Tumor immunology, immunomics and targeted immunotherapy for ...

Tumor immunology, immunomics and targeted immunotherapy for central nervous system malignancies Haumith R. Khan-Farooqi*, Robert M. Prins* and Linda M. Liau*{,{

Published by Maney Publishing (c) W. S. Maney & Son Limited

*Department of Neurosurgery, {the Jonsson Comprehensive Cancer Center and {the Brain Research Institute, David Geffen School of Medicine at UCLA, University of California at Los Angeles, Los Angeles, California, USA

Although the brain was traditionally considered as ‘immunologically privileged’, recent findings have implied an involvement of immune mechanisms in neurological disease and illness, including central nervous system (CNS) malignancies. In this review, we initially focus on aspects of the immune system critical for effective antitumor immunity, as an understanding of normal immunological functions and how they relate to tumor immunology will set a foundation for understanding the unique challenges facing the integration of neuro-oncology and neuroimmunology. We summarize current knowledge of immune responses in the ‘immunologically quiescent’ brain and its role in tumor immunology. We will then discuss the emerging field of ‘immunomics’ and recent advances in molecular technologies, such as DNA microarray, which are being applied to brain tumor antigen epitope discovery and patient stratification for brain cancer immunotherapy. This, in turn, should have significant importance for ultimately designing and developing efficient and focused strategies for anticancer immunotherapy. Finally, the current state of immune-based treatment paradigms and future directions will be discussed, paying particular attention to targeted antibody strategies, adoptive cellular immunotherapy, and tumor vaccine approaches that have been studied in clinical trials for CNS neoplasms. [Neurol Res 2005; 27: 692–702] Keywords: Brain tumor; cancer vaccines; immunomics; immunotherapy; microarray

INTRODUCTION Tumors arising within the central nervous system (CNS) present the immune system with challenging targets to recognize and eradicate. Anatomical constraints and distinctive immunoreactivity provide a complex microenvironment within the CNS. Tumor heterogeneity and immune defects seen in tumor-bearing hosts add to this complexity, comprising significant impediments for immune based therapies. The relative absence of immune reactivity in the brain and the dismal prognosis of patients with glioblastomas has often been ascribed to the ‘immune privilege’ of the CNS. However, our current understanding of neuroimmunology now reveals that immune responses can and do occur within the CNS and brain tumors (Figure 1). Immunotherapy for cancer represents a particularly appealing form of treatment because of its potential for highly specific antitumor cytotoxicity. This review will initially focus on aspects of the immune system critical for effective antitumor immunity. An understanding of normal immunological functions and how they relate to tumor immunology Correspondence and reprint requests to: Linda M. Liau, MD, PhD, UCLA Division of Neurosurgery CHS 74-145, Box 956901 10833, Le Conte Avenue, Los Angeles, CA 90095-6901, USA. [lliau@mednet. ucla.edu] Accepted for publication June 2005.

692 Neurological Research, 2005, Volume 27, October

will set a foundation for appreciating the unique challenges that brain tumor researchers currently face. We will then discuss the emerging field of ‘immunomics’ and recent advances in molecular technologies that are being applied to brain tumor antigen discovery and patient stratification for immunotherapy. This, in turn, should have significant importance for ultimately designing immunotherapeutic strategies in the future. INNATE AND ADAPTIVE IMMUNE RESPONES Innate immunity The innate immune response against neoplastic transformation is comprised of natural killer (NK) cells, macrophages (microglial cells in the CNS) and granulocytes. The innate immune response is thought to be critical for stunting the initial burst of tumor cell proliferation during the period of 4–7 days needed to generate an adaptive immune response. Directed against a broad range of targets, the innate immune system detects antigens known as pathogen-associated molecular patterns (PAMP). PAMP are conserved bacterial, viral, parasitic or tumor-derived molecules that are associated with some form of cellular stress1,2. Some examples of PAMP include lipopolysaccharides (Gram-negative bacteria), lipoteichoic acid or mannose # 2005 W. S. Maney & Son Ltd

10.1179/016164105X49490


Central nervous system malignancies: Haumith R. Khan-Farooqi et al.

CNS), granulocytes (neutrophils, eosinophils and basophils) and natural killer cells3. Upon binding of PAMP, TLR activate cells and elicit potent pro-inflammatory responses (e.g. cytokine release), activating other immune cells that are ultimately responsible for antigen destruction and removal. For example, a bacteriumrelated antigen may bind to TLR and cause the secretion of factors that induce lysis and phagocytosis. Both of these are accompanied by the release of IL-1, IL-6, IL12, tumor necrosis factor-alpha (TNF-a), oxygen radicals, nitric oxide, prostaglandins and leukotrines. TLR recognition of virally derived antigens, however, produces factors that induce cellular apoptosis and the release of type I interferons. In either circumstance, these secreted factors also upregulate DC activity by enhancing their migration to the lymphoid tissue and increasing major histocompatibility complex (MHC) presentation, as well as activating a Th1 response. In effect, the TLR-induced innate immune response serves to facilitate the efficient priming of the adaptive immune response. Adaptive immune response The adaptive immune response consists of two subtypes, the humoral and cellular immune responses. Both types are highly specific reactions consisting of four stages4: recognition/activation, proliferation, effector and memory. While both the humoral and cellular immune responses may act cooperatively and simultaneously during adaptive immunity, each one is suited for the removal of particular forms of antigen.

