Jonathan S. Berek
Cancer is caused by a series of events that include the accumulation of successive molecular lesions and alterations in the tumor microenvironment (1). Molecular lesions include overexpression, amplification, or mutations of oncogenes; deletion of tumor suppressor genes; and the inappropriate expression of growth factors and their cellular receptors. In addition to these molecular changes, the formation of new blood vessels (angiogenesis) and the lack of effective host antitumor immune responses create a microenvironment that supports the growth of cancer (2). Our improved understanding of these mechanisms presents an opportunity for the development of novel therapeutic approaches (3). This chapter provides an overview of biologic, targeted, and immunotherapeutic strategies for gynecologic cancers.
Biologic and Targeted Therapies
The growth of cancer cells is crucially dependent on oncogenic signal transduction pathways. Extracellular signals are transmitted to the cancer cell via transmembrane receptors. Activation of the epidermal growth factor receptors (EGFR, HER2, HER3, and HER4), for example, stimulates a cascade of intracellular proteins that ultimately lead to changes in gene expression. Novel therapeutics are targeted to modulate these signal transduction pathways by blocking the extracellular transmembrane receptors or interfering with intracellular proteins such as tyrosine kinases further downstream. This novel therapeutic approach is also termed molecular targeting (4). It is accomplished by either monoclonal antibodies that bind to transmembrane receptors and serum proteins such as vascular endothelial growth factor (VEGF) or chemical, small-molecule inhibitors that prevent activation of signal transduction proteins. Targeting the signaling cascade inhibits the proliferation of cancer cells, induces apoptosis, and blocks metastasis. The specificity of these molecules is based on the assumption that cancer cells are overexpressing various proteins in the signal transduction pathways, therefore presenting a preferred target compared to normal cells. Conceptually, this should result in more cancer cell-specific therapy and less clinical side effects because of sparing of normal tissue (5). At this time, a large variety of molecular-targeting strategies are being tested for efficacy in clinical trials (Table 2.1).
Table 2.1 Targeted Therapies in Cancer Disease
The formation of new blood vessels (neoangiogenesis) is a normal process during embryonic development, tissue remodeling, and wound healing (6). Malignant tumors are able to induce angiogenesis by secreting paracrine factors that promote the formation of new blood vessels. Angiogenesis is a complex process that is influenced by various pro-and antiangiogenic factors, including VEGF, interleukin 8, platelet-derived endothelial cell growth factor, and angiopoietins. Overexpression of these angiogenic factors leads to neovascularization and increased supply of nutrients and oxygen to the tumor.
Three main therapeutic strategies that target angiogenesis are currently being explored for the treatment of cancer patients (7). One group of agents targets VEGF (e.g.,bevacizumab, VEGF-Trap), the second group prevents VEGF from binding to its receptor (pertuzumab), and a third group of agents inhibits tyrosine kinase activation and downstream signaling in the angiogenesis signaling cascade (valatanib, sunitenib) (8).
Vascular Endothelial Growth Factor
VEGF is overexpressed in gynecologic malignancies, therefore presenting an excellent target for therapy (9). Inhibition of VEGF-induced angiogenic signaling decreases tumor microvascular density and causes death of solid tumors in various preclinical models. Several agents are now available for clinical use; all target the VEGF signaling pathway. The most widely used agent at this time is bevacizumab, a humanized, recombinant monoclonal antibody that binds to all isoforms of VEGF-A. Bevacizumab has been approved for the treatment of colorectal carcinoma based on improved overall survival in combination with chemotherapy in patients with metastatic colorectal cancer in (10,11). In addition,bevacizumab has shown promising effects in metastatic non-small cell lung cancer and recurrent or metastatic breast carcinoma in (12).
In ovarian carcinoma, various clinical trials have demonstrated the efficacy of bevacizumab treatment. In a study by the Gynecologic Oncology Group, 62 patients received single agent bevacizumab 15 mg/kg intravenously every 21 days (13). Thirteen patients (21%) showed clinical responses with two complete and 11 partial responses. The median response duration was 10 months, and 25 patients (41.3%) survived progression free for at least 6 months. In a second trial, bevacizumab treatment of 44 patients with recurrent, platinum-resistant ovarian carcinoma resulted in partial responses in seven patients (15.9%) and stable disease in 27 (61.4%) (14). Median progression-free survival was 4.4 months with a median survival of 10.7 months.
Bevacizumab has also been used in combination with other agents. In a phase II study of 13 patients with recurrent ovarian or primary peritoneal carcinoma, combination treatment with bevacizumab (15 mg/kg i.v. every 21 days) and erlotinib (150 mg/day orally) resulted in one complete response and one partial response for a total response rate of 15% (12). Seven patients had stable disease. Another trial investigated the combination of bevacizumab (10 mg/kg every 14 days) and oral cyclophosphamide (50 mg/day orally) in 70 patients with recurrent ovarian cancer (15). Partial responses were observed in 17 patients (24%), with a median time to progression and survival of 7.2 and 16.9 months, respectively. The probability of being alive and progression free at 6 months was 56%. Serum levels of VEGF and thrombospondin-1 were determined in four patients and showed a decrease over time but did not correlate with clinical outcome. The combination of bevacizumab and weekly taxane therapy was evaluated in a small series of patients (16). All nine evaluable patients had a decrease of serum CA125 and a significant improvement of their disease-related symptoms.
