that range from reactive oxygen human vaccines seem to induce a higher level of protection species to growth and angiogenic factors. The contribution from carcinogenic viruses, however clinical trials of HPV of chronic inflammation to viral and chemical carcinogenesis vaccines (see below) clearly showed a higher degree of in humans has been acknowledged for many years. More protection from chronic than from acute infection, a situation recently, genetically modified mouse models clearly illus- remindful of the high level/low level MDV viremia. In chic- Cancer Immunoprevention Current Drug Targets, 2011, Vol. 12, No. 13 1959 kens the lack of sterilyzing immunity allowed a progressive develop that vaccine was to prevent acute hepatitis, and selection of more virulent MDV and required the develop- cancer immunoprevention is just a (very useful) byproduct. ment of successive generations of improved vaccines com- Clearly this does not apply to HPV, which essentially causes bining non-virulent and attenuated viral strains. neoplastic pathologies, ranging from benign warts to malig- nant cervical and anogenital carcinomas [20]. Thus, HPV Hepatocellular Carcinoma vaccines were the first “pure” implementation of human cancer immunoprevention. Vaccination against hepatitis B virus (HBV) significantly reduces the incidence of post-hepatitis hepatocellular carci- The worldwide diffusion of vaccination programs against noma (HCC) [16]. A recent follow-up of the Taiwanese HPV began a few years ago, therefore long-term efficacy of study [16] confirmed a strong risk reduction for vaccinated cervical carcinoma immunoprevention at the population children; protection from HCC was directly related to level has not yet been established. However the results of completion of the vaccination course and to the use of the clinical trials were extremely favorable and predict an almost more effective recombinant vaccine (Table 1). complete prevention of carcinogenesis (Table 1) [21-23]. The success of HPV vaccines allows a better appraisal of the kind of issues raised by a cancer immunoprevention Table 1. Key Results of Human Cancer Immunoprevention programme. First, approved vaccines elicit protective immune responses against two or four HPV types, and there 1 HBV vaccine [16] Risk of HCC is evidence of partial cross-protection from further viral Vaccinated vs. non-vaccinated 0.31 (0.24-0.41) types [24], but broad spectrum immunity from HPV (more than 100 types) is a perspective for future generation Incomplete vaccination vs. complete 4.32 (2.34-7.91) vaccines [25]. Second, current vaccines are eminently Recombinant vaccine vs. non-recombinant 0.43 (0.19-0.97) prophylactic, but lack therapeutic efficacy [21, 24]. Further-

2 more, the natural course of human HPV infection in Western HPV vaccine [101] Efficacy countries is marked by a peak of incidence in young, Prevention of infection 97% (81.3-99.9) sexually active people, followed by a pronounced decay of HPV positivity over the years, leaving only a minority of Prevention of persistent infection 100% (52.2-100.0) adults at risk of cancer [26]. Hence vaccinations are usually Prevention of atypical squamous cells 96% (83.5-99.5) scheduled before the onset of sexual activity, a logical choice Prevention of cervical carcinoma 100% (42.4-100.0) that, however, led to some fears that vaccination might encourage sexual promiscuity among adolescents reassured 1Odds ratio and 95% confidence interval 2Vaccine efficacy % and 95% confidence interval of their immunity from HPV. Compounded with the current cost of the vaccine, this has slowed adoption of vaccination programmes in some countries and states. HBV vaccination prevents not only acute hepatitis, but also the sequelae that develop in a minority of patients, Sexual transmission of HPV ensures infection of both including persistent viral infection, chronic hepatitis, liver sexes, however only girls are currently undergoing vaccina- cirrhosis, and eventually hepatocellular carcinoma. Chronic tions, because tumors affect mainly females, however it inflammation plays a relevant role in the onset of liver would be more sensible to vaccinate also males, which are a cancer [17] that follows HBV infection. Hence the success of large reservoir of infection. A global vaccination programme HBV vaccine suggests that, once the triggering event is could bring a prophylactic benefit to males, because HPV- known, immunoprevention can be envisaged as a form of related oral cancers and genital warts affect both sexes, and primary prevention of carcinogenic chronic inflammation. is the only possibility to establish herd immunity and to realistically aim at the eradication of the virus [27, 28]. An effective vaccine against hepatitis C virus (HCV) infection is needed to complete the immunoprevention of Primary prevention of cervical carcinoma with HPV post-hepatitis HCC. HCV is less prevalent than HBV, but vaccines and molecular diagnosis of HPV infection are carries a higher risk of persistent infection, chronic hepatitis, starting to modify secondary prevention programmes based cirrhosis and HCC. The effort toward an HCV vaccine was on PAP test or other morphological methods. In the future, if hampered by the shortage of preclinical in vitro and in vivo vaccines hold the promise to eradicate cervical carcinoma, models and by the heterogeneity and mutability of the virus. mass screening could be altogether abandoned, but what Nonetheless, some promising candidates were eventually should we do in the meantime? Molecular screening is developed and are now undergoing early clinical trials [18]. effective at reducing cancer mortality, and can be even better Worldwide implementation of a combined HBV+HCV vac- than traditional methods [29]. Probably the best interim cination programme is bound to impact human carcino- policy will be to combine molecular and morphological early genesis, because HCC accounts for more than 4% of all diagnosis, for example using HPV status to include only human cancers, and 80% of cases are associated with viral women at risk in early diagnosis programmes [30]. infection [19]. In perspective, a worldwide implementation of multiva- lent HPV vaccines offers a concrete possibility to eradicate Cervical Carcinoma malignant and benign human tumors caused by HPV, much If human cancer immunoprevention were based only on as happened in the past with smallpox, leading for the first the HBV vaccine, one might object that the original thrust to time in human history to the disappearance of a prevalent neoplastic disease. 1960 Current Drug Targets, 2011, Vol. 12, No. 13 Lollini et al.

Perspectives and Targets for Immunoprevention of The diverse immunological treatments that were adminis- Infectious Tumors tered to healthy mice to prevent cancer onset at a later age can be grouped into -specific or non-antigen specific. Infectious tumors contribute an estimated 17.8% of all Antigen-specific agents, such as vaccines, target molecules human tumors [19]. The agents discussed above (HBV, associated with cancer development, whereas non-antigen HCV, HPV) account for more than 10% of all human specific treatments, like and immune stimulants, tumors, that is 57% of the whole infectious carcinogenesis. enhance a broad immune reactivity that includes anti-tumor What can be done for the remaining 43%? components. For a variety of reasons, not all types of infectious tumors In general, antigen-specific strategies are preferred. How- appear to be equally “vaccinable” (i.e. amenable to vaccine ever, in the context of cancer immunoprevention, a non- immunoprevention, akin to the use of “druggable” in phar- antigen specific strategy has a distinct translational edge, macology). For example, viral tumors occurring in advanced because it makes no assumption on the exact antigens exp- AIDS patients will certainly be better prevented by an ressed by future tumors. The success of immunopreventive effective human immune deficiency virus (HIV) vaccine, vaccines developed in genetically modified mouse models which could altogether prevent immune deficiency, whereas depended on the a priori knowledge that tumors will express cancer immunopreventive vaccines could be envisaged for a given antigenic target. For example, anti-ErbB2 vaccines organ transplant candidates [31]. Massive HIV vaccine work in ErbB2 transgenic mice because all nascent tumors research will indirectly benefit also the development of overexpress ErbB2. In the overwhelming majority of hu- vaccines against the human leukemogenic retrovirus HTLV- mans, however, we are currently unable to predict not only 1, which causes only a tiny proportion (0.1%) of human antigenic targets, but even the histotype of future tumors, cancers. thus precluding a straightforward translation of many Another case in which vaccines are not very actively successful preclinical results. pursued is the prevention of gastric cancer, even though the number of worldwide cases is on a par with liver and cer- Antigen Non-Specific Strategies of Cancer Immuno- vical carcinomas. Pharmacological eradication of Helico- prevention bacter pylori is the preferred preventive strategy, even though H. pylori therapy is not devoid of difficulties, and Significant prevention of tumor onset was obtained using does not prevent reinfection [32]. agents with widely different mechanisms of action, including interleukin 12 (IL-12), a of antigen presenting cell A good target is Epstein-Barr virus (EBV), that is impli- [44, 45], α-galactosyl ceramide, which activates NKT cells cated in a variety of diseases worldwide: infectious mono- [38], Treg depletion [39], circumvention of T cell negative nucleosis in Western countries, nasopharyngeal carcinoma in selection [40], DNA containing unmethylated CpG seq- Asia, Burkitt lymphoma in Africa, lympho–proliferative dis- uences [41, 42], which engage TLR9 cellular receptors in eases in immunodeficient patients. In a recent editorial the various leukocyte populations, or poly I:C, which engages development of EBV vaccines was described as “agoniz- TLR3 and activates NK cells [43]. The diversity of effective ingly slow”, for the lack of a consensus applicative strategy agents indicates that enhancement of immune surveillance [33], however some promising candidate vaccines [31, 34] and control of tumor onset can be obtained through different offer “an opportunity to make a difference in global health” immune mechanisms. [33], as EBV-related account for more than 1% of all human tumors [19]. Two critical issues are the potency of these treatments and their toxicity. When compared to vaccines in the same Finally, not to forget the veterinary roots of viral tumor mouse model, antigen non-specific strategies invariably immunoprevention, bovine papillomavirus (BPV) causes show a lower potency, and usually do not result in a sizeable tumors not only in cattle, but also in equids, and vaccines proportion of long-term tumor-free mice, but only delay were developed both with a prophylactic and with a carcinogenesis [38, 41, 42, 44, 45]. Potency is directly con- therapeutic intent [35]. nected to toxicity, because most immune responses cannot be kept indefinitely at high levels of activation without IMMUNOPREVENTION OF NON-INFECTIOUS inflicting damages to the host. This is clearly a problem in TUMORS cancer immunoprevention, because lifelong treatments are mostly unfeasible with current antigen non-specific agents. Will we ever replicate in human non-infectious carcino- Furthermore IL-12, which was effective in mice both against genesis the successes of cancer immunoprevention with carcinogen-induced and -driven carcinogenesis [44, antiviral vaccines? The pursuit of this challenging endeavor 45], is known to be more toxic for humans than for rodents is no longer based on wishful thinking, because in the last [46], hence preclinical results represent a fundamental proof- decade preclinical results clearly showed that distinct of-concept, but effective dosages are not directly applicable immunological interventions effectively limit tumor deve- to humans. lopment in mice carrying a genetic risk of cancer or exposed to chemical carcinogens [36, 37]. Antigen Specific Strategies of Cancer Immunoprevention In this section we will review the experimental evidence When the antigenic profile of forthcoming tumors can be of cancer immunoprevention, expound the main concepts predicted, antigen specific strategies offer the most effective and strategies that emerge from preclinical models of auto- cancer immunoprevention, because immunological respon- chthonous tumor growth, and address the translation to ses can be precisely focused on target antigens, immuno- human conditions. Cancer Immunoprevention Current Drug Targets, 2011, Vol. 12, No. 13 1961 logical resources are not wasted to elicit unrelated responses delayed carcinogenesis in a variety of different model and, at least for what concerns antibodies, long-term main- systems, thus confirming that immunoprevention of non- tenance of protective responses is devoid of limiting infectious tumors is a general concept applicable to every toxicities. tumor type. We summarized in Table 2 the main features and Most preclinical results were obtained with active stra- results obtained with active immunoprevention in gene- tically-modified mice. Note that Table 2 includes only the tegies, i.e. vaccines, however it must be stressed that also results of prophylactic vaccinations administered to hosts at administration of monoclonal antibodies (MAb) before risk of spontaneous tumor onset, not experiments in which tumor onset, which would be classified as “passive immuni- vaccinated mice were challenged with cultured or trans- zation”, protects from carcinogenesis [47-49], hence suggest- plantable tumor cells. The rest of this section is focused on ing that currently available therapeutic MAbs could be used for immunoprevention in individuals at risk of cancer. cancer-prone ErbB2 transgenic mice (BALBneuT mouse line especially), which were the most commonly studied model Vaccination of healthy mice at risk of cancer, mainly system (Table 2a). This permits a meaningful comparison of transgenic mice carrying , or knockout mice with different immunological modalities in the prevention of the altered tumor suppressor genes, effectively reduced or same carcinogenic process.