Humoral immune response

Figure 1: Photomicrographs of immunohistochemistry using antiCD4 (A) and anti-CD8 (B) antibodies, showing scattered perivascular T cells within a gliobastoma. Immunohistochemistry against the pan T-cell marker CD45RO (C) shows that T cells can be present with higher density in the perivascular space and infiltrate among glioblastoma tumor cells more extensively than in normal brain tissue. Original magnification 6200

(Gram-positive bacteria), mannan (fungi), viral envelopes, and heat shock proteins (in neoplastic cells)1. Primary transmembrane proteins, known as toll-like receptors, are expressed on the surfaces of immune cells and are critical for PAMP detection. The origin of the name ‘toll-like receptors’ (TLR) came from the family of Drosophila proteins called ‘Toll’, which were initially discovered as developmental genes in flies. Toll was later found to have antimicrobial immune functions by protecting Drosophila against fungal infections. Around the same time, it was shown that a similar protein of immunologic significance in mammals also signaled by way of toll-related mechanisms. Hence, the name ‘tolllike receptors’ was used to refer to the mammalian homologs of Toll. TLR can be found on tissue dendritic cells (DC), macrophages, monocytes (microglia in

The primary components of the humoral immune response are B cells and antibody-secreting plasma cells (Figure 2). CD4z helper T cells often assist with the humoral response. Naı¨ve CD4z Th cells (Th0) may differentiate into either CD4z Th1 cells or CD4z Th2 cells. The primary role of both CD4z subclasses is the release of cytokines that regulate the function of other immune cells. Th2 cells secrete the B cell-stimulating cytokines IL-4 and IL-10 after antigen-presenting cell (APC) interaction. Upon recognition of their specific antigen by B cell receptors on the cell surface, stimulated B cells will differentiate into antibodysecreting plasma cells. Antibodies can then eliminate antigen via complement-dependent lysis, antibodydependent cell mediated cytotoxicity and/or opsonization of antigen-coated particles. Antibodies detect free and native antigen, making them effective against bacteria or other foreign antigens. Because they lack the ability to cross the blood–brain barrier (BBB) and enter the brain parenchyma, antibodies are not usually efficient eliminators of tumor antigens in the CNS unless they are adoptively transferred directly into the brain. A subpopulation of activated, antigen-specific B lymphocytes also differentiate into long-lived memory cells. These memory B cells rapidly differentiate into antibody-secreting plasma cells when antigen is Neurological Research, 2005, Volume 27, October

693



Figure 2: Schematic summarizing the Th2 CD4z helper T cell-mediated response, which stimulates B cells, memory cells and antibodysecreting plasma cells

encountered later. The humoral response is critical for defense against the majority of extracellular pathogens, but does not provide effective immunity to intracellular pathogens within tumor cells.

Cellular immune response The cellular immune response exerts its effects primarily through CD4z Th1 helper T cells and CD8z cytotoxic T lymphocytes (CTL) (Figure 3). Unlike its humoral counterpart, the cellular immune response is adapted for the elimination of antigens derived from the cytoplasm of cells. Activated CD8z CTL precursors differentiate and gain the ability to directly lyse tumor or infected cells. Each form of T cell (CTL versus Th) can only recognize antigens in a particular format. This format is created by the MHC, which has two distinct classes—class I and class II. In terms of presenting antigen, the vast majority of nucleated cells express MHC class I molecules. CD8z T cells are ‘restricted’ to the recognition of antigens in the context of MHC class I, which is conferred by the differences in the expressed form of the T cell receptor. Also known as the direct pathway, MHC class I carries peptides of 8–12 amino acids in length. These peptides are derived from endogenous antigens such as cytoplasmic tumor proteins and viral pathogens. Th cells, however, are restricted to antigens in the context of MHC class II, 694 Neurological Research, 2005, Volume 27, October

which are normally related to exogenous antigens and are 12–20 amino acids in length. MHC class II complexes are commonly found on APC or macrophage cell surfaces; and this form of recognition is known as the indirect pathway, because the effector arm acts indirectly through the immune regulation caused by the Th cells. Naı¨ve T cells of both types continuously circulate through the blood and lymphatics, sampling all antigens within the environment through interactions with APC. Derived from hematopoetic progenitors, DC are professional APC situated at sites of interface with the environment. In the absence of inflammatory stimuli, they continuously sample both self and non-self. In the presence of pro-inflammatory signals or ‘danger signals’5, however, DC become activated and increase the uptake of antigens, which are loaded in the context of MHC. Commonly known danger signals include lipopolysaccharides (LPS), viral dsRNA, CpG oligonucleotides, mechanical stress and inflammatory cytokines such as IL-1, IL-2 and TNF-a6,7. Danger signals may lower the threshold for overcoming peripheral tolerance in self antigen-specific T cells, as well as increase DC trafficking, survival and cross presentation. Mature DC lower antigen uptake and upregulate the adhesion and co-stimulatory molecules. After maturation, DC migrate to the paracortical regions of the lymph nodes, where they can encounter naı¨ve CD8z and CD4z T cells.



Figure 3: Schematic summarizing the Th1 CD4z helper T cell-mediated response, which stimulates macrophages and CTL

Here, activation of DC induces cross presentation, increasing MHC expression and trafficking to T cell areas of lymphoid tissue, where DC can initiate effector T cell responses. The ‘danger signal’ mechanism is not the only safeguard that the immune system has against harmful autoimmunity. Because a breakdown in self-tolerance can be lethal, especially within the CNS, several other mechanisms exist to preserve peripheral tolerance. For instance, only the concomitant stimulation of at least two receptor types on the T cell will drive differentiation into effector cells (Figure 4). The first signal arises when antigens, in the context of the appropriate MHC, bind T cell receptors (TCR) on the surfaces of CD4z and CD8z cells. The second signal is derived from ligands on the APC (i.e. B7 for CTL and CD40 for Th cells). Binding of B7 by receptors on the surface of naı¨ve T cells (e.g. CD28 for CD8z T cells and CD40 for CD4z T cells) with concomitant TCR activation results in cytokine release, clonal proliferation and subsequent differentiation to cytotoxic T cells. These differentiated CTL are then ready to exert their effects as the effector arm of the cellular immune response. In the absence of costimulation, however, T cells can develop tolerance to the specific antigen bound to the T cell receptor, and the T cell can become unresponsive or ‘anergic’ to that antigen. To further guard against deleterious autoimmunity, a unique class of T cells, known as regulatory T cells (Treg), also help maintain peripheral self-tolerance. Treg are distinctive in that they express high levels of