Bevacizumab-related side effects include venous and arterial thrombosis, hemorrhage, nephrotic syndrome with proteinuria, hypertension, rare leukoencephalopathy, and bowel perforation. The risk factors in particular for bowel perforation under bevacizumab treatment are not well identified but include tumor involving the intestines and bowel resection. The incidence of bowel perforation in ovarian cancer patient during bevacizumab treatment is reported in several clinical trials. In one trial, five of 44 patients (11.4%) experienced treatment related gastrointestinal perforation (14), while others have not reported any such complication (13). Recent reviews have suggested that the incidence of bowel perforation for ovarian cancer patients receiving bevacizumab is approximately 5% to 7% (17,18).
The Gynecologic Oncology Group has initiated a clinical trial that will evaluate the addition of bevacizumab to first-line chemotherapy after primary tumor debulking (19). Patients receive bevacizumab in combination with carboplatin and paclitaxel followed by bevacizumab alone for consolidation. A similar trial by the Gynecologic Cancer InterGroup is designed to evaluate the safety and efficacy of adding bevacizumab to standard chemotherapy (carboplatin and paclitaxel) in patients with advanced epithelial ovarian or primary peritoneal cancer (ICON7) (19).
Various other targeting strategies of angiogenic pathways are currently being investigated (7). VEGF-Trap (AVE 0005) is a recombinant fusion protein that consists of the extracellular domain of VEGF receptors VEGFR1 and VEGFR2 fused to the FC portion of immunoglobulin G1 (20). VEGF is inactivated by binding to the ligand-binding domain of this fusion protein followed by destruction of this complex via immune system-mediated mechanisms. Smallmolecule tyrosine kinase inhibitors that target the VEGF pathway include pazopanib,vatalanib, and sunitinib. Pazopanib is a multitarget receptor tyrosine kinase inhibitor against VEGFR1, VEGFR2, VEGFR3, platelet-derived growth factor receptor (PDGFR), and c-kit(3). Similarly, vatalanib (PTK787) targets multiple VEGF-receptor tyrosine kinases. Sunitinib (SU11248) inhibits PDGR, VEGFR, c-kit, and SLT3 and has shown promising results in renal cell cancer and gastrointestinal stromal tumors. Trials in gynecologic malignancies are ongoing to test the efficacy of these agents in patients.
Epidermal Growth Factor Receptor Inhibitors
The epidermal growth factor receptor pathway plays an important role in regulation of growth and differentiation of epithelial cells through regulation of cell division, migration, adhesion, differentiation, and apoptosis (21). The epidermal growth factor receptor family consists of four members including EGFR (HER1), HER2, HER3, and HER4 (22). EGFR overexpression has been reported in 35% to 70% of patients with epithelial ovarian cancer (23,24). In endometrial cancer, EGFR is overexpressed in 43% to 67% of tumors and is associated with shortened disease-free and overall survival (25,26,27). In addition, amplification of the HER2 gene is commonly found in endometrial carcinoma. Overexpression of the HER2 receptor is more prevalent in nonendometrial cancer and is associated with an aggressive form of the disease. In uterine serous papillary carcinoma, HER2 gene amplification can be demonstrated in as many as 42% of cases (28).
Various agents directed against epidermal growth factor receptors are available (29). Trastuzumab is a humanized monoclonal antibody that binds to the extracellular domain of HER2 (30). Blockade of HER2 affects various molecules that ultimately decreases cell proliferation. Pertuzumab is another humanized monoclonal antibody that binds to a different epitope of HER2 compared to trastuzumab. Binding to HER2 prevents dimerization of the receptor, which is required for its function (31). Cetuximab is a chimeric monoclonal antibody that binds to EGFR, thereby preventing dimerization and activation (32). Gefitinib is a small-molecule tyrosine kinase inhibitor of EGFR that prevents phosphorylation of the receptor by binding to the intracellular ATP-binding domain of the receptor (33). Erlotinib is a small-molecule tyrosine kinase inhibitor of EGRF that prevents phosphorylation of the intracellular domain of the EGFR receptor. Lapatinib (GW572016) inhibits both EGFR and HER2 (Fig. 2.1).
Figure 2.1 Inhibition of epidermal growth factor receptor signaling.
Epidermal Growth Factor Receptor
Inhibition of EGFR signaling is accomplished by using either monoclonal antibodies against the extracellular receptor or small-molecule inhibitors against the intracellular kinase domain. Both strategies results in inhibition of phosphorylation or receptor activation.
Erlotinib is a potent reversible inhibitor of EGFR tyrosine kinase that blocks receptor autophosphorylation and has been used for the treatment of ovarian carcinoma. In one study, 34 patients were treated with single-agent erlotinib (150 mg/day orally) for as long as 48 weeks (34). Two patients showed a partial response lasting 8 and 17 weeks. Fifteen patients (44%) had stable disease, and 17 patients (50%) progressed under treatment. The side effects of erlotinib were mainly confined to tissues with strong expression of EGFR: Skin rashes and diarrhea were observed in 68% and 38% of patients, respectively.
Erlotinib has been used in combination with docetaxel and carboplatin as first-line treatment after surgical cytoreduction in patients with ovarian, fallopian tube, and primary peritoneal cancers (35). In this study, 23 evaluable patients showed five complete and seven partial responses. The treatment was well tolerated; main side effects were neutropenia and skin rashes. The study demonstrated the feasibility and tolerability of erlotinib in conjunction with chemotherapy.