Table 2a. Immunoprevention* Targeting the Cancer-Causing Gene Product. ErbB2

Carcinogenesis model Vaccine Results

Cancer- Prevention effects§ Refs. Mouse # causing Tumor(s) Type Antigen + adjuvant(s) Immune effects (and follow up in model gene wk of age)

rat ErbB2 and GM-CSF IFN-γ-releasing T cells Cell vaccine and transduced 3T3 cells. rat ErbB2-specific CTL Slightly delayed gene-based [102] Recombinant vaccinia virus tumor onset (40 wk) vaccine (viral) Anti-rat ErbB2 expressing rErbB2 gene antibodies Anti-human ERBB2 Dendritic cell Truncated human Delayed tumor antibodies (mainly [103] (DC) vaccine ERBB2+ transduced DC onset (43 wk) IgG2a)

Mammary Anti-rat ErbB2 Extracellular domain of proliferating T cells Delayed tumor rat ErbB2 carcinomas in [104] proto- FVBneuN about 100% of rat ErbB2 Anti-rat ErbB2 onset (65 wk) oncogene females by 35-50 antibodies wk of age Protein/peptide Chimeric human ERBB2 Anti-human ERBB2 Complete protection vaccine [105] peptides antibodies (52 wk)

human ERBB2 IFN-γ-releasing T cells Delayed tumor complexed with heat [106] Anti-rat ErbB2 IgG2a onset (69 wk) shock protein 110 antibodies Plasmid DNA expressing Gene-based Anti-human ERBB2 Delayed tumor full-length or EC-TM of [107] vaccine (plasmid) antibodies onset (52 wk) human ERBB2

rat ErbB2 Allogeneic fibroblasts T cell infiltration Complete protection activated Cell vaccine transfected with rat [108] Anti-rat ErbB2 (about 25 wk) oncogene ErbB2 gene antibodies human ERBB2 peptides (B-cell epitopes) Anti-human ERBB2 Slightly delayed Protein/peptide conjugated with tetanus and anti-rat ErbB2 tumor onset (about [109] vaccine toxoid + Gerbu adjuvant antibodies 40 wk) + systemic IL-12 Mammary carcinomas in Anti-rat ErbB2 FVBneuT about 100% of Plasmid DNA expressing proliferating T cells Delayed tumor females by 25-30 full-length or truncated IFN- -releasing T cells [110] γ onset (52 wk) wk of age rat ErbB2 Anti-rat ErbB2 Gene-based antibodies vaccine (plasmid) Anti-rat ErbB2 Electroporated plasmid antibodies Complete protection DNA expressing EC-TM [111] (52 wk) rat ErbB2 IFN-γ-producing CD8 T cells Gene-based Alphavirus replicon particles Anti-rat ErbB2 Complete protection [112] vaccine (viral) containing rat ErbB2 gene antibodies (35 wk) 1962 Current Drug Targets, 2011, Vol. 12, No. 13 Lollini et al. (Table 2a) Contd….. Carcinogenesis model Vaccine Results

Cancer- Prevention effects§ Refs. Mouse # causing Tumor(s) Type Antigen + adjuvant(s) Immune effects (and follow up in model gene wk of age)

BALBneuT Mammary Allogeneic rat ErbB2+ IFN-γ-releasing T cells [50, carcinomas in tumor cells + IL12 Complete protection 51, 100% of females (recombinant cytokine or Anti-rat ErbB2 IgG2a (52 wk) 55] by 20-25 weeks gene-transduced cells) and 2b antibodies of age Allogeneic rat ErbB2 + Anti-rat ErbB2 IL-12-expressing tumor Delayed tumor antibodies [79] Cell vaccine cell (short course) + onset (100 wk) tamoxifen IFN-γ-producing cells Delayed (rat ErbB2 human HER-1 (EGF-R), vaccine) or slightly -2 (ERBB2), -3, -4, rat Anti-rat ErbB2 delayed (human [113] ErbB2 transduced 3T3 antibodies ERBB2 vaccine) cells tumor onset (40 wk)

IFN-γ-releasing T DC transduced with cells Delayed tumor [114] EC-TM rat ErbB2 Anti-rat ErbB2 onset (28 wk) Dendritic cell antibodies (DC) vaccine DC transduced with Anti-rat ErbB2 truncated rat ErbB2 + antibodies Delayed tumor [115] IL-15 + IL-15 receptor No anti-rat ErbB2 growth (30 wk) alpha CTL human ERBB2 peptide vaccines (spanning Slightly reduced Anti-human ERBB2 [48, regions targeted by tumor volume (25 antibodies 116] Protein/peptide or wk) vaccine pertuzumab) rat ErbB2 peptide vaccine IFN- -producing CD8 Delayed tumor + TLR agonist adjuvant + γ [92] T cells onset (42 wk) Treg depletion Tumor infiltrating T cells Plasmid DNA expressing Delayed tumor No anti-rat ErbB2 CTL [117] EC-TM rat ErbB2 onset (38 wk) Anti-rat ErbB2 antibodies

IFN-γ-releasing T cells Tumor infiltrating T Electroporation of cells Complete protection [52, plasmid DNA expressing (52 wk) 53] EC-TM rat ErbB2 No anti-rat ErbB2 CTL Anti-rat ErbB2 antibodies Electroporated rat ErbB2 DNA vaccine (plasmids IFN- -producing CD8 Delayed tumor designed to elicit CD8 T γ [118] Gene-based T cells onset (40 wk) vaccine (plasmid) cell responses but no response)

Electroporated plasmid Anti-rat ErbB2 or anti- Delayed tumor onset DNA expressing EC-TM human ERBB2 (46 wk) Almost no rat ErbB2 (or human antibodies (not [119] protection by human ERBB2) + GM-CSF crossreacting) ERBB2 plasmid plasmid no role played by CD8 Electroporated EC-TM Anti-mouse Amot rat ErbB2 and human antibodies (in non Complete tumor angiomotin (human transgenic mice) [120] prevention (65 wk) Amot) combined causing anti- plasmids angiogenic effect Anti-rat ErbB2 Electroporated rat ErbB2 Delayed tumor antibodies [54] plasmid + Treg depletion onset (100 wk) low avidity CTL Cancer Immunoprevention Current Drug Targets, 2011, Vol. 12, No. 13 1963 (Table 2a) Contd….. Carcinogenesis model Vaccine Results

Cancer- Prevention effects§ Refs. Mouse # causing Tumor(s) Type Antigen + adjuvant(s) Immune effects (and follow up in model gene wk of age)

Gene-based IFN-γ-releasing CD4 T vaccine (viral) Adenovirus vector cells Complete protection encoding EC-TM rat No anti-rat ErbB2 CTL [56] (50 wk) ErbB2 Anti-rat ErbB2 IgG2a antibodies rat ErbB2 recombinant Anti-rat ErbB2 Delayed tumor [121] vaccinia virus antibodies onset (40 wk)

Gene-based Murine polyoma virus- IFN-γ-releasing T cells Complete protection vaccine (virus- like particles containing [122] No anti-human (50 wk) like particles) EC-TM human ERBB2 ERBB2 antibodies

Adenoviral (Ad) vectors Anti-rat ErbB2 Gene-based Delayed tumor and electroporated antibodies (Ad vector) vaccine (viral and onset (Ad vector) [123] plasmid encoding codon- plasmid) IFN-γ-producing cells (45 wk) optimized rat ErbB2 (Ad vector)

Syngeneic rat ErbB2 + Anti-rat ErbB2 Cell and gene- tumor cells challenge + antibodies Delayed tumor based vaccine Treg depletion, boost [93] onset (66 wk) (plasmid) with electroporated rat IFN-γ-producing T ErbB2 DNA cells Mammary Anti-rat ErbB2 Delayed tumor onset carcinomas in Electroporated rat antibodies (with with plasmids (BALBneuT Gene-based 100% of females ErbB2/human ERBB2 plasmids containing containing rat ErbB2 [124] x B6) F1 vaccine (plasmid) by 30 weeks of chimeric DNA plasmids rat ErbB2 sequences sequences at the age at the NH2-end) NH2-end (70 wk) *Table only reports active immunization studies of prevention of autochthonous tumors. #Immune effects elicited by the vaccines. These were not always formally demonstrated to be essential for cancer prevention. §When more than one protocol was studied, only the best results are reported. Complete protection = 80-100% protection. EC-TM = Extracellular and transmembrane domains; TLR = Toll-like receptor