negative co-stimulatory molecules, which attenuate activation of CTL. For example, Treg express cytotoxic T-lymphocyte antigen 4 (CTLA-4), which has a higher affinity for CD28 than does B7, but does not costimulate T cell activation upon binding. This disrupts normal, positive co-stimulation and increases T cell anergy. Treg also constitutively express CD25 and express high levels of FoxP3, which acts as a transcription factor that is thought to direct the cellular machinery responsible for the tolerance mechanism8,9. TUMOR IMMUNOLOGY AND THE CENTRAL NERVOUS SYSTEM Once tolerance has been overcome and the effector arm activated, the immune system should theoretically eliminate tumor antigens. In the CNS, however, special environmental circumstances may hinder the actions of the typical progression of antitumor immune responses (Table 1). Although there are currently several examples of successful immunotherapeutic approaches for tumors outside the CNS, the unique properties of the brain have made the direct application of basic immunologic principles more challenging for tumor rejection within the CNS. Such unique immunological properties of the brain initially led researchers to refer to the CNS as ‘immunologically privileged’. However, it is now widely accepted that the CNS should be thought of more as an ‘immunologically quiescent’ site that is unable to initiate immune responses efficiently, rather than a site that is completely devoid of effector immune responses altogether10,11. Neurological Research, 2005, Volume 27, October

695



Figure 4: Schematic view of the antigen-specific interactions between cytotoxic T lymphocytes, natural killer cells and tumor cells. Upon antigen presentation/recognition, the concomitant stimulation of at least two receptor types on the T cell (e.g. TCR/MHC 1 and CD28/B7) is necessary to drive activation and clonal proliferation of effector CTL

located in CNS, therefore preventing interactions of systemically activated immune cells with brain tumor cells. The initiation of an antitumor immune response may be further hindered by a lack of recognized tumor rejection antigens, lack of MHC expression and lack of TAP (transporter associated with antigen processing) molecules and professional APC within the brain. Furthermore, even if an antitumor immune response is initiated, the absence of an organized lymphatic system

The effective initiation of an antitumor immune response requires tumor antigens to be presented to immune cells. Tumors that arise within the confines of the CNS may have anatomical and biological constraints that limit such antigen presentation. Such limitations include the BBB, which, through tight junctions and glial cell barriers, regulates the inflow of immunological materials to a great extent. This may allow insufficient access of immune cells to tumors Table 1: Possible mechanisms of immune escape by tumors in the CNS Mechanism

Examples

Tumor antigens are not presented to immune cells

Lack Lack Lack Lack

Tumor antigens are presented, but do not elicit an immune response

Tumor peptide density below the threshold level required for T cell activation Oncofetal antigen with established immune tolerance Co-stimulatory signal absent or with inhibitory action (e.g. Treg) Appearance of new epitopes that induce tolerance instead of immunogenicity

Lack of efficacy of systemic immune response

Insufficient access of immune cells to CNS

Tumor microenvironment suppresses immune responses

Soluble factors (TGF-b2, IL-10, PGE2, etc.) Cell–cell contact (Fas/FasL, galectin)


of of of of

tumor antigen(s) on tumor cells MHC class I or II molecules in brain parenchyma TAP (transporter associated with antigen processing) molecules professional APC in local CNS tumor environment



within the CNS may hinder further immune activity downstream of initiation. In addition to the constraints imposed by the anatomical segregation of the CNS itself, malignant brain tumors may also induce defects in adaptive antitumor immune responses. Thus, although the CNS can demonstrate the presence of effector immune responses, there are special mechanisms that work antagonistically to hamper the efficacy of such responses. These include immunosuppressive soluble factors (e.g. TGF-b2, IL-10 or PGE2)12,13 or cell–cell interactions (e.g. Fas/FasL, galectin-1) that lead to T cell death14,15. There is evidence that profound local and systemic immune defects can be observed in patients with CNS malignancies, which may further exacerbate brain tumorigenesis and progression. In spite of these numerous obstacles, there is recent evidence that the immune system can overcome the immunoreactivity limitations of the CNS during illness and disease (i.e. multiple sclerosis, encephalitis, tumor and transplantion)16. The exact mechanisms by which this change in immunoreactivity occurs are yet unknown. However, the possibility that the immune system can mediate interactions with the CNS during disease, such as in the development of brain tumors, validates further investigation of immunotherapeutic modes of treatment for brain neoplasms. BRAIN TUMOR ‘IMMUNOMICS’ AND REVERSE IMMUNOLOGY To maximize the utility of brain tumor immunotherapy, however, the identification of relevant tumor rejection antigens is needed. The unveiling of the human genome sequence, along with rapid advances in genomic/ proteomic expression profiling, bioinformatics tools, and immunological assay methods, have now made it possible to screen large numbers of tumor-associated proteins for immunogenic epitopes. These technological advances in ‘‘immunomics’’ will likely lead to the design of more rational and specific strategies for CNS immunotherapy in the future. Microarray technology and tumor-associated antigens (‘reverse immunology’) Despite immune escape mechanisms elaborated by brain neoplasms (Table 1), the concept of effective anti-glioma immune responses has been bolstered by the recent identification of several tumor-associated antigens (TAA) expressed by human brain tumors (Table 2). These include a mutated epidermal growth factor receptor variant (EGFRvIII)17,18, mutant p5319, AIM-220, SART21, IL-13Ra222, survivin23,24, hTERT25,26 and melanoma antigen-encoding genes (MAGE-1, MAGE-3, gp100, TRP-1 and TRP2)27,28. Screening peptide and cDNA libraries for T cell reactivity can be conducive to identifying additional tumor-related antigens for further study. With the advent of microarray chip technology that can scan tumors in a high throughput manner for tumorspecific gene expression, more potential glioma-specific