Cetuximab (C225, Erbitux) is a chimerized monoclonal antibody against EGFR. Treatment of patients with primary ovarian or peritoneal cancer using cetuximab has shown only modest activity in screened patients with EGFR-positive tumors. Cetuximab in combination with carboplatin resulted in three complete (10.7%) and six partial (21.4%) responses in 28 patients with recurrent ovarian cancer (36). Twenty-six of these 28 patients (92.8%) had EGFR-positive tumors. The combination of paclitaxel, carboplatin, and cetuximab for first-line chemotherapy of stage III ovarian cancer patients resulted in progression-free survival of 14.4 months and was therefore not significantly prolonged compared to historical data (37).
Gefitinib (ZD1839 Iressa) is a low molecular weight quinazoline derivative that inhibits the activation of EGFR tyrosine kinase via competitive binding of the ATP-binding domain of the receptor. In a clinical trial by the Gynecologic Oncology Group, 27 patients with recurrent or persistent epithelial ovarian cancer were treated with 500 mg gefitinib daily (38). Four patients (14.8%) survived progression free for more than 6 months with one objective response (3.6%). Commonly observed toxicity included skin rash and diarrhea. Interestingly, EGFR expression was associated with longer progression-free survival and possibly longer survival. The patient with the only objective antitumor response had a tumor with a mutation in the catalytic domain of the tumor's EGFR (2235dEL15). This patient received 29 cycles of gefitinib and had a progression-free survival of approximately 27 months.
In a separate trial, 24 patients with recurrent epithelial ovarian cancer were treated with gefitinib 500 mg daily (39). All tumor samples had detectable levels of EGFR and PGFR. Of 16 patients who completed more than two cycles of therapy, no complete or partial responses were observed. However, analysis of clinical samples showed that gefitinib inhibited phosphorylation of EGFR, thereby providing a conceptual proof of targeted therapy.
Treatment of patients with recurrent ovarian cancer using the combination of gefitinib, carboplatin, and paxitaxel resulted in a high overall response rate of 63% (40). Interestingly, antitumor responses were observed in 35% of patients with platinum-resistant disease compared to a 73% response rate in patients with platinum-sensitive disease. Based on the preliminary data, none of the 18 patients treated showed EGFR receptor mutations.
Gefitinib has also been used in combination with tamoxifen. In 56 patients with primary ovarian or fallopian tube cancer, treatment with tamoxifen (40 mg/day) and gefitinib (500 mg/day) did not result in objective antitumor responses, but 16 patients had stable disease (41). In squamous and adenocarcinoma of the cervix, gefitinib (500 mg/day) treatment resulted in disease stabilization in six of 28 patients (20%) but no clinical responses (42). The median duration of stable disease was 111.5 days with a median overall survival of 107 days.
Lapatinib is a small-molecule inhibitor of both the HER2 and EGFR tyrosine kinase receptor.
The rationale for using lapatinib in endometrial carcinoma is supported mainly by studies in human cancer cell lines. Its efficacy in endometrial cancer is being investigated currently in clinical trials (28).
The HER-2/neu receptor is activated by homo- or heterodimerization, resulting in tyrosine phosphorylation and subsequent activation of various downstream signals that among other functions control cellular proliferation, migration, and invasion. Trastuzumab is a recombinant, humanized IgG1 monoclonal antibody that is specific for the extracellular domain of HER-2/neu. Binding of the antibody to HER-2/neu prevents activation of the receptor with a subsequent increase of apoptosis in vitro and in vivo, impaired DNA damage repair, and inhibition of tumor neovascularization (30). Preclinical models have suggested that the therapeutic activity may also depend on innate immune effector cells that mediate antibodydependent cellular cytotoxicity. In addition, trastuzumab influences the adaptive immune response and augments antigen processing and presentation (43). In breast cancer, the addition of trastuzumab to adjuvant chemotherapy for patients with HER-2/neu-positive tumors significantly decreases the hazard ratio for recurrence and subsequently improves survival (44).
The HER-2/neu oncogene is overexpressed in several gynecologic malignancies, including 20% to 30% of ovarian cancers (45). The largest clinical trial evaluating HER2/neu as a target in ovarian or primary peritoneal carcinoma was conducted by the Gynecologic Oncology Group (46). Of 837 tumor samples screened for HER2/neu expression, 95 patients (11.4%) were found to have tumors with HER/neu overexpression. Forty-one patients with HER2/neu-positive tumors received trastuzumab weekly. Single-agent treatment resulted in one complete (2.4%) and two partial (4.9%) responses, with a median duration of response of 8 weeks (range 2 to 104 weeks). The authors concluded that single-agenttrastuzumab in recurrent ovarian cancer was of limited value because of the low frequency of HER-2/neu overexpression and the low rate of clinical antitumor response.
HER2/neu overexpression is infrequent in cervical cancer. In one study, only one of 35 (2.9%) cervical carcinomas showed strong expression of HER2/neu (47).
In uterine papillary serous carcinoma, 12 of 68 (18%) tumors showed HER2/neu overexpression; this was associated with a worse overall prognosis (48). In a separate study, five of 19 specimens (26%) stained strongly for HER2/neu protein receptor (49).
Mitogen-Activated Protein Kinase Pathways
The mitogen-activated protein (MAP) kinase cascades are activated by various cofactors, inflammatory cytokines, and stress (50). The signaling cascades include various molecules, including RAS, MEK1/2, ERK1/2, and p38 MAPK. Various molecules have been developed that target this pathway but are mostly still under investigation. Sorafenib is among the first of the agents with clinically proven efficacy. Sorafenib is a competitive inhibitor of raf that has been approved for treatment of renal cell carcinoma and hepatocellular carcinoma (51). Besides targeting raf, sorafenib also inhibits VEGFR2 and VEGFR3, FT3, c-kit, and PDGFR-β.
