Previously, the development of immune-based therapies has primarily focused on vaccines and cytokines, yielding benefit in a small percentage of patients. Recent advances in our understanding of the function of costimulatory molecules have revitalized enthusiasm in the development of immune therapies for cancer. This family of proteins possesses properties involved in both lymphocyte activation and immune-inhibitory functions. The costimulatory molecule with the greatest translation into the clinic thus far is CTL-associated antigen-4 (CTLA-4). CTLA-4 engagement leads to T-cell inhibition by two principle mechanisms. The first involves competitive binding with CD28 for B7 on the antigen-presenting cell. The second is direct intracellular inhibitory signals mediated by the CTLA-4 cytoplasmic tail. Numerous clinical trials testing the blockade of CTLA-4 signaling with fully human monoclonal antibodies have treated a variety of cancers, with the most experience in the treatment of metastatic melanoma. Significant antitumor activity as well as potential autoimmune-related toxicities have been observed. Further clinical investigation with CTLA-4 blockade, planned clinical trials testing manipulation of other costimulatory molecules, and continued improvement in understanding of costimulatory pathways present a new era of immune therapies for cancer patients.
The development of an effective immune response to an antigen requires the presence of stimulatory signals while evading negative regulatory or inhibitory ones. The immunoglobin superfamily of costimulatory molecules plays significant roles in the development of innate immunity as well as the balance of anergy to self-antigens. Importantly, alterations in the costimulatory pathways have been implicated in the development of autoimmune diseases. Relevant to cancer, these mechanisms may hinder the ability to develop effective antitumor immune responses to malignant cells that primarily possess self-antigens. As a result, means of overcoming such inhibitory signals offer exciting opportunities for the development of novel therapeutics for cancer. CTLA-4 has recently been the first member of this family to be extensively investigated. The blockade of CTLA-4 with the use of monoclonal antibodies in murine models revealed significant antitumor efficacy and such strategies are demonstrating tremendous promise in early clinical trials for a variety of cancer types.
Regulation of T-cell activation. Multiple signals are required for lymphocyte activation. Antigen-presenting cells (APC) when activated express high levels of costimulatory molecules and possess an enhanced ability to process antigens. To prime a T cell to a particular antigenic epitope, the professional APC must first process the full-length antigen protein via the immunogenic proteosome into peptides that are presented on the cell surface within the groove of the MHC to the T-cell receptor (TCR; Fig. 1A). As part of antigen-specific priming, the B7 molecule on the APC must also engage CD28 on the naïve T cell. This, along with other costimulatory molecular engagements, results in antigen-specific T-cell proliferation and cytokine production (1). Subsequent to its priming, the T-cell increases CTLA-4 transcription and expression with protein directed to the cell surface by its association with activator protein 1 (AP-1) within the Golgi apparatus. In addition to increased cell surface expression, TCR signaling also leads to phosphorylation of CTLA-4 by SRC kinases, accomplished through responding Lck and Zap-70 (2, 3). As a result of this tyrosine phosphorylation, binding of the cytoplasmic tail of CTLA-4 with AP50, a subunit of the clathrin adaptor AP-2 protein, is prevented (2, 4, 5). This further limits CTLA-4 internalization, assuring persistent cell surface expression. Additional data suggest that T-cell activation also leads to increased cell surface expression of CTLA-4 from endosomes of previously internalized protein (2, 4), but this principle remains in dispute. Overall, as a result of increased CTLA-4 expression, the maximum inhibitory effects of its engagement are significantly greater in primed rather than naïve T cells.
Within the T cell, the binding of B7 to CD28 along with TCR engagement results in activation of phosphatidylinositol 3-kinase, which subsequently leads to activation of AKT/protein kinase B as well as nuclear factor-κB (6). This cascade promotes improved T-cell survival, cell proliferation, and cytokine production. However, CTLA-4 has higher binding affinity to B7. With subsequent increased expression of CTLA-4 on the surface of the T cell, the B7 and CTLA-4 interaction leads to a variety of intracellular signals resulting in reduction of TCR-mediated activation, including decreases in members of the mitogen-activated protein kinase pathway and c-Jun-NH2-kinase (ref. 7; Fig. 1B). There is evidence that the CTLA-4 receptor may associate with the serine/threonine phosphatase PP2A, thereby modulating these intracellular signaling cascades (8, 9). Particularly affected are the reductions in a number of transcription factors, including nuclear factor-κB, AP-1, and nuclear factor of activated cells (6, 10). The net result of CTLA-4 engagement is the decrease in cytokine production, particularly interleukin 2 and the expression of its receptor, and cell cycle arrest at G1 (11–13).
CTLA-4 has also recently been implicated in increasing the motility of T cells. Increased CTLA-4 expression, therefore, counteracts the TCR signal, limiting cell migration and affording proper contact between APC and T cells during activation (14). This physical property is a potentially complementary means that CTLA-4 contributes to decrease in T-cell proliferation and cytokine production by mechanically limiting cell to cell contact.