antigens will probably be discovered and investigated in the near future29–34. Furthermore, recent advances in our understanding of tumor immunology can link the information generated from gene-expression profiling of CNS tumors to the identification of precise immunogenic epitopes that can serve as useful targets for specific immunotherapy. For instance, since T cells can only recognize antigens in the context of MHC, tumorassociated genes can be scrutinized for amino acid sequences that can interact with relevant MHC molecules. Candidate tumor-specific peptide epitopes for CTLs can now be identified using available computer algorithms that predict the MHC binding affinities of specific peptide and epitope sequences (http://bimas. dcrt.nih.gov/molbio/hla_bind/,http://www.uni-tuebingen. de/uni/kxi/). By testing the stability of MHC molecules with candidate tumor-specific peptides, an estimation of the potential immunogenicity of a peptide can be determined in vitro35. The immunogenicity of human cancer genes or their predicted CTL peptide epitopes can also be assessed in animal models using HLA transgenic mice36, providing even better estimates for immunotherapy targets in vivo. Recently, experiments involving the screening of human tumor samples for CTL reactivity have identified human glioma-associated antigen epitopes specifically recognized by the immune system22. Based on these recent advances, it is suggested that cancer genomics can be directly linked to tumor immunotherapy by ‘reverse immunology’, which paves the way for the design of new, more Table 2: Tumor-associated antigens in primary CNS malignancies

Category of antigen Tumor antigen

Documented CD8z T cell response to HLA restriction CNS tumor

Unique mutant

EGFRvIII p53 p97

HLA-A2 Multiple HLA Unknown

2 2 2

Cancer/testis antigens

MAGE GAGE IL-13a2 SSX

Multiple HLAs HLA-A29, Cw6 HLA-A2 HLA-A2

z 2 z 2

Differentiation antigens (tissue of origin-related Ag)

Tyrosinase

HLA-A1, A2, A24, B44 HLA-A31 HLA-A2, A31, A33, Cw8 HLA-A2, A3, A24, Cw8

2 2 z z

HLA-A2 HLA-A2 HLA-A2 HLA-A24 HLA-A2 HLA-A1

2 z 2 z 2 z

TRP-1 TRP-2 gp100

Normal proteins with selective or over-expression

Wild-type EGFR Her2/neu hTERT SART1, SART3 Survivin AIM-2

Neurological Research, 2005, Volume 27, October

697



specific targets for brain tumor immunotherapy. This concept of ‘tumor immunomics’37 (illustrated in Figure 5) has given new momentum to the daunting task of adding characterized TAA to the currently sparse list of effective antigenic targets for brain tumor immunotherapy.

overcome the identified problems will allow for the future design of more rational immunotherapeutics for CNS malignancies. With new genetic tools, such as DNA microarray analysis, significant progress in the treatment of neoplasms in the ‘immunologically privileged’ brain should be forthcoming42.

Microarrays and patient stratification for immunotherapy In addition to target antigen discovery, DNA microarray analysis of brain tumor specimens can also be used for molecular classification and patient stratification for immunotherapy. Although numerous immunotherapuetic approaches may hold great promise for the treatment of patients with brain cancer, a major challenge to the realization of clinical efficacy using these treatments is to determine which individual patients are most likely to benefit from the various forms of immunotherapy. DNA microarray-based technologies, which allow simultaneous analysis of the expression of thousands of genes, have already begun to uncover previously unrecognized patient subsets of brain cancer that differ in their survival31,32,38. The fact that DNA microarrays can be used to detect molecular subsets that strongly predict survival indicates that it will soon be possible to develop gene-based predictors of therapeutic response, including responses to specific immunotherapies39. One of the most promising applications of microarray technology is class distinction and patient selection for clinical trials (including immunotherapy protocols) through gene expression profiling as a diagnostic, therapeutic and prognostic tool. Recent studies have shown that gene-expression signatures are relatively robust across an individual tumor, regardless of the spatial location of the tumor sample. Thus, despite the spatial heterogeneity in the histological morphology of cancer cells within a tumor, the gene-expression profiles between spatially distinct biopsies of the same tumor are relatively consistent40. It has also been found that the variation in gene expression is much greater between samples from different patients, despite the same histological diagnosis, than from multiple samplings of a given tumor from an individual patient38,40. Therefore, the varied immunogenicity of shared tumor antigens among different patients owing to heterogeneous levels of antigen expression and differing levels of immunoreactivity of patients indicates that individualized immunotherapy should ideally be performed41. Systematic gene expression analysis by DNA chip/ microarray technology will lead to a better understanding of the nature of brain tumor antigens and changes in antigen gene expression before and after immunotherapy. This information will allow for the development and improvement of immunotherapeutic strategies by identifying patients who are most likely to benefit from immunotherapy and those who harbor immunotherapeutic escape mechanisms. Based on this information, more rational hypotheses about the various mechanisms of tumor immune escape and how to

METHODS OF TARGETED IMMUNOTHERAPY In addition to the identification of immunogenic tumor rejection antigens and the selection of appropriate patients, successful brain cancer immunotherapy also depends on the development of effective methods of delivering antitumor immunity to the CNS. Various types of immunotherapy strategies, both passive and active, have been used to target intracranial neoplasms.


Passive immunotherapy In passive (adoptive) immunotherapy, immunity is exogenously ‘transferred’ by injecting either antibodies or T cells specific for an antigen. Therefore, a host ‘passively’ acquires immunity, thereby not having to effectively generate immunity using endogenous immune cells. This mode of therapy can be advantageous when a patient is severely immunocompromised as a result of their disease. Serotherapy is a form of passive immunotherapy that involves the administration of antigen-specific antibodies. Tumor-specific antibodies used to target CNS glioma cells and deliver radiochemicals or immunotoxins to the tumor site have been extensively investigated43,44. Clinical experiences with radiolabeled anti-tenascin mAb 81C6 have resulted in statistically significant increases in survival and prolonged disease stabilization45. Clinical trials of immunotoxins for CNS neoplasia have been conducted using TFN-CRM107 (human diferric transferrin z diphtheria toxin mutant CRM107), and Pseudomonas aeruginosa exotoxin A mutants conjugated to interleukin-4 (IL-4-PE38KDEL)46, interleukin-13 (IL-13-PE38QQR)47–49, or transforming growth factor-alpha (TP-38)50. These initial trials of passive serotherapy have shown some significant antitumor effects, which have spurred further multicenter phase II/III trials. Adoptive transfer of lymphokine activated killer (LAK) cells, allogeneic fibroblasts and expanded systemic T cells have also been used to treat malignant glioma patients. Early studies focused on the adoptive transfer of non-antigen-specific LAK cells and rhIL-2 for CNS gliomas51. Even though sporadic cases of improved survival were reported, the non-specificity of LAK cells and the neurotoxicity associated with high-dose IL-2 hampered further studies. In pre-clinical animal models, prolonged survival was observed in mice treated with semi-allogeneic fibroblasts secreting IL-252. In recent years, greater tumor specificity was achieved with adoptive immunotherapy using tumor-sensitized or alloreactive CTLs, which have shown improved survival in pre-clinical models and in phase I clinical trials for CNS gliomas53,54. However, poor tumor-specific trafficking and rapid disappearance of transferred T cells