The PI3-kinase/Akt/mTOR Pathway
The phosphoinositide3-kinase (PI3-kinase)/Akt/mTOR pathway is a major oncogenic signaling pathway in various cancers (52). Activation of this pathway can be demonstrated in more than 80% of endometrial cancers, 50% to 70% of epithelial ovarian cancers, and approximately 50% of cervical cancers (53,54,55). Activation of PI3-kinase by various growth factors such as platelet-derived growth factor (PDGF) or insulin growth factor results in phosphorylation and therefore activation of the central oncogenic protein Akt. Activated Akt is released from the membrane and elicits downstream effects mainly by phosphorylating signal transduction proteins such as BAD, FKHR, Caspase 9, and mammalian target of rapamycin (mTOR). Activation of these downstream signals leads to an increase in cellular proliferation, invasiveness, drug resistance, and neoangiogenesis. The PTEN (phosphatase and tensin homologue deleted on chromosome 10) gene is a tumor suppressor gene that is located on chromosome 10q23 and encodes a dual-specificity phosphatase for both lipid and protein substrates (56). PTEN decreases the activation of Akt in the PTEN/PI3-kinase/Akt pathway.
Several inhibitors of PI3-kinase/Akt/mTOR signaling are currently in clinical trials (57,58). Rapamycin or rapamycin analogues, for example, block the activity of mTOR, a protein complex responsible for increasing protein synthesis and cellular proliferation (59). Several mTOR inhibitors, including RAD001 and CCI779, and specific PI3-kinase inhibitors are currently under development in preclinical models and clinical trials. PI3-kinase/Akt/mTOR inhibitors have been used in endometrial cancer with limited benefit (60). However, the results from clinical trials using mTOR inhibitors in renal cell carcinomas and glioblastomas are very encouraging (61).
Failure of functional immunity contributes to the genesis of virus-associated cancers, such as those caused by human papilloma virus (HPV) or Epstein-Barr virus. The greatest success story involving the enhancement of immunity to combat gynecologic cancer is the development of vaccines against HPV, which are highly effective for the prevention of cervical dysplasia and cancer (62). Although many effective induced antitumor immune responses have been described, the relative role of natural antitumor immune responses in the detection and destruction of cancer cells, at least as was envisioned originally when the concept of immune surveillance was first defined (63), is still unclear. Some researchers suggest that immune responses are mainly involved in protection from virus-associated cancers but not other forms of cancer (64).
Cancer is a common disease, and overt immune deficiency certainly is not necessary for its development. However, recent studies have shown that many cancers, including those that are not known to have a viral etiology, are seen with increased frequency in patients who have dysfunctional immunity. In a recent metaanalysis of cancer incidence in populations known to be immune deficient (e.g., organ-transplant recipients, patients with HIV infection), Grulich and co-workers (65) found an increased incidence of several common cancers, suggesting that impaired immunity can contribute to the development of cancer.
Components of the Immune System Involved in Antitumor Responses
Various types of human immune responses can target tumor cells. Immune responses can be categorized as humoral or cellular, a distinction based on the observation in experimental systems that some immune responses could be transferred by serum (humoral) and others by cells (cellular). In general, humoral responses refer to antibody responses;antibodies are antigen-reactive, soluble, bifunctional molecules composed of specific antigen-binding sites associated with a constant region that directs the biologic activities of the antibody molecule, such as binding to effector cells or complement activation (Fig. 2.2). Cellular immune responses generally refer to cytotoxic responses mediated directly by activated immune cells rather than by the production of antibodies (Fig. 2.3).
Nearly all immune responses involve both humoral and cellular components and require the coordinated activities of populations of lymphocytes operating in concert with each other and with antigen-presenting cells. These activities result in various effector functions such as antibody production, cytokine secretion, and the stimulation and expansion of cytotoxic T cells. Cellular interactions involved in immune responses include direct cell-cell contact, as well as cellular interactions mediated by the secretion of, and response to, cytokines. The latter are biologic messenger molecules that play important roles in the genesis, amplification, and effector functions of immune responses.
T lymphocytes play a pivotal role by acting as helper cells in the generation of humoral and cellular immune responses and by acting as effector cells in cellular responses. Cytotoxic T cells are effector T cells that can directly interact with, and kill, target cells by the release of cytotoxic molecules and the induction of target cell apoptosis. T-lymphocyte precursors mature into functional T lymphocytes in the thymus, where they learn to recognize antigen in the context of the major histocompatibility complex (MHC) molecules of the individual. Most T lymphocytes with the capability of responding to self-antigens are removed during thymic development. T lymphocytes are distinguished from other types of lymphocytes by their biologic activities and by the expression of distinctive cell surface molecules, including the T-cell antigen receptor and the CD3 molecular complex. T lymphocytes recognize specific antigens by interactions that involve the T-cell antigen receptor (Fig. 2.2) (66).
Figure 2.2 The basic immunoglobulin structure.
Figure 2.3 Cell-mediated cytotoxicity: two different types of cell-mediated cytotoxicity.