Immunoregulatory cells. In addition to its effects on antigen presentation, CTLA-4 exhibits another mechanism of influence by its presence on regulatory immune cells. Tremendous advances in our understanding of immunoregulation and characterization of cells responsible for this function have recently been made. A subset of regulatory cells, identified as CD4+CD25+ cells, are selected by high-affinity encounters in the thymus and express significant levels of CTLA-4 on their surface (15–18). One working hypothesis is that as a result of chronic exposure to self-antigens, increased expression of CTLA-4 on these regulatory cells halts proliferation and cytokine production for a population of high-affinity cells (1). Overall, the combined roles of CTLA-4 in its effects on antigen presentation and immunoregulation make it an attractive target for therapeutic intervention.
Although the innate role of costimulatory molecules like CTLA-4 may be primarily to prevent the development of autoimmunity, this necessary function also likely hinders a host's ability to mount an effective antitumor immune response. The understanding of the mechanisms of CTLA-4 function led to the suggestion that blocking its engagement would first lead to unopposed CD28 activation of T cells while suppressing or depleting the activity of CD4+CD25+ regulatory cells.
Preclinical animal models. The development of a lethal lymphoproliferative disorder in young CTLA-4–deficient mice illuminates the pivotal role of CTLA-4 in immune homeostasis (19, 20). As CD4+ but not CD8+ T-cell depletion reduces autoimmune disease in these mice, the activities of CTLA-4 are essential for normal helper T-cell regulation (21). From noting this importance of CTLA-4, means to improve antitumor immunity have been investigated in numerous animal models. The administration of antibodies that inhibit CTLA-4 function has elicited impressive tumor regression for a variety of immunogenic tumors (22). For example, injection of anti–CTLA-4 antibodies led to rapid rejection of both colon carcinoma 51Blim10 tumors that were engineered to express B7 and of wild-type tumors. Administration of anti–CTLA-4 after tumor implantation also resulted in tumor rejection (22). Similar results were revealed in the fibrosarcoma SA1N model.
In contrast to the biological activity witnessed in immunogenic murine tumors, CTLA-4 antibody blockade alone elicits minimal effects in poorly immunogenic transplantable tumor models. However, the combination of CTLA-4 blockade and vaccination with irradiated tumor cells engineered to secrete granulocyte-macrophage colony stimulating factor (GM-CSF) has resulted in improved antitumor effects in these models. Concurrent CTLA-4 blockade and vaccination is highly efficacious in the B16 melanoma, SM1 breast carcinoma, and transgenic adenocarcinoma of mouse prostate (TRAMP) models at protection against a subsequent challenge of wild-type tumor cells or the eradication of small burdens of preexisting tumor (23–26). Mice remained resistant to rechallenge 4 months later, indicating that rejection was accompanied by the induction of potent immunologic memory.
Whereas preclinical studies investigating the combination of CTLA-4 blockade and GM-CSF tumor cell vaccines reveal striking therapeutic synergies, the immune response in some cases involved the loss of tolerance to normal tissues. The majority of mice in the B16 melanoma model developed progressive fur depigmentation as the result of infiltration of mononuclear cells into the skin resulting in the destruction of normal melanocytes (23, 27). The TRAMP model was subsequently used to investigate this combination in a spontaneous tumor model (28). In this system, the androgen-regulated rat probasin promotor drives the SV40 Tag oncogene in prostatic epithelium, thus leading to the development of adenocarcinoma in mice by 14 to 16 weeks of age. Treatment with GM-CSF–secreting tumor cell vaccination and CTLA-4 blockade in tumor-bearing mice revealed marked antitumor immune responses, but with evidence of immune responses to normal prostate epithelium. Whereas similar antitumor synergies with this combination therapy were observed in the SM1 breast adenocarcinoma model, there was no evidence of immune reactivity to normal tissues. This suggests that potent antitumor immunity can be generated with this combination without the loss of tolerance.
A recent study in the B16 melanoma model examining the mechanisms of combining GM-CSF–secreting tumor cell vaccination with CTLA-4 blockade unexpectedly revealed that this combination therapy induced increased tumor infiltration of both T effector as well as T regulatory cells (29). This finding offers several possibilities for CTLA-4 blockade function associated with tumor rejection. First, it suggests that CTLA-4 blockade does not deplete T regulatory cells or alter their function. Therefore, there is likely the need for chronicity of exposure to CTLA-4 blockade in the continued presence of antigen. It further questions the limitations in CTLA-4 blockade on T regulatory cells. Because both effector and regulatory populations of infiltrating T cells increase as a result of CTLA-4 blockade, one possible mechanism that will need to be delineated further is the importance of the ratio of tumor-infiltrating T effector cells to T regulatory cells in the tumor microenvironment and the resulting antitumor effects.