Figure 5: Diagram showing the necessary steps to link genomics with tumor immunology for the discovery of novel tumor antigens for CNS malignancies. Genes that are over-expressed in gliomas are identified initially by differential gene expression analysis (using DNA microarray, SAGE, etc.). These genes are analysed for their role in tumorigenesis and/or for their selective expression in brain tumors. Candidate antigens are tested further, using methods of epitope deduction. This includes peptide amino acid prediction, binding to HLA, and T cell repertoire analysis in animal models and human samples. (Figure adapted from ref. 35 by the kind permission of Elsevier Publishers)


699



still limited the usefulness of this strategy. To improve tumor-specific targeting, a recent study utilizing the adoptive transfer of interleukin-13-zetakine re-directed cytolytic T cells for tumor-specific recognition and killing of glioblastoma cells has shown notable success in xenograft models, and phase I clinical trials in recurrent glioblastoma patients have been initiated to test the feasibility of this treatment modality55. Active immunotherapy Active (vaccine) immunotherapy seeks to generate or augment endogenous host immunity by specifically boosting the host’s own immune response to antigen(s). Many studies of cancer vaccines have used cytokine gene therapy approaches to generate antitumor immunity. These pioneering studies were performed over a decade ago now and demonstrated that cytokinesecreting tumor cells could induce antitumor immunity56. With CNS tumor models, cytokine gene therapy studies have been able to demonstrate efficacious antitumor immunity to intracranial gliomas secreting IL-2, IL-4, IL-6, IL-7, IL-12, GM-CSF, mM-CSF and IFNa/b/c57–59. These studies reveal the complexity of the multiple mechanisms by which antitumor immunity can be generated. DC-based vaccine mechanisms are yet another form of active immunotherapy. As mentioned above, inflammatory signals that are elaborated by professional APC, such as DC, can subsequently prime potent tumorspecific T cell responses. Thus, investigators have sought to enhance tumor-specific antigen presentation to T cells using ex vivo generated DC. This therapy exploits the ability of DC to prime T cells to tumorderived antigens. DC are generated ex vivo and loaded with any of a wide array of tumor-derived antigens, and are then re-infused into the host to augment the immune response. Preclinical animal studies have demonstrated that DC pulsed with tumor peptides, RNA and/or tumor lysates can elicit antitumor immune response against CNS neoplasms in vivo60–63. To date, the safety and feasibility of DC therapy has been supported by recent data from several phase I clinical trials39,64–68, and a multicenter phase II clinical trial of DC vaccination for patients with newly diagnosed glioblastoma is underway. Although the clinical data to date offers too limited information to make any conclusions about efficacy, the potential advantages of DC-based immunotherapy, along with its safety and feasibility, make this a promising immunotherapeutic option. Despite such promise, however, there are still some practical and theoretical problems for the clinical development of DC-based vaccines, as several variables are at play within the context of therapy modulation. Examples include the source of the antigens used for priming (e.g. tumor lysates, specific peptide epitopes, or RNA), origin of the DC used (e.g. spleen or bone marrow-derived peripheral blood mononuclear cells), and DC maturation status (e.g. immature versus mature)69,70. In addition, dose, frequency and route of injection are also important considerations of the effectiveness of the 700 Neurological Research, 2005, Volume 27, October

vaccine and require careful manipulation. Furthermore, adjuvants to DC vaccines have also shown promising preliminary data. By exploiting the mechanisms of the danger hypothesis (described above), adjuvants such as LPS71, Poly IC:LC (Ampligen RTM)72, CpG73 and the toll-like receptor-7 (TLR-7) agonist imiquimod (AldaraTM)74 can potentiate DC activation, strengthen antigen presentation, and even enhance therapeutic efficacy. Another modality by which to deliver tumor antigens to the immune system involves the use of attenuated, recombinant viruses and/or bacteria. Viral/bacterial infections and the resulting tissue damage can also provide the appropriate ‘danger signals’ to attract professional APC necessary for adequate antigen presentation. Using recombinant techniques, tumor antigens have been introduced into viral and bacterial vectors (e.g. Vaccinia, Listeria) to improve APC recognition of the tumor epitopes and subsequent APC antigen presentation to CTL75–78. Direct infection of APC by viruses/bacteria expressing tumor antigens could result in endogenous processing of tumor epitopes in the context of MHC class I, as well as antigen presentation via MHC class II pathways. This improved antigen processing/presentation, along with the danger signals associated with infection, may break tumor tolerance and enhance specific antitumor immunity within the CNS. Furthermore, vaccine strategies that stimulate the presentation of tumor-specific antigens to T lymphocytes may be particularly beneficial for CNS tumors because they have been found to induce determinant spreading, which is a phenomenon whereby T cells can recognize shared endogenous glioma epitopes along with the targeted antigen(s)76. Given the heterogeneity of human brain tumors and the possibility of immune escape of tumor cells, exploiting the process of epitope spreading will obviously be valuable in the clinical context of designing tumor vaccines in the future. CONCLUSION AND FUTURE DIRECTIONS The virtual absence of significantly improved survival for patients with CNS tumors in the last decades indicates that novel strategies for treating these neoplasms must be explored intensively. Experimental immunotherapeutic strategies represent attractive complements to treatments based on surgery, radiation and traditional chemotherapy, all of which have proved rather disappointing. The impressive advances that have been achieved in the understanding of tumor immunology and tumor genomics/immunomics provide novel ways to attack brain cancer. Because of encouraging results in many phase I studies, phase II clinical trials of immunotherapy should be pursued. However, it should be noted that, as with any other targeted treatment modality for brain tumors, immunotherapy trials might only realize significant potential clinical efficacy if given to the appropriate subgroup of patients and/or if administered in combination with other therapies. With the current experience in cancer treatments, it appears that simultaneously