There are two major subsets of T lymphocytes: T helper/inducer cells, which express the CD4 cell surface marker; and T suppressor/cytotoxic cells, which express the CD8 marker. CD4 T lymphocytes can provide help to B lymphocytes, resulting in antibody production, and also can act as helper cells for other T lymphocytes. Much of the helper activity of T lymphocytes is mediated by the production of cytokines. CD4 T cells have been further subdivided into TH1 (cellular immunity/proinflammatory) and TH2 (antibody response-promoting) subsets, based on the pattern of cytokine production and the biological properties of these cells. Recent studies have identified a subset of T cells that inhibit autoreactive cells, perhaps acting to prevent autoimmune responses (67). This subset of T cells has been called regulatory T cells. Other recently described T-cell subsets include TH17 cells, which are important in driving immune responses to bacteria and fungi (68,69).
The CD8 T-lymphocyte subset includes cells that are cytotoxic and can directly kill target cells. A major biologic role of such cytotoxic T lymphocytes is the lysis of virus-infected cells. However, cytotoxic T lymphocytes can directly mediate the lysis of tumor cells. Effector T cells also can contribute to antitumor immune responses by producing cytokines, such as tumor necrosis factor (TNF), that induce tumor cell lysis and can enhance other antitumor cell effector responses.
Both CD4 and CD8 T cells respond to antigen only when it is presented in the context of MHC molecules on antigen-presenting cells or target cells or both. The T-cell receptor on CD4 T cells is restricted to responding to antigen plus MHC class II molecules; the receptor on CD8 T-cells is restricted to responding to antigen plus MHC class I molecules. In addition, both T-cell subsets require a second simultaneous costimulatory signal for optimal stimulation in the absence of which the T cells may be induced to enter a state of unresponsiveness or even apoptosis. Therefore, provision of effective costimulatory signals is necessary for the induction of effective antitumor responses by activated T cells.
B lymphocytes are the cells that produce and secrete antibodies, which are antigenbinding molecules (Fig. 2.2). B lymphocytes develop from pre-B cells and, after exposure to antigen and appropriate activation signals, differentiate to become plasma cells—cells that produce large quantities of antibodies. Mature B lymphocytes use cell-surface immunoglobulin molecules as antigen receptors. In addition to producing antibodies, B lymphocytes play another important role: They can serve as efficient antigen-presenting cells for T lymphocytes. Although the production of antitumor antibodies does not appear to play a central role in host antitumor immune responses, monoclonal antibodies reactive with tumor-associated antigens have proved to be very useful in antitumor therapy, as well as in the detection of tumors or of tumor-associated molecules. Unfortunately, no truly unique tumor-specific antigens have been identified, and most tumor-related antigens are expressed to some extent on nonmalignant tissues. Also, because some monoclonal antibodies are of murine and not human origin, the host's immune system can recognize and respond to murine monoclonal antibodies. This has led to the development of“humanized” monoclonal antibodies (genetically engineered monoclonal antibodies composed of human constant regions with specific antigen-reactive murine variable regions), with the aim of avoiding many of the problems associated with the administration of murine monoclonal antibodies.
Macrophages and dendritic cells also play key roles in the generation of adaptive, lymphocyte-mediated immune responses by acting as antigen-presenting cells. Helper/inducer (CD4) T lymphocytes, bearing a T-cell receptor of appropriate antigen and self-specificity, are activated by antigen-presenting cells that display processed antigen combined with self-MHC molecules (Fig. 2.2). Antigen-presenting cells also provide costimulatory signals that are important for the induction of T-lymphocyte activation. In addition to serving as antigen-presenting cells, macrophages can ingest and kill microorganisms and act as cytotoxic antitumor killer cells. These cells also produce various cytokines, including IL-1, IL-6, chemokines, IL-10, and TNF, which are involved in many immune responses. These monocyte-produced cytokines can have direct effects on tumor cell growth and development, both as growth-inducing and growth-inhibiting factors.
Natural killer (NK) cells are cells that have large granular lymphocytic morphology, do not express the CD3 T-cell receptor complex, and do not respond to specific antigens. NK cells can lyse target cells, including tumor cells, unrestricted by the expression of antigen or self-MHC molecules on the target cell. Therefore, NK cells are effector cells in an innate (non-antigen-restricted) immune response and may play a vital role in immune responses to tumor cells. The cells that can effect antibody-dependent cellular cytotoxicity (ADCC) are NKlike cells.
Cytokines are soluble mediator molecules that induce, enhance, or effect immune responses. Cytokines are produced by various types of cells and play critical roles not only in immune responses but also in biologic responses outside of the immune response, such as hematopoiesis or the acute-phase response. T helper 1(TH1) and TH2 cells, which control the nature of an immune response by secreting characteristic and mutually antagonistic sets of cytokines (9,10,11), are defined by the cytokines they produce. TH1 clones produce IL-2 and IFN-, whereas TH2 clones produce IL-4, IL-5, IL-6, and IL-10. TH1 cytokines promote cellmediated and inflammatory responses, whereas TH2 cytokines enhance antibody production. Most immune responses involve both TH1 and TH2 components.
Research has identified CD4-positive T cells that participate in the maintenance of immunologic self-tolerance by actively suppressing the activation and expansion of self-reactive lymphocytes. These cells are called regulatory T cells, or Treg cells. Treg cells are characterized by the expression of CD25 (the IL-2 receptor-chain) and the transcription factor FoxP3 (70,71). Treg cell activity is thought to be important in preventing the development of autoimmune diseases. Removal of Treg also may enhance immune responses against infectious agents or cancer. Although much remains to be learned about the role of Treg activity in antitumor immunity, it is clear that such cells may play a role in modulating host responses to cancer.