Table 2b. Immunoprevention* Targeting the Cancer-Causing Gene Product. Ret and SV40

Carcinogenesis model Vaccine Results

Prevention effects§ Refs. Immune Cancer-causing gene Mouse model Tumor(s) Type Antigen + adjuvant(s) (and follow up in effects# wk of age) ret peptide vaccine + hRET proto-oncogene Antibodies to ret MT/ret 304/B6 Subcutaneous CpG adjuvant + 1-methyl Slightly delayed under mouse Protein/pepti CTL transgenic tumors (head and tryptophan (indoleamine tumor growth [125] metallothionein 1 de vaccine mice neck) after 4 wk 2,3-dioxygenase IFN-γ-producing (14 wk) promoter/enhancer inhibitor, per os) cells Insulinomas. SV40 Tag under rat Abortive Tag-specific Delayed tumor RIP1-Tag4 Average lifespan 38 SV40 virus [126] insulin promoter infection CTL onset (100 wk) weeks. Allogeneic SV40- Antibodies Delayed tumor Cell vaccine transformed cells + [127] IFN-γ- onset (67 wk) Prostate systemic IL12 producing cells SV40 Tag under rat TRAMP adenocarcinomas by probasin promoter Histological evi- 40 wk Dendritic cell Tag-IV peptide-pulsed Tag-specific dence of reduced [128] (DC) vaccine and LPS-treated DC CTL disease prog- ression (16 wk) SV40 Tag under Carcinomas, Tag-, IL-7-, and B7.1- IFN-γ-pro- Complete chimeric -actin/ - β β LoxP-TAg sarcomas, leukemias Cell vaccine transduced syngeneic ducing T cells protection [129] globin promoter by 40-80 wk sarcoma cells (104 wk) (conditional) CTL SV40 Tag under Delayed tumor Hepatoma (13 wk Dendritic cell TagIV peptide - TagIV-specific albumin promoter AST C–L onset (43 wk after [130] after induction) (DC) vaccine loaded DC CTL (conditional) induction) *Table only reports active immunization studies of prevention of autochthonous tumors. #Immune effects elicited by the vaccines. These were not always formally demonstrated to be essential for cancer prevention. §When more than one protocol was studied, only the best results are reported. Complete protection = 80-100% protection. Tag = SV40 large T antigen 1964 Current Drug Targets, 2011, Vol. 12, No. 13 Lollini et al.

Table 2c. Immunoprevention* Targeting Non Engineered or Undefined Antigens

Carcinogenesis model Vaccine Results Prevention Cancer-causing Mouse Immune effects§ (and Refs. Tumor(s) Type Antigen + adjuvant(s) gene model effects# follow up in wk of age)

Polyoma virus Multifocal mammary MTAG middle T antigen carcinogenesis in 100% Allogeneic sarcoma cell Vaccine Delayed tumor (MMTV- Cell vaccine [131] (under MMTV female mice by 23 wk of challenge rejection growth (23 wk) PyMT/B6) promoter) age Diffuse C cell hyperplasia Ret (mutated at xenogenic (human or CD8 Ret/Cal in 100% mice at 3 wk, then 60-70% codon 634, driven Dendritic cell salmon) calcitonin-pulsed [132, transgenic 100% diffuse medullary IFN-γ- decreased tumor by calcitonin vaccine dendritic cells (from non 133] mice thyroid carcinoma at 26 releasing growth (26 wk) promoter) transgenic C57BL6 mice) wk of age cells Electroporated plasmid Gene-based Slightly delayed encoding a soluble vaccine CD8 tumor growth [134] mTERT (inactivated by 2 (plasmid) (60 wk) Prostate aminoacid substitutions) adenocarcinomas in SV40 Tag (under mPSCA (Prostate stem TRAMP 100% male mice, Delayed tumor probasin promoter) cell antigen) plasmid leading to death within Gene-based IFN- - growth (no (prime, gene gun) and γ 50 wk vaccine (viral releasing induction of [135] Venezuelan Equine and plasmid) cells rheumatoid Encephalitis replicon factor) (50 wk) particle (boost)

Apc (germ-line Multiple adenomatous IL4- Fusion cells between tumor-suppressor Min+/- polyps of gastro-intestinal releasing APC Dendritic cell dendritic cells and + Decreased tumor mutation at codon tract by 10 weeks of age CD4 cells [136] and APC1309 vaccine APC1309intestinal tumor growth (22 wk) 850 or truncated at (leading to death within cells + IL-12 Anti-tumor codon 1309) 16 wk of age) antibodies mCyclin B1 plasmid Anti- p53 (homozygous Mainly lymphoma Gene-based Slightly delayed (prime, gene gun) and CyclinB1 tumor-suppressor p53-/- (death within 37 wk of vaccine (viral tumor growth [137] recombinant protein IgG knockout) age) and plasmid) (37 wk) (boost) antibodies

Ptch1 After exposure to IFN-γ- 42% reduction in (heterozygous ionizing radiation (at 8 releasing T the number of Protein/peptide mHip1 recombinant germ-line tumor- Ptch1+/- wk of age), microscopic cells microscopic basal [138] vaccine polypeptides suppressor basal cell carcinomas in Anti-Hip1 cell carcinomas knockout) 100% mice at 30 wk antibodies (32 wk) *Table only reports active immunization studies of prevention of autochthonous tumors. #Immune effects elicited by the vaccines. These were not always formally demonstrated to be essential for cancer prevention. §When more than one protocol was studied, only the best results are reported. Complete protection = 80-100% protection. TERT = telomerase

BALBneuT mice carry a mutant ErbB2 transgene of rat potent adjuvants and with the use of intensive vaccination origin under the transcriptional control of a tissue-specific protocols. The Triplex vaccine was made of cells expressing mammary tumor virus element, resulting in an aggressive the ErbB2 oncoprotein (p185), IL-12 and allogeneic class I mammary carcinogenesis. Around six months of age all major histocompatibility complex (MHC) antigens (a female mice develop ErbB2+ multifocal mammary carcino- polyclonal T cell inducer) [50, 51]. The DNA vaccine, which mas which grow rapidly and metastasize. Immunoprevention encoded ErbB2 in plasmids containing CpG sequences, was based on antigen non-specific treatments (see above) or on administered by electroporation, which enhances antigen specific vaccines of low potency delay mammary carcino- expression and immunogenicity [52]. Short vaccination genesis, but mice usually develop progressive tumors before protocols delayed carcinogenesis for several months, but one year of age. Successive improvements led to the only repeated boostings over long periods resulted in long- development of two highly active anti-ErbB2 vaccines, one term tumor-free mice. The situation is remindful of tetanus based on cells and one on DNA, that keep all mice free from immunity, which protects humans for many years, but not mammary carcinoma at least until one year of age [50-54], forever, without periodic vaccine booster shots. when these long-term experiments are usually concluded. In- Somewhat surprisingly, in ErbB2 trangenic mice the depth studies of the protective mechanisms elicited by these various vaccines elicit a minor T-cell mediated cytotoxic vaccines led to similar conclusions and will be discussed response. This is partly due to the lack of high-avidity T cells together. and to the negative control by Treg cells [54]. Most of the Quantum jumps in cancer immunopreventive activity anti-tumor action rests on the multiple direct and indirect were obtained with the combination of the antigen with antitumor activities of vaccine-elicited anti-ErbB2 antibodies Cancer Immunoprevention Current Drug Targets, 2011, Vol. 12, No. 13 1965

Table 2d. Immunoprevention* in Models with Multiple Engineered Genes

Carcinogenesis model Vaccine Results Prevention Immune effects§ (and Refs. Genes Mouse model Tumor(s) Type Antigen + adjuvant(s) effects# follow up in wk of age)

Prime with vaccinia Proliferating + virus, expressing CEA, CD4 T cells Complete human CEA Multiple adenomatous B7.1, ICAM-1, LFA-3 IFN-γ- + recombinant GM-CSF (celecoxib) or (carcino-embryonic polyps of gastro- Gene-based releasing partial (without [139, CEA.Tg/MIN + antigen) transgenic intestinal tract, leading ± celecoxib. Boost with CD8 T cells celecoxib) mice crossed with to death within 25-28 vaccine (viral) fowlpox virus 140] Min expressing CEA, B7.1, CTL protection (50- APC wk 77 wk) ICAM-1, LFA-3 and Anti-CEA GM-CSF antibodies IFN-γ- releasing T cells Complete rat ErbB2+ allogeneic Tumor mammary carcinoma (females) or infiltrating T partial (males) cells + IL12 cells Salivary gland Cell Vaccine (recombinant cytokine protection from [80] carcinomas in 100% of or gene-transduced No anti-rat carcinomas, ErbB2 CTL delayed onset of rat ErbB2 (mutated females by 15-18 cells) weeks of age, followed sarcoma (52 wk) oncogene, under Anti-rat MMTV promoter) BALB- by mammary ErbB2 and p53 +/- carcinomas by 20 wk. antibodies neuT/p53 Salivary gland heterozygous Complete knockout carcinomas (by 15-18 wk) and/or Electroporated DNA protection from Antibodies salivary rhabdomyo–sarcomas vaccine (combination of in all males (by 20 wk) Gene-based IFN-γ- carcinoma plasmids carrying releasing T vaccine extracellular and trans- (females), [82] (plasmid) cells delayed tumor membrane domains of onset for rat ErbB2, allogeneic Th1 polarized H-2Dq and IL-12) response mammary carcinoma (65 wk) rat ErbB2 (mutated Complete Parotid gland rat ErbB2 EC- (females) or oncogene, under Gene-based MMTV promoter) BALB- carcinomas in 100% of TMelectroporated partial (males) and dominant 172R-H females by 10-12 wk of vaccine plasmid combined with Antibodies protection from [81] neuT/p53 age, followed by (plasmid) systemic recombinant negative mutated parotid heterozygous p53 mammary carcinomas. IL-12 carcinoma (40 wk) rat ErbB2 (mutated decreased oncogene, under Mammary carcinomas human ERBB2 p435 intratumoral BALB-neuT × in 100% of females by Protein/peptide and p776 peptides + Treg cells Delayed tumor MMTV promoter), A2.1/DR1 vaccine onset (52 wk) [141] HLA-A2.1, HLA- 22 weeks of age IFA + recombinant GM-CSF specific CTL DR1 and Th cells Polyoma middle T Double Mammary carcinoma Dendritic cells fused CTL (under MMTV Dendritic cell + Delayed tumor transgenic in 100% of females by with human MUC1 Anti-MUC1 [142] promoter) and MMT mice 15 weeks of age vaccine tumor cells onset (26 wk) human MUC1 antibodies triple Complete transgenic Progression from human MUC1 peptides IFN-γ- (P48Cre bred protection from K-ras (mutated preneoplastic + HBV peptide + CpG releasing T with LSL- cells carcinoma, oncogene, intraepithelial lesion to Protein/peptide + GM-CSF + Freund [143] conditional), KRASG12D delayed and then invasive pancreatic vaccine combined with Cox-2 CTL progression of human MUC1 ductal adenocarcinoma inhibitor and anti-MUC1 preneoplastic mated to at 26-39 wk gemcitabine hMUC1.Tg antibodies lesions (39 wk) mice) anti-MUC1 IgG Inflammatory bowel disease (IBD) starting CTL Delayed IBD human MUC1 decrease of and about 70% transgenic mice MUC1+/IL10- at at 6-8 weeks of age Protein/peptide Synthetic human MUC1 [13] /- and progressing to vaccine peptide + TLR4 agonist neutrophils prevention of bred to IL-10-/- colitis-associated colon and myeloid- CACC (16 wk) cancer (CACC) derived suppressor cells *Table only reports active immunization studies of prevention of autochthonous tumors. #Immune effects elicited by the vaccines. These were not always formally demonstrated to be essential for cancer prevention. §When more than one protocol was studied, only the best results are reported. Complete protection = 80-100% protection.