targeting several components essential to the neoplastic process should provide maximal chances of tumor control. Therefore, therapies based on immunoenhancement and cancer vaccines could be combined with the traditional surgery, radiation and chemotherapies. Such integrated treatment strategies may prove to be of low toxicity and should be synergistic. In addition, the combined use of conventional treatments within the context of clinical trials of immunotherapy will allow evaluation of efficacy, yet retain the ethical requirements for human investigation. ACKNOWLEDGEMENTS


We are grateful to the generous support of the Kenneth A. Jonsson and Phillip R. Jonsson Foundations, the Neidorf Family Foundation and the Musella Foundation for Brain Tumor Research.

REFERENCES 1 Medzhitov R, Janeway CA. Decoding the patterns of self and nonself by the innate immune system. Science 2002; 296: 298– 300 2 Diefenbach A, Jamieson AM, Liu SD, et al. Ligands for the murine NKG2D receptor: Expression by tumor cells and activation of NK cells and macrophages. Nat Immunol 2000; 1: 119–126 3 Prins RM, Liau LM. Immunology and immunotherapy in neurosurgical disease. Neurosurgery 2003; 53: 144–153 4 Janeway C, Travers P, Walport M, et al. Immunobiology: The immune system in health and disease, New York: Current Biology Publications, 2001 5 Matzinger P. The danger model: A renewed sense of self. Science 2002; 296: 301–305 6 Skoberne M, Beignon AS, Bhardwaj N. Danger signals: A time and space continuum. Trends Mol Med 2004; 10: 251–257 7 Liu L, Zhou X, Liu H, et al. CpG motif acts as a "danger signal" and provides a T helper type-1-biased microenvironment for DNA vaccination. Immunology 2005; 115: 223–230 8 Fontenot JD, Rudensky AY. A well-adapted regulatory contrivance: Regulatory T cell development and the forkhead family transcription factor Foxp3. Nat Immunol 2005; 6: 331–337 9 Sakaguchi S. Naturally arising Foxp3-expressing CD25(z)CD4(z) regulatory T cells in immunological tolerance to self and non-self. Nat Immunol 2005; 6: 345–352 10 Lampson LA. Brain tumor immunotherapy: An immunologist’s perspective. J Neurooncol 2003; 64: 3–11 11 Nguyen MD, Julien JP, Rivest S. Innate immunity: the missing link in neuroprotection and neurodegeneration? Nat Rev Neurosci 2002; 3: 216–227 12 Liau LM, Fakhrai H, Black KL. Prolonged survival of rats with intracranial C6 gliomas by treatment with TGF-b2 antisense gene. Neurol Res 1998; 20: 742–747 13 Yang L, Ng KY, Lillehei KO. Cell-mediated immunotherapy: a new approach to the treatment of malignant glioma. Cancer Control 2003; 10: 138–147 14 Ichinose M, Masuoka J, Shiraishi T, et al. Fas ligand expression and depletion of T-cell infiltration in astrocytic tumors. Brain Tumor Pathol 2001; 18: 37–42 15 Hahn HP, Pang M, He J, et al. Galectin-1 induces nuclear translocation of endonuclease G in caspase- and cytochrome cindepdendent T-cell death. Cell Death Differ 2004; 11: 1277– 1286 16 Tambur A. Transplantation immunology and the central nervous system. Neurol Res 2004; 26: 243–255 17 Heimberger AB, Crotty LE, Archer GE, et al. Epidermal growth factor receptor VIII peptide vaccination is efficacious against established intracerebral tumors. Clin Cancer Res 2003; 9: 4247– 4254 18 Wikstrand CJ, Cole VR, Crotty LE, et al. Generation of antiidiotypic reagents in the EGFRvIII tumor-associated antigen system. Cancer Immunol Immunother 2002; 50: 639–652