There is great interest in developing effective biologic and immune therapies for gynecologic malignancies. For example, patients with small-volume or microscopic residual peritoneal ovarian cancer are attractive candidates for immunotherapy or biologic therapy, especially approaches based on regional peritoneal immunotherapy or biotherapy (72,73). Also, many patients with advanced disease are immunocompromised, suggesting a role for immuneenhancing therapeutic approaches. Dysplastic and cancerous cervical epithelial cells infected with HPV, an oncogenic virus, also present an attractive target for immune enhancement-based therapeutic strategies, including the development of therapeutic vaccines for HPV. Advances in molecular biology, biotechnology, immunology, and cytokine biology have resulted in the availability of many new, promising immunotherapeutic approaches for gynecologic cancers.
The state of the art in immunotherapy for ovarian cancer and other gynecologic malignancies has been considered and described in detail in several recent excellent reviews (74,75,76,77), and the reader is referred to these publications for more detailed information on immunotherapy for gynecologic cancers. Examples of the current use of immunotherapy in clinical trials, both monoclonal antibody-based and therapies-based on enhanced cellular immune responses, are provided below.
Monoclonal Antibodies and Antibody-Based Immunotherapy
Monoclonal antibodies have played an important role in both the development of immunotherapeutic agents and tumor markers. Monoclonal antibodies also have been used for radioimmunodetection (78,79) and are being used for treatment. Monoclonal antibodies can potentially induce antitumor responses in various ways: (i) by complement activation and subsequent tumor cell lysis; (ii) by directly inducing antiproliferative effects, perhaps by interaction with tumor cell surface signaling molecules; (iii) by enhancing the activity of phagocytic cells, which can interact with immune complexes containing monoclonal antibodies; and (iv) by mediating ADCC via interactions of the Fc portion of monoclonal antibodies with Fc receptors on cells that mediate ADCC (80). In addition, monoclonal antibodies can be labeled with either radioactive particles or antitumor drugs and used to focus these agents onto tumor cells (81). In fact, some monoclonal antibody-based drugs are currently approved and being used for the treatment of cancer with great success. These FDA-approved monoclonal antibody-based anticancer drugs include: bevacizumab (Avastin) for treatment of colon and lung cancer, cetuximab (Erbitux) for the treatment of colon and head and neck cancer, gemtuzumab (Mylotarg) for the treatment of acute myelogenous leukemia, rituximab(Rituxan) for the treatment of non-Hodgkin's lymphoma, and trastuzumab (Herceptin) for the treatment of breast cancer.
Several clinical trials have utilized monoclonal antibodies directed against ovarian cancer antigens, including CA125, folate receptor, MUC1 antigen, and tumor-associated glycoprotein 72 (76). Evidence that CA125 can act as a tumor antigen that stimulates humoral and cellular immune responses is derived from various in vitro studies and clinical trials. Oregovomab (B43.13) is a murine monoclonal antibody to CA125 that has been used for the treatment of ovarian cancer. The antibody binds to circulating CA125, resulting in the formation of immune complexes (antibody-antigen complexes). These immune complexes are recognized as foreign, mainly because of the murine component. They are taken up by antigen-presenting cells, allowing the processing of the autologous CA125 antigen, ultimately leading to induction of CA125-specific antibodies, helper T cells, and cytolytic T cells.
In 2004, Berek et al. reported on the use of oregovomab for maintenance therapy in patients with ovarian cancer after first-line treatment. A subgroup of patients with favorable prognostic factors had a significantly longer time to relapse compared to patients in the placebo group (82). A five-year follow-up report in 2008 documented a median survival of 57.5 months for the oregovomab group and 48.6 months for the placebo group (83). These differences were not statistically significant. However, the velocity of the rise in CA125 levels at relapse was found to be a highly significant predictor of postrelapse outcome.
Another antibody network-based strategy has employed anti-idiotype vaccines in patients with relapsed ovarian cancer. ACA125 is a murine anti-idiotypic antibody that mimics an antigenic epitope on CA125 (75,76,84,85). Therefore, antibodies generated to ACA125 have the potential to react with antigenic epitopes on CA125, with ACA125 serving as an antiidiotype vaccine that would enhance immune responses to CA125 (75). Treatment with ACA125 resulted in both humoral and cellular responses, and those patients who had detectable anti-ACA125 responses showed a longer mean survival time than those who did not develop responses (86,87).
Abagovomab is an anti-idiotypic antibody that mimics the CA125 antigen. The initial results of abagovomab treatment in patients with ovarian cancer were reported by Sabbatini et al. and showed that all patients developed an anti-idiotypic antibody response (Ab3) (88). In addition, the generation of T-cell immunity to CA125 was demonstrated in five patients. While patients had measurable serum CA125 levels in both trials, neither trial analyzed CA125 expression in tumor tissue. A large international, multicenter trial is underway to investigate the effect of abagovomab as consolidation treatment in patients with ovarian cancer.
Adoptive immunotherapy involves the ex vivo expansion of antitumor immune cells followed by the administration of such effector cells. It has provided another immune system-based approach for antitumor therapy (74,75,89,90,91,92). Adoptive immunotherapy, involving the infusion of large numbers of autologous ex vivo-activated immune effector cells, has been shown to produce tumor regression in various animal and human tumors (90), and it has been seen to produce the best results seen to date in tumor immunotherapy (74). This approach can provide large numbers of tumor-specific T cells with the capacity to specifically kill tumor cells and can potentially lead to the complete elimination of residual tumor cells (74).