[55-57]. The role of humoral response seems to be parti- , which is both the target of the cularly prominent because of the double role of ErbB2 vaccine and a membrane-exposed receptor regulating cell 1966 Current Drug Targets, 2011, Vol. 12, No. 13 Lollini et al. growth. The direct activities of vaccine-elicited anti-ErbB2 upheld for long time periods, ideally for the entire life span antibodies that affect the transformed phenotype of ErbB2+ of the host. In the meantime, genetic instability will foster tumors include block of mitogenic signal transduction the onset of poorly immunogenic transformed cells that can through inhibition of receptor dimerization and induction of evade immune surveillance and give rise to malignant internalization and recycling. In the ErbB2+ tumor cells tumors. Downmodulation or loss of the target antigen by these antibodies cause a drastic reduction of the basal level neoplastic cells is an obvious cause of immune evasion, of Akt phosphorylation in the absence of an impairment of however the most prevalent mechanism is the downmodu- PI3K enzymatic activity and induce an increase of PTEN lation of molecules involved in peptide antigen processing, phosphatase activity that correlated with a reduced PTEN in particular class I MHC glycoproteins [62, 63], which tyrosine phosphorylation [58]. However, vaccine-elicited affects more than 80% of all human tumors and prevents anti-ErbB2 antibodies also affect tumor growth by mediating recognition and lysis of tumor cells by cytotoxic T cells. antibody dependent cell-mediated cytotoxicity (ADCC) [59] We defined two essential features of target tumor and complement-dependent cytotoxicity. In immunized antigens that preempt the loss of immune recognition. The BALBneuT mice, IgG2a is the dominant isotype of serum antigen must a) have an essential role in tumor growth or antibodies to ErbB2. This indicates that vaccination elicits progression, and b) be a target of antibodies. Antigens like the activation of T helper cells producing - (IFN- γ the ErbB2 oncogene in transgenic mice are the drivers of the γ), the primary switch factor for IgG2a. Indeed, IFN-γ carcinogenic process, and antigen loss variants can occur, release was the most conspicuous T cell response present in but in the early phases of cancer progression their tumori- vaccinated ErbB2 transgenic mice, in the absence of detect- genic potential will be low or nil [64, 65]. Similar criteria are able cytotoxic T lymphocytes (CTL) [50]. The complete now favored in the choice of targets for immunotherapy [61], disappearance of protection in transgenic mice lacking IFN- γ because the emergence of antigen-loss variants is omnipre- [55] confirms the pivotal role of this cytokine. sent in tumor immunology when the antigen is just a passen- In summary, effective vaccines elicited helper cytokines ger of the neoplastic process. It is clear that, in the course of and protective antibodies that provided long-term protection tumor progression, the driving role can be subsequently from the development of mammary carcinomas driven by taken by different genes [66, 67], which in turn can offer ErbB2. The comparison of immunoprevention with cancer further antigenic targets. We have seen that tumor cells can immunotherapy, which is mainly based on effector T cells, is maintain a high expression of oncogenic antigens and still again remindful of viral immunity, in which cytotoxic T cells evade T cell recognition through downmodulation of antigen cure infection and antibodies, elicited by primary infection processing machinery and MHC, but antibody recognition of or by vaccines, provide long-term prevention of reinfection. accessible molecules (e.g. on the cell surface) is not affected and ensures functional inhibition of the target together with A global view of Table 2 shows that immunoprevention complement mediated cytotoxicity and ADCC. is a general approach that can be pursued with different immunological strategies against a variety of cancer predis- We called oncoantigens those tumor antigens that fulfill positions. The analysis of immune mechanisms indicates that the requirements described above [36], however the term the target antigen is a major determinant of the type of was used also to indicate antigens that satisfy the first protective immune responses, in particular of the balance (essential role in tumor growth), but not the second (target of between B and T cell responses. For example, unlike ErbB2 antibodies). This suggests a hierarchy of the two criteria that transgenic mice (Table 2a), immunoprevention in TRAMP we wish to incorporate here into a general definition of onco- mice (Table 2b) was effected by cytotoxic T cells, because antigens according to their localization, outlined in Table 3. the driving oncogene, SV40 large T, is a nuclear protein Class I oncoantigens correspond to the original definition, mainly recognized by T cells. In the following section, the whereas Class III oncoantigens are those molecules that molecular and immunological nature of the target was used cannot be reached by antibodies because of their intracellular to define a novel classification of tumor antigens relevant to localization. We took the opportunity to explicitely define a cancer immunoprevention. class of extracellular oncoantigens (Class II) that was pre- viously part of the original definition of (Class I) oncoantigens [68]. ONCOANTIGENS - TARGETS FOR CANCER IMMUNOPREVENTION It should be noted that splicing or mutation can give rise to different form of the same gene product pertaining to Many decades of research contribute to current databases different oncoantigenic classes. For example the receptor of tumor antigens, which comprise hundreds of different ALK is a class I oncoantigen expressed on molecules [60]; see also databases at http://www. canceri- the surface of various solid tumors, however in anaplastic mmunity.org. By definition tumor antigens are recognized large cell lymphomas the NPM-ALK translocation gives rise by the immune system of the host, however the variety of to an intracellular class III oncoantigen [69-72]. An even discovery approaches does not guarantee that the immune more extreme case is the idiotype of B cell neoplasms, which response will be in any way significant for the host, and in can belong to secreted (class II), surface (class I) or the field of there is now a need of intracellular (class III) immunoglobulins. additional sorting criteria, to concentrate clinical efforts on promising targets [61]. Cancer immunoprevention poses The three classes of oncoantigens define also a gradient additional constraints, not only because the target cell is of antibody-mediated antitumor mechanisms. Cells expres- preneoplastic or early neoplastic, but also because preclinical sing class I oncoantigens will be prone both to functional results indicate that protective immune responses must be inactivation of the target and to immune cytotoxicity, class II antigens undergo antibody-mediated inactivation, but are not Cancer Immunoprevention Current Drug Targets, 2011, Vol. 12, No. 13 1967 directly conducive to complement or cell-mediated cyto- ventive treatments can be combined with current chemo- toxicity, class III oncoantigens in intact cells are not affected preventive modalities. by specific antibodies, even though the patients can develop The first problem will be in most cases to define the detectable serological responses, presumably against deter- target oncoantigens. Under this respect, preclinical models of minants released by dead tumor cells [71, 73]. carcinogenesis driven by activated oncogenes are different from human cancer syndromes caused by the loss of tumor Table 3. Classification of Oncoantigens suppressor genes, an event which frequently does not pro- duce antigenic targets. It has been shown that vaccines can Oncoantigen Oncogenicity Localization Examples prevent carcinogenesis caused by a combination of oncogene class expression and loss of heterozygosity of a tumor suppressor gene [80-82], therefore the challenge for human immuno- receptors prevention is to investigate the oncoantigens of early G protein- coupled preneoplastic cells, which might represent suitable targets Class I Essential Cell surface receptors even in cancer syndromes fueled by the loss of tumor sup- Death receptors pressor genes [83]. In addition to oncoantigen search strate- Adhesion molecules gies, we will examine two issues relevant to human transla- Mucins tion of cancer immunoprevention, the risks of unwanted side effects and the interrelationships between cancer immuno- Growth factors Class II Essential Extracellular prevention and cancer immunotherapy. Angiogenic factors Non-receptor tyrosine Oncoantigen Search Strategies kinases Class III Essential Intracellular Cell cycle molecules The availability of vast databases and high-throughput Transcription factors methods allow search strategies to pinpoint and rank eligible target antigens for cancer immunoprevention and immuno- therapy [61, 84, 85]. When considering oncoantigens, how- Tumor heterogeneity for several features, including ever, it must be considered that the oncoantigen concept antigenicity, among individual tumors and over time in the embeds a functional property (to have an essential role in same patient represents a critical point also for immune tumor growth) which is not addressed by in silico or purely approaches. So-called cancer-initiating cells, able to self- phenotypic search strategies, unless the oncogenic role of the renew and to reestablish tumor heterogeneity, have been candidate is known a priori on the basis of existing found in most human cancers along with differentiated cells functional studies. [74, 75]. The identification of oncoantigens involved in the To overcome the limitations described above, a novel maintenance of the stem-cell like features opens a promising oncoantigen search strategy was devised combining animal therapeutic opportunity for anti-cancer vaccines. Cancer- models, high throughput technologies and human data [86, initiating cells of most human and mouse carcinomas are 87]. First the transcriptome of mice bearing preneoplastic very efficient in generating tumors when injected in mice, lesions (caused by carcinogens or by genetic defects) is and can be isolated on the basis of their ability to grow as analysed to reveal a set of putative oncoantigens showing non-adherent cell spheroid aggregates that can be propagated increased expression in tumor progression. The subset that for serial in vitro passages [76, 77]. Thus, oncoantigens of has a human orthologue highly expressed in the transcrip- the three classes, selectively expressed by cancer-initiating tome of human neoplastic samples is then used to immunize cells, can be identified by comparing genome-wide trans- mice at risk of cancer, to determine their immunopreventive cription profiles of the tumor with that of succeeding 1 potency. The implementation of this search strategy [87] spheroids passages . yields a wealth of novel candidate oncoantigens that are expressed by human tumors and can be targeted by CANCER IMMUNOPREVENTION: FROM MICE TO immunopreventive vaccines. HUMANS A further criterion that could guide the choice of candi- Promising preclinical results indicate that cancer im- date oncoantigens is the finding of specific antibodies in munoprevention could be fruitfully applied to humans to cancer patients, providing prima facie evidence of immune reduce the risk of cancer. One could imagine that in the recognition and breakage of tolerance [71, 88, 89]. future cancer vaccines might be administered to the general population, as it is happening now to prevent infectious The Risks of Immunoprevention tumors. In a more realistic perspective, there are several Immunoprevention of infectious tumors is expected to be human groups at risk of cancer that could benefit from devoid of significant side effects caused by cross reactivity, specific vaccines, especially for what concerns genetic risk, because target antigens are of viral origin. In some instances, preneoplastic syndromes, cohorts of individuals previously also oncoantigens are selectively expressed in neoplastic exposed to environmental carcinogens and cancer survivors development, for example antigens resulting from oncogenic with increased risk of a new primary tumor (e.g. breast mutations or chromosomal translocations (e.g. NPM-ALK), cancer). A few instances in which combined chemo- aberrant glycosylation (MUC-1) or clonal rearrangements immunopreventive approaches were evaluated in preclinical (idiotypes of B and T cell lymphomas) [90]. In most cases, models [78, 79] illustrate the concept that most immunopre- however, tumor cells express oncoantigens cross-reactive 1968 Current Drug Targets, 2011, Vol. 12, No. 13 Lollini et al. with normal cells, hence the antitumor immune response is thus indicating that cancer progression per se was not the autoimmune in nature, and the possibility of harmful side cause of failure. effects must be taken into account, particularly when long- Fig. (1) outlines the potency of immunological treatments term immunity is sought [36, 85, 91]. relative to cancer progression in the ErbB2 model system. It Most vaccination studies of long-term cancer immuno- can be said that immunoprevention is an effective form of prevention in genetically engineered mice do not show primary, secondary and tertiary prevention, but can be evidence of significant adverse effects. The recent wave of ineffective against established lesions. This knowledge can immune potentiation strategies based on Treg depletion or guide the design of early trials toward clinical situations in inactivation [39, 54, 92, 93] effectively targets one physio- which efficacy is maximized [88, 96, 97]. logical mechanism of immune tolerance, hence it carries an additional risk of autoimmunity [93]. In any case, a clear From Immunotherapy Back to Immunoprevention in distinction must be made between autoimmunity, which is a Humans widespread condition, and , which is activated by specific conditions such as inflammation [94]. The results of preclinical studies mentioned above In general it must be kept in mind that transgenic mice indicate that under appropriate conditions cancer immuno- expressing xenogeneic (e.g. human or rat) genes under the preventive vaccines can display a sizeable therapeutic acti- control of artificial promoters are inappropriate models for vity. This offers an opportunity for first-in-human thera- the analysis of toxic side effects, because the target might peutic studies, because the huge size of current chemopre- not play a role in the normal biology of the host, and ventive studies in human practically forbids a immune responses of uncertain cross-reactivity might not direct transition from preclinical to human prevention. We significantly affect the products of the endogenous gene. A have previously suggested that cancer preventive vaccines more meaningful appraisal of autoimmunity in cancer could follow in the path traced by tamoxifen, which went immunoprevention awaits the analysis of immune responses from an established role in breast cancer therapy to large elicited against endogenous (as opposed to transgenic) chemopreventive studies eventually leading to its approval targets, like some of the studies listed in Table 2c. for breast cancer prevention [98]. Eventually the issue of side effects will have to be add- The dramatic loss of efficacy observed against estab- ressed in human studies (see below). One important deter- lished murine tumors (Fig. 1) suggests that timing of vacci- minant is predicted to be the biological role of target nation relative to cancer progression is of paramount impor- antigens in the adult host, mostly as it happens with ther- tance, and predicts that early clinical studies in advanced apeutical monoclonal antibodies. Oncoantigens like ErbB2 cancer patients, in which most vaccines are currently being or ALK, that have an important role in the embryo, but are tested [99], will probably yield valuable information con- much less relevant during adult life, are likely conducive to cerning vaccine safety, and possibly some indications from milder and less frequent toxicities than other oncoantigens, careful immune monitoring, but are unlikely to provide evi- like EGFR, that play a fundamental role in epithelial cell dence of relevant clinical benefit. A meaningful therapeutic homeostasis. test will be more likely provided by the addition of vaccines to adjuvant therapeutic regimens. Under this respect breast In conclusion, to minimize the risks of harmful auto- cancer expressing estrogen receptor provides a suitable immunity, cancer immunoprevention should preferentially testbed, because most patients are included in long-term (5- target oncoantigens that are either structurally different in 10 years) therapeutic regimens with anti-estrogenic drugs in tumor cells than in normal cells [84, 90], or those that play a which repeated vaccinations could be easily integrated. marginal role in the adult host. The following step will be the first true test of these prophylactic vaccines in humans at risk. Also in this case From Cancer Immunoprevention to Immunotherapy murine models provide compelling clues as to timing that Vaccines developed for prophylactic purposes are not will need to be verified in humans, in particular for what necessarily endowed with therapeutic activity, as we have concerns the activity against early neoplastic lesions. Anti- seen above in the case of HPV vaccines, however the simul- ErbB2 prophylactic vaccines effectively blocked the carcino- taneous presence of cancer preventive and therapeutic genic progression of preneoplastic lesions, such as atypical activities would greatly increase the intrinsic value of a hyperplasia, whereas the development of in situ carcinomas given treatment, and would also ease the transition from and later lesions was delayed, but not blocked. If this preclinical studies to early clinical trials. happens to be the case also in humans, then various early neoplastic conditions would not be amenable to cancer In the ErbB2 transgenic model it was found that the potency of different immunopreventive treatments was immunoprevention. In any case, there are many humans at risk of cancer in which prophylactic cancer vaccines could inversely correlated with the advancement of tumor prog- be usefully tested: people exposed to carcinogens [45, 100], ression. In other words, therapeutic efficacy against estab- for recreational (tobacco), occupational (asbestos, radiations) lished tumors was negligible. This might be caused by an or medical (cancer survivors) reasons. In many instances intrinsic resistance of tumor cells to immune defenses, or to individuals at risk will have either measurable lesions (e.g. the structural and microenvironmental barriers that accom- pany the formation of sizeable tumor masses. It was found oral leukoplakia), or biochemical correlates (e.g. increased PSA level). Finally, a straightforward translation would be to that the Triplex vaccine was effective in the therapy of humans carrying genetic alterations that increase the risk of micrometastases of malignant mammary carcinoma [95], cancer, such as BRCA1/2 families. In most human synd- Cancer Immunoprevention Current Drug Targets, 2011, Vol. 12, No. 13 1969