19 Lutzker SG, Lattime EC. Use of dendritic cells to immunize against cancers overexpressing p53. Clin Cancer Res 2001; 7: 2–4 20 Liu G, Yu JS, Zeng G, et al. AIM-2: A novel tumor antigen is expressed and presented by human glioma cells. J Immunother 2004; 27: 220–226 21 Murayama K, Kobayashi T, Imaizumi T, et al. Expression of the SART3 tumor-rejection antigen in brain tumors and induction of cytotoxic T lymphocytes by its peptides. J Immunother 2000; 23: 511–518 22 Okano F, Storkus WJ, Chambers WH, et al. Identification of a novel HLA-A*0201-restricted, cytotoxic T lymphocyte epitope in a human glioma-associated antigen, interleukin 13 receptor alpha2 chain. Clin Cancer Res 2002; 8: 2851–2855 23 Kajiwara Y, Yamasaki F, Hama S, et al. Expression of survivin in astrocytic tumors: correlation with malignant grade and prognosis. Cancer 2003; 97: 1077–1083 24 Katoh M, Wilmotte R, Belkouch MC, et al. Survivin in brain tumors: An attractive target for immunotherapy. J Neurooncol 2003; 64: 71–76 25 Vonderheide RH, Domchek SM, Schultze JL, et al. Vaccination of cancer patients against telomerase induces functional antitumor CD8z T lymphocytes. Clin Cancer Res 2004; 10: 828–839 26 Maes L, Lippens E, Kalala JP, et al. The hTERT-protein and Ki-67 labelling index in recurrent and non-recurrent meningiomas. Cell Prolif 2005; 38: 3–12 27 Liu G, Ying H, Zeng G, et al. HER-2, gp100, and MAGE-1 are expressed in human glioblastoma and recognized by cytotoxic T cells. Cancer Res 2004; 64: 4980–4986 28 Prins RM, Odesa SK, Liau LM. Immunotherapeutic targeting of shared melanoma-associated antigens in a murine glioma model. Cancer Res 2003; 63: 8487–8491 29 Watson MA, Perry A, Budhjara V, et al. Gene expression profiling with oligonucleotide microarrays distinguishes World Health Organization grade of oligodendrogliomas. Cancer Res 2001; 61: 1825–1829 30 Liau LM, Lallone RL, Seitz RS, et al. Identification of a human glioma-associated growth factor gene, granulin, using differential immuno-absorption. Cancer Res2000; 60: 1353–1360 31 Shai R, Shi T, Kremen TJ, et al. Gene expression profiling identifies molecular subtypes of gliomas. Oncogene 2003; 22: 4918–4923 32 Mischel PS, Shai R, Shi T, et al. Identification of molecular subtypes of glioblastoma by gene expression profiling. Oncogene 2003; 22: 2361–2373 33 Mischel PS, Nelson SF, Cloughesy TF. Molecular analysis of glioblastoma: Pathway profiling and its implications for patient therapy. Cancer Biol Ther 2003; 2: 242–247 34 Liau LM, Yang I. Microarrays and the genetic analysis of brain tumors. Current Genomics 2003; 3: 33–41 35 Schultze JL, Vonderheide RH. From cancer genomics to cancer immunotherapy: Toward second-generation tumor antigens. Trends Immunol 2001; 22: 516–523 36 Butterfield LH, Meng WS, Koh A, et al. T-cell responses to HLAA*0201–restricted peptides derived from human alpha fetoprotein. J Immunol 2001; 166: 5300–5308 37 Maecker B, von B, Anderson KS, et al. Linking genomics to immunotherapy by reverse immunology—’immunomics’ in the new millennium. Curr Mol Med 2001; 1: 609–619 38 Freije WA, Castro-Vargas FE, Fang Z, et al. Gene expression profiling of gliomas strongly predicts survival. Cancer Res 2004; 64: 6503–6510 39 Liau LM, Prins RM, Kiertscher SM, et al. Dendritic cell vaccination in glioblastoma patients induces systemic and intracranial T-cell responses modulated by the local CNS tumor microenvironment. Clin Cancer Res 2005; 11: 5515–5525 40 Mehrian-Shai R, Reichardt JK, Ya-Hsuan H, et al. Robustness of gene expression profiling in glioma specimen samplings and derived cell lines. Brain Res Mol Brain Res 2005; 136: 99–103 41 Kawakami K, Fujita T, Matsuzaki Y, et al. Identification of human tumor antigens and its implications for diagnosis and treatment of cancer. Cancer Sci 2004; 95: 784–791 42 Virasch N, Kruse CA. Strategies using the immune system for therapy of brain tumors. Hematol Oncol Clin North Am 2001; 15: 1053–1071


701


Central nervous system malignancies: Haumith R. Khan-Farooqi et al. 43 Hall WA. Targeted toxin therapy for malignant astrocytoma. Neurosurgery 2000; 46: 544–551 44 Wikstrand CJ, Zalutsky MR, Bigner DD. Radiolabeled antibodies for therapy of brain tumors. In: Liau LM, Becker DP, Cloughesy TF, Bigner DD, eds. Brain Tumor Immunotherapy, Totowa: Humana Press, 2001: pp. 205–229 45 Reardon DA, Akabani G, Coleman RE, et al. Phase II trial of murine (131)I-labeled antitenascin monoclonal antibody 81C6 administered into surgically created resection cavities of patients with newly diagnosed malignant gliomas. J Clin Oncol 2002; 20: 1389–1397 46 Rand RW, Kreitman RJ, Patronas N, et al. Intratumoral administration of recombinant circularly permuted interleukin-4Pseudomonas exotoxin in patients with high-grade glioma. Clin Cancer Res 2000; 6: 2157–2165 47 Kunwar S. Convection-enhanced delivery of IL13-PE38QQR for treatment of recurrent malignant glioma: Presentation of interim findings from ongoing Phase I studies. Acta Neurochir Suppl 2003; 88: 105–111 48 Parney IF, Kunwar S, McDermott M, et al. Neuroradiographic changes following convection-enhanced delivery of the recombinant cytotoxin interleukin 13–PE38QQR for recurrent malignant glioma. J Neurosurg 2005; 102: 267–275 49 Kioi M, Kawakami K, Puri RK. Analysis of antitumor activity of an interleukin-13 (IL-13) receptor-targeted cytotoxin composed of IL13 antagonist and Pseudomonas exotoxin. Clin Cancer Res 2004; 10: 6231–6238 50 Sampson JH, Akabani G, Archer GE, et al. Progress report of a Phase I study of the intracerebral microinfusion of a recombinant chimeric protein composed of transforming growth factor (TGF)-a and a mutated form of the Pseudomonas exotoxin termed PE-38 (TP-38) for the treatment of malignant brain tumors. J Neurooncol 2003; 65: 27–35 51 Merchant RE, Grant AJ, Merchant LH, et al. Adoptive immunotherapy for recurrent glioblastoma multiforme using lymphokine activated killer cells and recombinant interleukin-2. Cancer 1988; 62: 665–671 52 Glick RP, Lichtor T, de Zoeten E, et al. Prolongation of survival of mice with glioma treated with semi-allogeneic fibroblasts secreting interleukin-2. Neurosurgery 1999; 45: 867–874 53 Kruse CA, Cepeda L, Owens B, et al. Treatment of recurrent glioma with intracavitary alloreactive cytotoxic T lymphocytes and interleukin-2. Cancer Immunol Immunother 1997; 45: 77–87 54 Merchant RE, Baldwin NG, Rice CD, et al. Adoptive immunotherapy of malignant glioma using tumor-sensitized T lymphocytes. Neurol Res 1997; 19: 145–152 55 Kahlon KS, Brown C, Cooper LJ, et al. Specific recognition and killing of glioblastoma multiforme by interleukin 13-zetakine redirected cytolytic T cells. Cancer Res 2004; 64: 9160–9166 56 Dranoff G, Jaffee E, Lazenby A, et al. Vaccination with irradiated tumor cells engineered to secrete murine granulocyte-macrophage colony-stimulating factor stimulates potent, specific, and longlasting anti-tumor immunity. Proc Natl Acad Sci U S A 1993; 90: 3539–3543 57 Giezeman-Smits KM, Okada H, Brissette-Storkus CS, et al. Cytokine gene therapy of gliomas: Induction of reactive CD4z T cells by interleukin-4-transfected 9L gliosarcoma is essential for protective immunity. Cancer Res 2000; 60: 2449–2457 58 Okada H, Villa L, Attanucci J, et al. Cytokine gene therapy of gliomas: Effective induction of therapeutic immunity to intracranial tumors by peripheral immunization with interleukin-4 transduced glioma cells. Gene Ther 2001; 8: 1157–1166 59 Okada H, Pollack IF. Cytokine gene therapy for malignant glioma. Expert Opin Biol Ther 2004; 4: 1609–1620 60 Heimberger AB, Crotty LE, Archer GE, et al. Bone marrow-derived dendritic cells pulsed with tumor homogenate induce immunity