Early approaches used peripheral blood mononuclear cells exposed to IL-2 ex vivo to lead to the generation of lymphokine-activated killer (LAK) cells that are cytotoxic for a variety of tumor cells (93,94). Although experimental treatment of human subjects with LAK cells and IL-2 yielded some responses, considerable toxicity was seen (72,89,90,91,92,94,95,96,97,98). Given that (1) the overall response rate to LAK treatment is low, (2) this type of adoptive immunotherapy can result in high morbidity, and (3) it is impractical in most medical settings, adoptive immunotherapy with LAK cells does not appear to be a practical option for the treatment of ovarian cancer. The use of immunotherapy based on ex vivo-stimulated tumor-infiltrating lymphocytes or tumor-associated lymphocytes from ascites, with or without added IL-2, also has been examined in ovarian cancer (74,92,97,98).
It is clear that optimization of such adoptive immunotherapies is needed in terms of the cell source, the forms of stimulation, the methods for ex vivo expansion, and the cytokines that are given during such treatment (74). Potentially important refinements of these approaches include (i) the use of dendritic cells (DCs) as antigen-presenting cells (APCs) to stimulate T cells, (ii) the provision of effective costimulatory signals to the responding T cells by APC, and (iii) host conditioning with immunosuppressive chemotherapy before the adoptive transfer of cells (74,99).
Dendritic Cell and Tumor Vaccine Therapy
Cancers may develop or progress because immune system cells are not given a strong enough signal to become activated to destroy the tumor cells. In some cases, cancers are able to down regulate immune responses as cytokines or other molecules produced by tumor cells, such as IL-10 (100), can inhibit antitumor immune responses. It may be possible to counter this lack of antitumor immune responsiveness by enhancing APC activity, providing tumor-associated antigens in a manner that can better induce the generation of anti-tumor effector T cells (tumor vaccine therapy), or both.
A thorough overview of the rationale and design of potential tumor vaccines for ovarian cancer can be found in recent review papers (74,75). At this time, tumor vaccine approaches include (i) vaccination with defined tumor-associated antigens, or DNA vaccines that encode for tumorassociated antigens; and (ii) vaccination with whole tumor cell preparations with and without the coadministration of antigen-presenting cells such as dendritic cells. Adjuvants have been used to enhance the immunogenicity of tumor vaccines (74).
Various tumor-associated antigens are potential immunogens for tumor vaccines, including (i) differentiation antigens, (ii) new antigens created by mutation of genes encoding host cell proteins, (iii) molecules that are overexpressed on tumor cells (i.e., HER2, NY-ESO-1, CA125), and (iv) viral antigens from oncogenic viruses (i.e., HPV-encoded antigens) (75).Experimental tumor vaccine therapy in ovarian cancer has been carried out using the NY-ESO-1 antigen. Nearly half of epithelial ovarian cancers are NY-ESO-1 positive (101). Vaccination with a peptide from NY-ESO-1 resulted in the generation of both cellular and humoral immunity to this antigen, in most vaccinated patients (102). Vaccines based on HER2 also have been tested in ovarian cancer patients, with such treatment resulting in the induction of specific T-cell responses in most patients (103).
Human papilloma virus—specifically, HPV subtypes 16, 18, 31, and 45—has been implicated as the major etiologic agent in cervical cancer. HPV-infected dysplastic and cancerous cervical epithelial cells consistently retain and express two of the viral genes, E6 and E7, that respectively interact with and disrupt the function of the p53 and retinoblastoma tumorsuppressor gene products. Factors other than infection with HPV, such as cellular immune function, play an important role in determining whether the infection of cervical epithelial cells regresses or progresses to cancer. This has led to the development of prophylactic and therapeutic vaccines to HPV, as well as treatment approaches based on the enhancement of host immune function.
Human papilloma virus vaccines have been shown to have an exceptional level of efficacy (62), clearly reducing the incidence of both HPV-16 and -18 infections and HPV-16 and -18-related cervical intraepithelial neoplasia. The HPV vaccines Gardasil and Cerverix use HPV-like particles as immunogens to generate neutralizing antibodies for HPV. These findings suggest that HPV-based therapeutic cancer vaccines may also be effective for the control of cervical cancer (77).
The choice of target antigens for therapeutic vaccines to HPV positive cancers is of great importance and requires careful consideration of the expression of virus-encoded antigens in tumor cells (77). HPV E6 and E7 are attractive antigens for use in therapeutic vaccines because these HPV-encoded proteins are involved in cellular transformation and therefore are consistently expressed in HPV-positive tumor cells. Candidate therapeutic HPV vaccines include DNA vaccines, with recent research aimed at enhancing the potency and delivery of such vaccines by linking vaccine DNA to MHC class II-associated invariant chain (77,104) or by fusion of HPV DNA to the sorting signal of lysosomal-associated membrane protein type 1 to enhance MHC class II-mediated antigen presentation (77,105,106). Clinical trails with such DNA-based vaccines are currently being planned (77).