Primary Secondary Therapy Tertiary prevention prevention prevention 100%

80%

60% f cancer f revention o 40%

20% Efficacy Efficacy immunop 0%

l a a u a s s si m si a sit no a in st Norma a a perpl m Hyperplasi met hy no ve carci y rci si rl ical a a C E Inva Disseminated cell Atyp →→→ TiTumor progression →→→→

Fig. (1). Efficacy of cancer immunopreventive treatments against successive phases of mammary carcinoma progression in Erbb2-trangenic BALBneuT mice. x-axis reports the stage of tumor progression at which Triplex vaccine administration started, while y-axis reports the corresponding efficacy. Data have been summarized from published papers [50, 51, 95, 98]. romes the genetic alterations are loss-of-function mutations HBV = Hepatitis B virus of tumor suppressor genes, which do not directly generate HCC = Hepatocellular carcinoma oncoantigens, therefore the search strategies outlined above will be instrumental in defining suitable antigenic targets in HCV = Hepatitis C virus each case. HIV = Human immune deficiency virus HPV = Human papillomavirus CONCLUSIONS HTLV-1 = Human T-cell leukemia virus-1 Immunoprevention of viral tumors is now implemented IFN- = -Interferon at the population level, thanks to the development of γ γ effective vaccines that are expected to decrease human tumor IL-12 = Interleukin 12 burden in the near future. Immunoprevention of non-viral MAb = Monoclonal antibodies tumors was successful in numerous mouse models of genetic and carcinogen-induced cancer risk, however the translation MDV = Marek’s disease virus of preclinical results to humans requires the definition of MHC = Major histocompatibility complex meaningful target antigens in preneoplasia and early neo- plasia. Rationalization of preclinical data led to the definition NK = Natural killer of oncoantigens, which are optimal targets for cancer im- Tag = SV40 T antigen munoprevention, and of strategies to search for novel im- munological targets in carcinogenesis. The rationale des- TERT = Telomerase cribed here provides a roadmap for the implementation of TLR = Toll-like receptor human immunopreventive endeavors, and could also benefit current cancer immunotherapy. Treg = Regulatory T cells

ABBREVIATIONS ACKNOWLEDGEMENTS ADCC = Antibody dependent cell-mediated cytotoxicity The work of the authors is supported by Associazione Italiana per la Ricerca sul Cancro, Ministero dell'Università e AIDS = Acquired immune deficiency syndrome della Ricerca, Universities of Torino and Bologna, BPV = Bovine papillomavirus Compagnia di San Paolo, Dipartimento di Patologia Sperimentale dell’Università di Bologna (fondo “Pallotti”), CTL = Cytotoxic T lymphocytes Regione Piemonte, Fondazione Carlo Denegri and the Italian EBV = Epstein-Barr virus Ministry of Health. EC-TM = Extracellular-transmembrane domain HBV = Hepatitis B virus FOOTNOTE 1Cavallo F, Forni G. Manuscript in preparation. 1970 Current Drug Targets, 2011, Vol. 12, No. 13 Lollini et al.