Clinical trials. Recent clinical trials using fully humanized monoclonal antibodies that block CTLA-4 have revealed significant biological activity. Most reported trials have been completed in patients with advanced melanoma (30–37). Doses of antibody ranged from 3 to 15 mg/kg and were administered as a single agent or in combination with peptide vaccines, interleukin 2, dacarbazine chemotherapy, or following vaccination with autologous, irradiated tumor cells engineered to secrete GM-CSF (GVAX). Overall, 6% to 21% of patients were noted to have a clinical response. Grade 3 or 4 toxicities, however, were primarily autoimmune and ranged in incidence from 15% to 43%. Diarrhea with bowel inflammation was the leading autoimmune event reported and included several episodes of bowel perforation. In addition, there have also been reported cases of immunomediated hypophysitis (38), hepatitis, nephritis, and symptomatic skin rashes that required medical intervention. Some investigators have correlated the development of such significant autoimmunity with favorable clinical outcomes. Multiple phase III studies in melanoma are currently being completed, testing varying doses of antibodies as well as combinations with peptide vaccines and chemotherapy in both first line and previously treated patients.
CTLA-4 blockade with fully human monoclonal antibodies has also been tested in renal cell cancer, prostate cancer, and ovarian cancer, but in significantly smaller numbers to that in melanoma. Of increasing clinical interest is the further testing of combination therapies. In hormone-refractory prostate cancer patients, for example, the combination of CTLA-4 blockade and systemic GM-CSF has shown clinical responses by prostate-specific antigen and radiographic examinations with apparently less incidence of severe toxicities to that witnessed in melanoma patients (39, 40). Our own group has reported administering CTLA-4 blockade to melanoma and ovarian cancer patients after autologous GVAX (30, 41). Biological activity was witnessed by biopsies of preexisting sites of disease in melanoma patients revealing extensive tumor necrosis associated with diffuse infiltration of multiple immune effector cells, as well as immune attack of the tumor vasculature. A majority of melanoma patients experienced significant reductions of tumor burden by standard response criteria or prolonged stable disease without developing clinically significant autoimmunity. In comparison, the ovarian cancer patients treated also experienced significant clinical activity, with a small percentage developing diarrhea and colitis.
Great promise for combination therapies. The goal of cancer therapeutics is to develop clinically effective and nontoxic treatments for patients. The understanding of the intracellular pathways involved with CTLA-4 signaling has led to extensive in vivo studies. Importantly, numerous clinical trials targeting CTLA-4 have revived interests to develop immunotherapies for the treatment of cancer. Despite such encouraging preliminary clinical results, not all patients benefit from CTLA-4 blockade and the possible toxicities of such treatment potentially limit its applicability. In addition, CTLA-4 represents one costimulatory target out of a complex family of biologically active molecules. To improve both antitumor efficacy and the therapeutic window of such therapy, investigation of combination therapies must be greatly expanded.
The obvious first consideration for combinations with CTLA-4 blockade is vaccines. With synergistic activity in preclinical animal models, clinical investigation thus far has only begun to understand these interactive mechanisms. Two key questions regarding combining vaccinations and CTLA-4 still require considerable investigation: (a) Does combining a vaccination strategy with blocking antibodies improve antitumor efficacy in patients? and (b) Does such a combination offer significant antitumor activity while limiting autoimmune toxicities by focusing on immune responses specific to tumors and within a therapeutic index? The designs of future clinical trials need to address these outstanding principles.
In addition to vaccines, synergistic means to improve the priming phase of the immune response in combination with CTLA-4 blockade need to be developed. The goal of such combination therapy would be to enhance and expand immune priming resulting in improved numbers and antitumor potency of immune effector cells. Cytokine combinations have been tried in a handful of patients, specifically using interleukin 2 and GM-CSF. Timing, dose, and schedule of combinations require rigorous study as these are likely important variables that influence biological outcomes. For example, GM-CSF contributes an important role to antigen presentation. Both preclinical murine models and early clinical development suggest synergy between vaccination with tumor cells engineered to secrete GM-CSF and CTLA-4 blockade. Interestingly, in prostate cancer, the combination of systemic GM-CSF and CTLA-4 blockade seems to have similar clinical benefits without significant autoimmune toxicities (39, 40). Therefore, this combination with various dose and schedules to maximize efficacy while limiting toxicity should be thoroughly investigated in prostate cancer, melanoma, as well as other tumor types. As another example, CD40 signaling is a critical pathway for activation of APCs. Clinical trials are currently under way in cancer patients using a CD40 agonist antibody. Preliminary results recently reported suggest clinical activity as a single agent in some patients, including a handful of melanoma patients (42). The improvement in APC function offered by CD40 activation suggests another important opportunity for combining with CTLA-4 blockade.
CTLA-4 is the clinical paradigm of manipulating costimulatory pathways with therapeutic intent (43). Currently, a number of additional family members are being targeted in early phase clinical trials, including PD-1, OX-40, and 41BB. The mechanisms of action in patients, clinical efficacy, and toxicity profiles are currently being determined. As the pathways involved are complex and have overlapping as well as counteractive effects, the goal of such combinations must be to improve efficacy and limit toxicities. A recent report in a preclinical murine model suggests that the combination of CTLA-4 blocking antibody with a 41BB agonist antibody provides similar antitumor activity without autoimmune effects (44). The future of clinical investigation is rich with such potential translation to benefit our patient populations.