Dendritic cells are highly effective antigen-presenting cells and play a central role in the induction of both CD4 and CD8 T-cell responses. Dendritic cells can be pulsed with tumor antigen peptides or bioengineered to express tumor antigens, allowing them to be used in experimental therapies that aim to enhance antitumor immunity. Exposure of T cells to dendritic cells pulsed with ovarian cancer-derived antigenic preparations resulted in the generation of cytolytic effector T cells that could kill autologous tumor cells in vitro(107,108,109). In a phase I clinical trial, Hernando and co-workers (110) showed that patients with advanced gynecological malignancies could be effectively vaccinated with dendritic cells pulsed with a nontumor test antigen, keyhole limpet hemocyanin (KLH), and autologous tumor antigens. Lymphoproliferative responses to KLH and to tumor lysate stimulation were noted. The treatment was safe, well tolerated, immunologically active, and generally devoid of significant adverse effects.
Given the exceptional ability of dendritic cells to serve as potent antigen-presenting cells in the induction of T-cell responses, an attractive approach would be to combine tumor vaccines with ex vivo-generated dendritic cell preparations. Dendritic cells can be generated ex vivo from peripheral blood mononuclear cells using various strategies (74,111,112,113,114). Such ex vivo-generated dendritic cells can be pulsed with tumor antigens or vaccines or with DNA-or RNA-encoding tumor antigens, before administration, and have resulted in the induction of antitumor responses in preclinical studies (74,115,116,117).
Several major challenges need to be overcome for the successful development of effective DCbased therapies for ovarian cancer: (i) the identification of tumor-associated or tumor-specific antigens, (ii) the development of means to induce optimal DC maturation after antigen uptake, (iii) the development of schemes for generation of DCs that maintain optimal antigen-presenting cell activity and do not produce immunosuppressive factors, and (iv) the development of ex vivo expansion techniques that provide sufficient numbers of DCs for effective immunotherapy (74,118,119,120). As more is learned about the immunogenicity of current tumor-associated antigens, novel cancer-associated antigens are identified, and the techniques of DC activation and antigen expression are better developed, DC-based immunotherapy may provide a therapeutic alternative for the treatment of these cancers.
Biologic Response Modifier and Cytokine Therapy: Modulation of Host Immunity
Most early experimental biologic therapies for metastatic ovarian cancer involved biologic response modifiers such as Corynebacterium parvum (a heat-killed, gram-negative anaerobic bacillus), bacillus Calmette-Guérin (BCG), or modifications of these agents (121,122,123). Exposure to C. parvum resulted in the nonspecific enhancement of host immune responses, including the induction of an acute inflammatory response (123). Biologic response modifier therapy for ovarian cancer, including treatment with C. parvum and BCG, was examined in several studies (121,123,124,125). However, intraperitoneal (IP) treatment with C. parvum induced a profound local reaction, including peritoneal fibrosis, and its toxicity precluded more widespread testing.
Malignancies that tend predominantly to grow in the peritoneal cavity, such as residual ovarian cancer, have been treated in many experimental trials with IP drugs, most frequently with cytotoxic chemotherapeutic agents (126). IP biologic response modifier therapy, immunotherapy with cytokines, and gene therapy have been proposed and used for similar reasons. These approaches have the additional advantage of potentially inducing the activation of regional immune effector mechanisms in the peritoneal cavity (73). This might be particularly true for cytokine-based treatment strategies or for adoptive immunotherapies because activated immune effector cells may require direct contact with the malignant target cells for most effective antitumor activity.
Various cytokines have been tested in clinical trials, to date producing mixed results (74). This includes trials of IFN-α and IFN-γ, TNF-α, and IL-2. In recent studies, treatment of patients with refractory ovarian cancer with intravenous recombinant human IL-12, a TH1-inducing cytokine, resulted in disease stabilization in about half of the treated patients (127). IP treatment with IL-12 in patients with carcinomatosis from mesotheliomas, müllerian or gastrointestinal carcinomas, showed a 10% complete response rate and disease stabilization in nearly half of the treated patients (128). In a recent phase II trial, treatment with subcutaneously administered IL-2 and oral retinoic acid was reported to improve survival in patients who had ovarian cancer responding to chemotherapy (129).
The recent identification of T-cell subpopulations that have potent immunoregulatory properties, such as Treg cells and TH17 cells, provides new opportunities for the design of host immune system-modulating therapies with the aim of enhancing immune responses to cancer. Treg cells, which are immunoinhibitory regulatory T cells, can inhibit the induction of cytotoxic T cells and may thereby inhibit host antitumor immune responses. Increased levels of Treg cells in ascites, blood and tumor, have been seen in patients with advanced ovarian cancer (130). These Treg cells can inhibit antitumor immunity and promote tumor cell growth in ovarian cancer (131). Therefore, blocking the action of these Treg cells is clearly an important target in the development of new immunotherapeutic approaches to combat cancer. Studies aimed at blocking Treg activity in patients with cancer have been initiated using monoclonal antibodies targeting CD25, a cell-surface molecule commonly expressed on these cells. To date, these studies have not resulted in enhanced anticancer immune responses (132), and more refined approaches to target Treg need to be developed.
TH17 cells are another recently identified regulatory T-cell subpopulation, characterized by the secretion of IL-17. They have the ability to modulate Treg activity (68,69,133). The role of TH17 cells in cancer has not been clearly defined. In a recent animal study, it was reported that provision of IL-2 in the tumor microenvironment was associated with decreased TH17 activity and increased Treg activity (133). Although much more work needs to be done to define the role of TH17 cells in down regulating Treg activity, and perhaps in enhancing antitumor responses, future experimental treatment strategies aimed at enhancing TH17 activity may be of value in enhancing antitumor immune responses.