REFERENCES [26] De Vuyst H, Clifford G, Li N, Franceschi S. HPV infection in Europe. Eur J Cancer 2009; 45: 2632-9. [1] Dunn GP, Old LJ, Schreiber RD. The three Es of cancer [27] Michels KB, zur Hausen H. HPV vaccine for all. Lancet 2009; 374: immunoediting. Annu Rev Immunol 2004; 22: 329-60. 268-70. [2] Stutman O. Tumor development after 3-methylcholanthrene in [28] Gardasil approved to prevent genital warts in men and boys. AIDS immunologically deficient athymic-nude mice. Science 1974; 183: Patient Care STDS 2009; 23: 991. 534-6. [29] Sankaranarayanan R, Nene BM, Shastri SS, et al. HPV screening [3] Shankaran V, Ikeda H, Bruce AT, et al. IFNgamma and for cervical cancer in rural India. N Engl J Med 2009; 360: 1385- lymphocytes prevent primary tumour development and shape 94. tumour immunogenicity. Nature 2001; 410: 1107-11. [30] Franco EL. A new generation of studies of human papillomavirus [4] Finn OJ. Cancer immunology. N Engl J Med 2008; 358: 2704-15. DNA testing in cervical cancer screening. J Natl Cancer Inst 2009; [5] Cavallo F, De Giovanni C., Nanni P, Forni G, Lollini PL. 2011: the 101: 1600-1. immune hallmarks of cancer. Cancer Immunol Immunother 2011; [31] Rees L, Tizard EJ, Morgan AJ, et al. A phase I trial of epstein-barr 60: 319-26. virus gp350 vaccine for children with chronic kidney disease [6] Hanahan D, Weinberg RA. Hallmarks of cancer: the next awaiting transplantation. Transplantation 2009; 88: 1025-9. generation. Cell 2011; 144: 646-74. [32] Arora S, Czinn SJ. Vaccination as a method of preventing [7] Grulich AE, van Leeuwen MT, Falster MO, Vajdic CM. Incidence Helicobacter pylori-associated gastric cancer. Cancer Epidemiol of cancers in people with HIV/AIDS compared with Biomarkers Prev 2005; 14: 1890-1. immunosuppressed transplant recipients: a meta-analysis. Lancet [33] Balfour HH, Jr. Epstein-Barr virus vaccine for the prevention of 2007; 370: 59-67. infectious mononucleosis--and what else? J Infect Dis 2007; 196: [8] Schulz TF. Cancer and viral infections in immunocompromised 1724-6. individuals. Int J Cancer 2009; 125: 1755-63. [34] Sokal EM, Hoppenbrouwers K, Vandermeulen C, et al. [9] Vajdic CM, van Leeuwen MT. Cancer incidence and risk factors Recombinant gp350 vaccine for infectious mononucleosis: a phase after solid organ transplantation. Int J Cancer 2009; 125: 1747-54. 2, randomized, double-blind, placebo-controlled trial to evaluate [10] Koebel CM, Vermi W, Swann JB, et al. Adaptive immunity the safety, immunogenicity, and efficacy of an Epstein-Barr virus maintains occult cancer in an equilibrium state. Nature 2007; 450: vaccine in healthy young adults. J Infect Dis 2007; 196: 1749-53. 903-7. [35] Nasir L, Campo MS. Bovine papillomaviruses: their role in the [11] Mantovani A, Romero P, Palucka AK, Marincola FM. Tumour aetiology of cutaneous tumours of bovids and equids. Vet Dermatol immunity: effector response to tumour and role of the 2008; 19: 243-54. microenvironment. Lancet 2008; 371: 771-83. [36] Lollini PL, Cavallo F, Nanni P, Forni G. Vaccines for tumour [12] Muller-Hermelink N, Braumuller H, Pichler B, et al. TNFR1 prevention. Nat Rev Cancer 2006; 6: 204-16. signaling and IFN-gamma signaling determine whether T cells [37] Spisek R, Dhodapkar MV. Immunoprevention of cancer. Hematol induce tumor dormancy or promote multistage carcinogenesis. Oncol Clin North Am 2006; 20: 735-50. Cancer Cell 2008; 13: 507-18. [38] Hayakawa Y, Rovero S, Forni G, Smyth MJ. Alpha- [13] Beatty PL, Narayanan S, Gariepy J, Ranganathan S, Finn OJ. galactosylceramide (KRN7000) suppression of chemical- and Vaccine against MUC1 Antigen Expressed in Inflammatory Bowel oncogene-dependent carcinogenesis. Proc Natl Acad Sci USA Disease and Cancer Lessens Colonic Inflammation and Prevents 2003; 100: 9464-9. Progression to Colitis-Associated Colon Cancer. Cancer Prev Res [39] Ambrosino E, Spadaro M, Iezzi M, et al. Immunosurveillance of (Phila Pa) 2010; 3: 438-46. Erbb2 carcinogenesis in transgenic mice is concealed by a [14] Osterrieder N, Kamil JP, Schumacher D, Tischer BK, Trapp S. dominant regulatory T-cell self-tolerance. Cancer Res 2006; 66: Marek's disease virus: from miasma to model. Nat Rev Microbiol 7734-40. 2006; 4: 283-94. [40] Zhou P, Fang X, McNally BA, et al. Targeting lymphotoxin- [15] Baigent SJ, Smith LP, Nair VK, Currie RJ. Vaccinal control of mediated negative selection to prevent prostate cancer in mice with Marek's disease: current challenges, and future strategies to genetic predisposition. Proc Natl Acad Sci USA 2009; 106: 17134- maximize protection. Vet Immunol Immunopathol 2006; 112: 78- 9. 86. [41] Sfondrini L, Besusso D, Rumio C, et al. Prevention of spontaneous [16] Chang MH, You SL, Chen CJ, et al. Decreased incidence of mammary adenocarcinoma in HER-2/neu transgenic mice by hepatocellular carcinoma in hepatitis B vaccinees: a 20-year foreign DNA. FASEB J 2002; 16: 1749-54. follow-up study. J Natl Cancer Inst 2009; 101: 1348-55. [42] Mastini C, Becker PD, Iezzi M, et al. Intramammary application of [17] Guidotti LG, Chisari FV. Immunobiology and pathogenesis of viral non-methylated-CpG oligodeoxynucleotides (CpG) inhibits both hepatitis. Annu Rev Pathol 2006; 1: 23-61. local and systemic mammary carcinogenesis in female BALB/c [18] Strickland GT, El-Kamary SS, Klenerman P, Nicosia A. Hepatitis Her-2/neu transgenic mice. Curr Cancer Drug Targets 2008; 8: C vaccine: supply and demand. Lancet Infect Dis 2008; 8: 379-86. 230-42. [19] Boyle P, Levin B. World Cancer Report. Lyon: IARC 2008. [43] Chin AI, Miyahira AK, Covarrubias A, et al. Toll-like receptor 3- [20] zur Hausen H. Papillomaviruses and cancer: from basic studies to mediated suppression of TRAMP prostate cancer shows the critical clinical application. Nat Rev Cancer 2002; 2: 342-50. role of type I in tumor immune surveillance. Cancer Res [21] Schiller JT, Castellsague X, Villa LL, Hildesheim A. An update of 2010; 70: 2595-603. prophylactic human papillomavirus L1 virus-like particle vaccine [44] Boggio K, Nicoletti G, Di Carlo E, et al. Interleukin 12-mediated clinical trial results. Vaccine 2008; 26 Suppl 10: K53-K61. prevention of spontaneous mammary adenocarcinomas in two lines [22] Ault KA. Effect of prophylactic human papillomavirus L1 virus- of Her-2/neu transgenic mice. J Exp Med 1998; 188: 589-96. like-particle vaccine on risk of cervical intraepithelial neoplasia [45] Noguchi Y, Jungbluth A, Richards EC, Old LJ. Effect of grade 2, grade 3, and adenocarcinoma in situ: a combined analysis interleukin 12 on tumor induction by 3-methylcholanthrene. Proc of four randomised clinical trials. Lancet 2007; 369: 1861-8. Natl Acad Sci USA 1996; 93: 11798-801. [23] Joura EA, Leodolter S, Hernandez-Avila M, et al. Efficacy of a [46] Colombo MP, Trinchieri G. Interleukin-12 in anti-tumor immunity quadrivalent prophylactic human papillomavirus (types 6, 11, 16, and immunotherapy. Cytokine Growth Factor Rev 2002; 13: 155- and 18) L1 virus-like-particle vaccine against high-grade vulval 68. and vaginal lesions: a combined analysis of three randomised [47] Katsumata M, Okudaira T, Samanta A, et al. Prevention of breast clinical trials. Lancet 2007; 369: 1693-702. tumour development in vivo by downregulation of the p185neu [24] Paavonen J, Naud P, Salmeron J, et al. Efficacy of human receptor. Nat Med 1995; 1: 644-8. papillomavirus (HPV)-16/18 AS04-adjuvanted vaccine against [48] Allen SD, Garrett JT, Rawale SV, et al. Peptide vaccines of the cervical infection and precancer caused by oncogenic HPV types HER-2/neu dimerization loop are effective in inhibiting mammary (PATRICIA): final analysis of a double-blind, randomised study in tumor growth in vivo. J Immunol 2007; 179: 472-82. young women. Lancet 2009; 374: 301-14. [49] Stagg J, Sharkey J, Pommey S, et al. Antibodies targeted to TRAIL [25] Bosch FX. Broad-spectrum human papillomavirus vaccines: new receptor-2 and ErbB-2 synergize in vivo and induce an antitumor horizons but one step at a time. J Natl Cancer Inst 2009; 101: 771- immune response. Proc Natl Acad Sci USA 2008; 105: 16254-9. 3. Cancer Immunoprevention Current Drug Targets, 2011, Vol. 12, No. 13 1971