61

62 63

64

65

66

67

68 69 70 71 72

73 74

75

76 77 78

against syngeneic intracerebral glioma. J Neuroimmunol 2000; 103: 16–25 Heimberger AB, Archer GE, Crotty LE, et al. Dendritic cells pulsed with a tumor-specific peptide induce long-lasting immunity and are effective against murine intracerebral melanoma. Neurosurgery 2002; 50: 158–164 Liau LM, Black KL, Prins RM, et al. Treatment of intracranial gliomas with bone marrow-derived dendritic cells pulsed with tumor antigens. J Neurosurg 1999; 90: 1115–1124 Ashley DM, Faiola B, Nair S, et al. Bone marrow-generated dendritic cells pulsed with tumor extracts or tumor RNA induce antitumor immunity against central nervous system tumors. J Exp Med 1997; 186: 1177–1182 Caruso DA, Orme LM, Neale AM, et al. Results of a phase 1 study utilizing monocyte-derived dendritic cells pulsed with tumor RNA in children and young adults with brain cancer. Neuro-Oncology 2004; 6: 236–246 Kobayashi T, Yamanaka R, Homma J, et al. Tumor mRNA-loaded dendritic cells elicit tumor-specific CD8(z) cytotoxic T cells in patients with malignant glioma. Cancer Immunol Immunother 2003; 52: 632–637 Yamanaka R, Abe T, Yajima N, et al. Vaccination of recurrent glioma patients with tumour lysate-pulsed dendritic cells elicits immune responses: results of a clinical phase I/II trial. Br J Cancer 2003; 89: 1172–1179 Yu JS, Wheeler CJ, Zeltzer PM, et al. Vaccination of malignant glioma patients with peptide-pulsed dendritic cells elicits systemic cytotoxicity and intracranial T-cell infiltration. Cancer Res 2001; 61: 842–847 Yu JS, Liu G, Ying H, et al. Vaccination with tumor lysate-pulsed dendritic cells elicits antigen-specific, cytotoxic T-cells in patients with malignant glioma. Cancer Res 2004; 64: 4973–4979 Walker PR, Calzascia T, de Tribolet N, et al. T-cell immune responses in the brain and their relevance for cerebral malignancies. Brain Res Rev 2003; 42: 97–122 Fecci PE, Mitchell DA, Archer GE, et al. The history, evolution, and clinical use of dendritic cell-based immunization strategies in the therapy of brain tumors. J Neurooncol 2003; 64: 161–176 Yang Y, Huang CT, Huang X, et al. Persistent Toll-like receptor signals are required for reversal of regulatory T cell-mediated CD8 tolerance. Nat Immunol 2004; 5: 508–515 Adams M, Navabi H, Jasani B, et al. Dendritic cell (DC)-based therapy for cervical cancer: Use of DC pulsed with tumour lysate and matured with a novel synthetic clinically non-toxic double stranded RNA analogue poly [I]:poly [C(12)U] (Ampligen R). Vaccine 2003; 21: 787–790 Schneeberger A, Wagner C, Zemann A, et al. CpG motifs are efficient adjuvants for DNA cancer vaccines. J Invest Dermatol 2004; 123: 371–379 Shackleton M, Davis ID, Hopkins W, et al. The impact of imiquimod, a Toll-like receptor-7 ligand (TLR7L), on the immunogenicity of melanoma peptide vaccination with adjuvant Flt3 ligand. Cancer Immun 2004; 4: 9–11 Sewell DA, Shahabi V, Gunn GR, et al. Recombinant Listeria vaccines containing PEST sequences are potent immune adjuvants for the tumor-associated antigen human papillomavirus-16 E7. Cancer Res 2004; 64: 8821–8825 Liau LM, Jensen ER, Kremen TJ, et al. Tumor immunity within the central nervous system stimulated by recombinant Listeria monocytogenes vaccination. Cancer Res 2002; 62: 2287–2293 Brockstedt DG, Giedlin MA, Leong ML, et al. Listeria-based cancer vaccines that segregate immunogenicity from toxicity. Proc Natl Acad Sci U S A 2004; 101: 13832–13837 Meyer RG, Britten CM, Siepmann U, et al. A phase I vaccination study with tyrosinase in patients with stage II melanoma using recombinant modified vaccinia virus Andara (MVA-hTyr). Cancer Immunol Immunother 2004; 54: 453–467