[50] Nanni P, Nicoletti G, De Giovanni C, et al. Combined allogeneic [72] Mastini C, Martinengo C, Inghirami G, Chiarle R. Anaplastic tumor cell vaccination and systemic interleukin 12 prevents lymphoma kinase: an oncogene for tumor vaccination. J Mol Med mammary carcinogenesis in HER-2/neu transgenic mice. J Exp 2009; 87: 669-77. Med 2001; 194: 1195-205. [73] Anderson KS, Wong J, Vitonis A, et al. p53 autoantibodies as [51] De Giovanni C, Nicoletti G, Landuzzi L, et al. Immunoprevention potential detection and prognostic biomarkers in serous ovarian of HER-2/neu transgenic mammary carcinoma through an cancer. Cancer Epidemiol Biomarkers Prev 2010; 19: 859-68. interleukin 12-engineered allogeneic cell vaccine. Cancer Res [74] Reya T, Morrison SJ, Clarke MF, Weissman IL. Stem cells, cancer, 2004; 64: 4001-9. and cancer stem cells. Nature 2001; 414: 105-11. [52] Quaglino E, Iezzi M, Mastini C, et al. Electroporated DNA vaccine [75] Clarke MF, Fuller M. Stem cells and cancer: two faces of eve. Cell clears away multifocal mammary carcinomas in Her-2/neu 2006; 124: 1111-5. transgenic mice . Cancer Res 2004; 64: 2858-64. [76] Dontu G, Abdallah WM, Foley JM, et al. In vitro propagation and [53] Curcio C, Khan AS, Amici A, et al. DNA immunization using transcriptional profiling of human mammary stem/progenitor cells. constant-current electroporation affords long-term protection from Genes Dev 2003; 17: 1253-70. autochthonous mammary carcinomas in cancer-prone transgenic [77] Bussolati B, Grange C, Sapino A, Camussi G. Endothelial cell mice. Cancer Gene Ther 2008; 15: 108-14. differentiation of human breast tumour stem/progenitor cells. J Cell [54] Rolla S, Ria F, Occhipinti S, et al. ErbB2 DNA vaccine combined Mol Med 2009; 13: 309-19. with Treg cell deletion enhances antibody response and reveals [78] Nanni P, Nicoletti G, De Giovanni C, et al. Prevention of HER- latent low-avidity T cells. Potential and limits of its therapeutic 2/neu transgenic mammary carcinoma by tamoxifen plus efficacy. J Immunol 2010; 184: 6124-32. interleukin 12. Int J Cancer 2003; 105: 384-9. [55] Nanni P, Landuzzi L, Nicoletti G, et al. Immunoprevention of [79] De Giovanni C, Nicoletti G, Landuzzi L, et al. Tamoxifen mammary carcinoma in HER-2/neu transgenic mice is IFN-gamma combined to anti-HER-2/neu cell vaccine does not hamper cancer and B cell dependent. J Immunol 2004; 173: 2288-96. immunopreventive efficacy. Vaccine 2009; 27: 2065-9. [56] Park JM, Terabe M, Sakai Y, et al. Early role of CD4+ Th1 cells [80] Croci S, Nicoletti G, Landuzzi L, et al. Immunological prevention and antibodies in HER-2 adenovirus vaccine protection against of a multigene cancer syndrome. Cancer Res 2004; 64: 8428-34. autochthonous mammary carcinomas. J Immunol 2005; 174: 4228- [81] Pannellini T, Spadaro M, Di CE, et al. Timely DNA vaccine 36. combined with systemic IL-12 prevents parotid carcinomas before [57] Quaglino E, Rolla S, Iezzi M, et al. Concordant morphologic and a dominant-negative p53 makes their growth independent of HER- gene expression data show that a vaccine halts HER-2/neu 2/neu expression. J Immunol 2006; 176: 7695-703. preneoplastic lesions. J Clin Invest 2004; 113: 709-17. [82] De Giovanni C, Nicoletti G, Palladini A, et al. A multi-DNA [58] Porzia A, Lanzardo S, Citti A, et al. Attenuation of PI3K/Akt- preventive vaccine for p53/Neu-driven cancer syndrome. Hum mediated tumorigenic signals through PTEN activation by DNA Gene Ther 2009; 20: 453-64. vaccine-induced anti-ErbB2 antibodies. J Immunol 2010; 184: [83] Finn OJ. Premalignant lesions as targets for cancer vaccines. J Exp 4170-7. Med 2003; 198: 1623-6. [59] Spadaro M, Ambrosino E, Iezzi M, et al. Cure of mammary [84] Warren RL, Holt RA. A census of predicted mutational epitopes carcinomas in Her-2 transgenic mice through sequential stimulation suitable for immunologic cancer control. Hum Immunol 2010; 71: of innate (neo-adjuvant IL-12) and adaptive (DNA vaccine 245-54. electroporation) immunity. Clin Cancer Res 2005; 11: 1941-52. [85] Kurosawa Y. Comprehensive Identification of Tumor-associated [60] Novellino L, Castelli C, Parmiani G. A listing of human tumor Antigens via Isolation of Human Monoclonal Antibodies that may antigens recognized by T cells: March 2004 update. Cancer be Therapeutic. Immune Netw 2009; 9: 4-7. Immunol Immunother 2005; 54: 187-207. [86] Cavallo F, Calogero RA, Forni G. Are oncoantigens suitable targets [61] Cheever MA, Allison JP, Ferris AS, et al. The prioritization of for anti-tumour therapy? Nat Rev Cancer 2007; 7: 707-13. cancer antigens: a national cancer institute pilot project for the [87] Calogero RA, Quaglino E, Saviozzi S, Forni G, Cavallo F. acceleration of translational research. Clin Cancer Res 2009; 15: Oncoantigens as anti-tumor vaccination targets: the chance of a 5323-37. lucky strike? Cancer Immunol Immunother 2008; 57: 1685-94. [62] Garrido F, Algarra I. MHC antigens and tumor escape from [88] Disis ML, Wallace DR, Gooley TA, et al. Concurrent trastuzumab immune surveillance. Adv Cancer Res 2001; 83: 117-58. and HER2/neu-specific vaccination in patients with metastatic [63] Seliger B. Molecular mechanisms of MHC class I abnormalities breast cancer. J Clin Oncol 2009; 27: 4685-92. and APM components in human tumors. Cancer Immunol [89] Reipert BM, Tanneberger S, Pannetta A, et al. Increase in Immunother 2008; 57: 1719-26. autoantibodies against Fas (CD95) during carcinogenesis in the [64] Nanni P, Pupa SM, Nicoletti G, et al. p185(neu) protein is required human colon: a hope for the immunoprevention of cancer? Cancer for tumor and anchorage-independent growth, not for cell Immunol Immunother 2005; 54: 1038-42. proliferation of transgenic mammary carcinoma. Int J Cancer 2000; [90] Finn OJ. Immunological weapons acquired early in life win battles 87: 186-94. with cancer late in life. J Immunol 2008; 181: 1589-92. [65] Weinstein IB, Joe A. Oncogene addiction. Cancer Res 2008; 68: [91] Hrelia P, Tanneberger S. Immunoprevention of cancer poses a 3077-80. challenge to pharmacological research. Pharmacol Res 1997; 35: [66] Kmieciak M, Knutson KL, Dumur CI, Manjili MH. HER-2/neu 391-401. antigen loss and relapse of mammary carcinoma are actively [92] Nava-Parada P, Forni G, Knutson KL, Pease LR, Celis E. Peptide induced by T cell-mediated anti-tumor immune responses. Eur J vaccine given with a Toll-like receptor agonist is effective for the Immunol 2007; 37: 675-85. treatment and prevention of spontaneous breast tumors. Cancer Res [67] Jones KL, Buzdar AU. Evolving novel anti-HER2 strategies. 2007; 67: 1326-34. Lancet Oncol 2009; 10: 1179-87. [93] Jacob JB, Kong YC, Nalbantoglu I, Snower DP, Wei WZ. Tumor [68] Lollini PL, Nicoletti G, Landuzzi L, De Giovanni C, Nanni P. New regression following DNA vaccination and regulatory T cell target antigens for cancer immunoprevention. Curr Cancer Drug depletion in neu transgenic mice leads to an increased risk for Targets 2005; 5: 221-8. autoimmunity. J Immunol 2009; 182: 5873-81. [69] Palmer RH, Vernersson E, Grabbe C, Hallberg B. Anaplastic [94] Bridle BW, Li J, Jiang S, et al. Immunotherapy can reject lymphoma kinase: signalling in development and disease. Biochem intracranial tumor cells without damaging the brain despite sharing J 2009; 420: 345-61. the target antigen. J Immunol 2010; 184: 4269-75. [70] Chiarle R, Martinengo C, Mastini C, et al. The anaplastic [95] Nanni P, Nicoletti G, Palladini A, et al. Antimetastatic activity of a lymphoma kinase is an effective oncoantigen for lymphoma preventive . Cancer Res 2007; 67: 11037-44. vaccination. Nat Med 2008; 14: 676-80. [96] Driggers L, Zhang JG, Newcomb EW, et al. Immunotherapy of [71] Ait-Tahar K, Damm-Welk C, Burkhardt B, et al. Correlation of the pediatric brain tumor patients should include an immunoprevention autoantibody response to the ALK oncoantigen in pediatric strategy: a medical hypothesis paper. J Neurooncol 2010; 97: 159- anaplastic lymphoma kinase-positive anaplastic large cell 69. lymphoma with tumor dissemination and relapse risk. Blood 2010; [97] Ryu WS, Son GS. Cancer vaccines targeting HER2/neu for early 115: 3314-9. breast cancer. J Breast Cancer 2010; 13: 5-13. 1972 Current Drug Targets, 2011, Vol. 12, No. 13 Lollini et al.

[98] Lollini PL, De Giovanni C, Pannellini T, et al. Cancer [119] Jacob J, Radkevich O, Forni G, et al. Activity of DNA vaccines immunoprevention. Future Oncol 2005; 1: 57-66. encoding self or heterologous Her-2/neu in Her-2 or neu transgenic [99] Ladjemi MZ, Jacot W, Chardes T, Pelegrin A, Navarro-Teulon I. mice. Cell Immunol 2006; 240: 96-106. Anti-HER2 vaccines: new prospects for breast cancer therapy. [120] Holmgren L, Ambrosino E, Birot O, et al. A DNA vaccine Cancer Immunol Immunother 2010; 59: 1295-312. targeting angiomotin inhibits angiogenesis and suppresses tumor [100] Berta GN, Sprio AE, Iezzi M, et al. A DNA Vaccine against growth. Proc Natl Acad Sci USA 2006; 103: 9208-13. ERBB2 Impairs Chemical Carcinogenesis in Random-Bred [121] Masuelli L, Marzocchella L, Focaccetti C, et al. Local delivery of Hamsters. Cancer Prev Res (Phila) 2011; 4: 994-1001. recombinant vaccinia virus encoding for neu counteracts growth of [101] Harper DM, Franco EL, Wheeler CM, et al. Sustained efficacy up mammary tumors more efficiently than systemic delivery in neu to 4.5 years of a bivalent L1 virus-like particle vaccine against transgenic mice. Cancer Immunol Immunother 2010; 59: 1247-58. human papillomavirus types 16 and 18: follow-up from a [122] Tegerstedt K, Lindencrona JA, Curcio C, et al. A single randomised control trial. Lancet 2006; 367: 1247-55. vaccination with polyomavirus VP1/VP2Her2 virus-like particles [102] Reilly RT, Gottlieb MB, Ercolini AM, et al. HER-2/neu is a tumor prevents outgrowth of HER-2/neu-expressing tumors. Cancer Res rejection target in tolerized HER-2/neu transgenic mice. Cancer 2005; 65: 5953-7. Res 2000; 60: 3569-76. [123] Gallo P, Dharmapuri S, Nuzzo M, et al. Adenovirus vaccination [103] Nabekura T, Nagasawa T, Nakauchi H, Onodera M. An against neu oncogene exerts long-term protection from immunotherapy approach with dendritic cells genetically modified tumorigenesis in BALB/neuT transgenic mice. Int J Cancer 2007; to express the tumor-associated antigen, HER2. Cancer Immunol 120: 574-84. Immunother 2008; 57: 611-22. [124] Quaglino E, Mastini C, Amici A, et al. A better immune reaction to [104] Esserman LJ, Lopez T, Montes R, et al. Vaccination with the Erbb-2 tumors is elicited in mice by DNA vaccines encoding extracellular domain of p185neu prevents mammary tumor rat/human chimeric proteins. Cancer Res 2010; 70: 2604-12. development in neu transgenic mice. Cancer Immunol Immunother [125] Zeng J, Cai S, Yi Y, et al. Prevention of spontaneous tumor 1999; 47: 337-42. development in a ret transgenic mouse model by ret peptide [105] Dakappagari NK, Douglas DB, Triozzi PL, Stevens VC, Kaumaya vaccination with indoleamine 2,3-dioxygenase inhibitor 1-methyl PT. Prevention of mammary tumors with a chimeric HER-2 B-cell tryptophan. Cancer Res 2009; 69: 3963-70. epitope peptide vaccine. Cancer Res 2000; 60: 3782-9. [126] Ye X, McCarrick J, Jewett L, Knowles BB. Timely immunization [106] Manjili MH, Wang XY, Chen X, et al. HSP110-HER2/neu subverts the development of peripheral nonresponsiveness and chaperone complex vaccine induces protective immunity against suppresses tumor development in simian virus 40 tumor antigen- spontaneous mammary tumors in HER-2/neu transgenic mice. J transgenic mice. Proc Natl Acad Sci USA 1994; 91: 3916-20. Immunol 2003; 171: 4054-61. [127] De Giovanni C, Croci S, Nicoletti G, et al. Inhibition of prostate [107] Pupa SM, Iezzi M, Di Carlo E, et al. Inhibition of mammary carcinogenesis by combined active immunoprophylaxis. Int J carcinoma development in HER-2/neu transgenic mice through Cancer 2007; 121: 88-94. induction of autoimmunity by xenogeneic DNA vaccination. [128] Degl'Innocenti E, Grioni M, Boni A, et al. Peripheral T cell Cancer Res 2005; 65: 1071-8. tolerance occurs early during spontaneous prostate cancer [108] Cefai D, Morrison BW, Sckell A, et al. Targeting HER-2/neu for development and can be rescued by dendritic cell immunization. active-specific immunotherapy in a mouse model of spontaneous Eur J Immunol 2005; 35: 66-75. breast cancer. Int J Cancer 1999; 83: 393-400. [129] Willimsky G, Blankenstein T. Sporadic immunogenic tumors avoid [109] Wagner S, Jasinska J, Breiteneder H, et al. Delayed tumor onset destruction by inducing T cell tolerance. Nature 2005; 437: 141-6. and reduced tumor growth progression after immunization with a [130] Stahl S, Sacher T, Bechtold A, et al. Tumor agonist peptides break Her-2/neu multi-peptide vaccine and IL-12 in c-neu transgenic tolerance and elicit effective CTL responses in an inducible mouse mice. Breast Cancer Res Treat 2007; 106: 29-38. model of hepatocellular carcinoma. Immunol Lett 2009; 123: 31-7. [110] Amici A, Venanzi FM, Concetti A. Genetic immunization against [131] Stewart TJ, Abrams SI. Altered immune function during long-term neu/erbB2 transgenic breast cancer. Cancer Immunol Immunother host-tumor interactions can be modulated to retard autochthonous 1998; 47: 183-90. neoplastic growth. J Immunol 2007; 179: 2851-9. [111] Smorlesi A, Papalini F, Amici A, et al. Evaluation of different [132] Papewalis C, Wuttke M, Seissler J, et al. Dendritic cell vaccination plasmid DNA delivery systems for immunization against with xenogenic polypeptide hormone induces tumor rejection in HER2/neu in a transgenic murine model of mammary carcinoma. neuroendocrine cancer. Clin Cancer Res 2008; 14: 4298-305. Vaccine 2006; 24: 1766-75. [133] Wuttke M, Papewalis C, Meyer Y, et al. Amino acid-modified [112] Wang X, Wang JP, Maughan MF, Lachman LB. Alphavirus calcitonin immunization induces tumor epitope-specific immunity replicon particles containing the gene for HER2/neu inhibit breast in a transgenic mouse model for medullary thyroid carcinoma. cancer growth and tumorigenesis. Breast Cancer Res 2005; 7: Endocrinology 2008; 149: 5627-34. R145-R155. [134] Mennuni C, Ugel S, Mori F, et al. Preventive vaccination with [113] Masuelli L, Focaccetti C, Cereda V, et al. Gene-specific inhibition telomerase controls tumor growth in genetically engineered and of breast carcinoma in BALB-neuT mice by active immunization carcinogen-induced mouse models of cancer. Cancer Res 2008; 68: with rat Neu or human ErbB receptors. Int J Oncol 2007; 30: 381- 9865-74. 92. [135] Garcia-Hernandez ML, Gray A, Hubby B, Klinger OJ, Kast WM. [114] Sakai Y, Morrison BJ, Burke JD, et al. Vaccination by genetically Prostate stem cell antigen vaccination induces a long-term modified dendritic cells expressing a truncated neu oncogene protective immune response against prostate cancer in the absence prevents development of breast cancer in transgenic mice. Cancer of autoimmunity. Cancer Res 2008; 68: 861-9. Res 2004; 64: 8022-8. [136] Iinuma T, Homma S, Noda T, et al. Prevention of gastrointestinal [115] Steel JC, Ramlogan CA, Yu P, et al. Interleukin-15 and its receptor tumors based on adenomatous polyposis coli gene mutation by augment dendritic cell vaccination against the neu oncogene dendritic cell vaccine. J Clin Invest 2004; 113: 1307-17. through the induction of antibodies partially independent of CD4 [137] Vella LA, Yu M, Phillips AB, Finn OJ. Immunity against cyclin B1 help. Cancer Res 2010; 70: 1072-81. tumor antigen delays development of spontaneous cyclin B1- [116] Garrett JT, Rawale S, Allen SD, et al. Novel engineered positive tumors in p53 (-/-) mice. Ann N Y Acad Sci 2009; 1174: trastuzumab conformational epitopes demonstrate in vitro and in 68-73. vivo antitumor properties against HER-2/neu. J Immunol 2007; [138] Vogt A, Chuang PT, Hebert J, et al. Immunoprevention of basal 178: 7120-31. cell carcinomas with recombinant hedgehog-interacting protein. J [117] Rovero S, Amici A, Carlo ED, et al. DNA vaccination against rat Exp Med 2004; 199: 753-61. her-2/Neu p185 more effectively inhibits carcinogenesis than [139] Greiner JW, Zeytin H, Anver MR, Schlom J. Vaccine-based transplantable carcinomas in transgenic BALB/c mice. J Immunol therapy directed against demonstrates 2000; 165: 5133-42. antitumor activity on spontaneous intestinal tumors in the absence [118] Cho HI, Niu G, Bradley N, Celis E. Optimized DNA vaccines to of autoimmunity. Cancer Res 2002; 62: 6944-51. specifically induce therapeutic CD8 T cell responses against [140] Zeytin HE, Patel AC, Rogers CJ, et al. Combination of a poxvirus- autochthonous breast tumors. Cancer Immunol Immunother 2008; based vaccine with a cyclooxygenase-2 inhibitor (celecoxib) elicits 57: 1695-703. Cancer Immunoprevention Current Drug Targets, 2011, Vol. 12, No. 13 1973

antitumor immunity and long-term survival in CEA.Tg/MIN mice. [142] Xia J, Tanaka Y, Koido S, et al. Prevention of spontaneous breast Cancer Res 2004; 64: 3668-78. carcinoma by prophylactic vaccination with dendritic/tumor fusion [141] Gritzapis AD, Voutsas IF, Lekka E, Papamichail M, Baxevanis cells. J Immunol 2003; 170: 1980-6. CN. Peptide vaccination breaks tolerance to HER-2/neu by [143] Mukherjee P, Basu GD, Tinder TL, et al. Progression of pancreatic generating vaccine-specific FasL(+) CD4(+) T cells: first evidence adenocarcinoma is significantly impeded with a combination of for intratumor apoptotic regulatory T cells. Cancer Res 2010; 70: vaccine and COX-2 inhibition. J Immunol 2009; 182: 216-24. 2686-96.

Received: April 22, 2010 Revised: July 15, 2010 Accepted: July 18, 2010

PMID: 21158706