Encyclopedia of cancer
Specific immunotherapy for melanoma
Melanoma vaccines aim at inducing or enhancing cellular and humoral immune responses specifically against melanoma cells.
The earliest concepts for immunotherapy of malignant tumors were developed at the beginning of the twentieth century, when Paul Ehrlich suggested that the newly discovered antibodies be directed as “magic bullets” specifically against cancer cells. Various approaches to stimulate tumor immune defense have been experimentally developed in the last few decades. Patients with malignant melanoma have been of particular interest in the field of tumor immunology because clinical observations such as spontaneous regressions of primary tumors in the skin supported the idea that the immune system is capable of recognizing and eliminating melanoma cells. Immunological approaches were believed to be particularly suited to prevent the early metastatic spread of melanoma cells, which is associated with high mortality. Initial clinical studies performed with bacterial extracts derived from Corynebacterium parvum or bacillus Calmette-Guйrin as an adjuvant immunotherapy were initiated in the 1970s but were largely unsuccessful. Recombinant cytokines such as interferon-alpha (IFNα) and interleukin-2 (IL-2) were introduced for immunotherapy of melanoma in the 1980s. Encouraging results were noted in some patients with inoperable metastatic melanoma, which showed complete and long-lasting remissions following treatment with high-dose IL-2. The occurrence of vitiligo-like depigmentation and other autoimmune phenomena in patients treated with IFNα or IL-2 appears to correlate with a favorable clinical course, suggesting that an activation of the immune system against pigment cells is involved.
Tumor immune defense against transplantable tumors in mouse models
MacFarlane Burnet proposed in the 1950s that the immune system is principally able to recognize and eliminate cancer cells. This “immunosurveillance hypothesis” was controversially discussed for many years. The biologic understanding of immune defense against malignant tumor cells was widely studied in experimental mouse models involving the transplantation of chemically or virally transformed tumor cells. The analysis of humoral immunity against transplanted carcinogen-induced sarcomas revealed that the immune system is in principle capable of detecting molecular events associated with transformation such as the aberrant expression of p53. Adoptive transfer experiments showed that cellular components of the immune system were responsible for the rejection of transplanted tumor cells. In the 1980s it was discovered that cytotoxic T cells (CTL) recognize peptide fragments derived from intracellularly synthe- sized proteins bound to major histocompatibility (MHC) class I molecules on the cell surface and are able to specifically detect and destroy not only virus-infected but also tumor cells. Various experimental studies using well-characterized model antigens showed that recombinant vaccines were able to induce immunity against viral infection and transplanted tumor cells, provided that the tumor cells express the model antigen themselves.
Culture of melanoma-specific human T cells and identification of melanoma antigens
In the 1980s melanoma-specific CTL could be isolated from peripheral blood or tumor tissue of melanoma patients and propagated in vitro with the help of recombinant IL-2, a T cell growth factor. In the 1990s a number of melanoma antigens recognized by these melanoma-specific CTL could be identified on the molecular level. These belonged to three principal categories: (i) tumor-specific antigens such as the MAGE family of proteins which were termed “cancer-testis antigens” because they are expressed by tumor cells as well as by germ cells in the testes; (ii) lineage differentiation tumor antigens such as the tyrosinase family of enzymes, gp100, or MART/Melan-A which are expressed by melanoma cells as well as melanocytes; and (iii) individually mutated tumor antigens which in some cases even participated in malignant transformation. The identification of melanoma antigens recognized by T cells provided the scientific basis for the development of T cell-directed melanoma vaccines.
Rational development of melanoma vaccine strategies in animal models
Strategies for melanoma vaccination were experimentally developed over many years using the trans- plantable mouse B16 melanoma cell line in the syngeneic, inbred C57BL/6 mouse strain. Unlike experiments with carcinogen-induced sarcomas, vaccinations with irradiated B16 melanoma cells were not able to protect mice against a subsequent challenge with viable B16 melanoma cells showing that these melanoma cells were poorly immunogenic. Cytokine gene transfer such as the transduction of B16 melanoma cells with GM-CSF significantly improved their immunogenicity. This was later shown to be due to the ability of GM-CSF to recruit and activate antigen-presenting dendritic cells (DC) which are critical for the induction of an effective cellular antitumor immune response. The establishment of protocols for the ex vivo culture of DC from precursors in the bone marrow or peripheral blood with the help of recombinant GM-CSF enabled their use as a biologic vaccine adjuvant. Injections with GM-CSF-derived DC pulsed with B16 tumor cell lysates were also able to promote protective immunity against a subsequent challenge with viable B16 melanoma cells. With the discovery of MHC class I-restricted antigen recognition by T cells and the molecular identification of different melanoma antigens, active specific approaches for melanoma vaccination were developed using well-characterized model antigens. Vaccines employing defined synthetic MHC class I-binding peptides with appropriate adju- vants, recombinant plasmid DNA, or recombinant viral vectors were able to successfully protect against a subsequent challenge with antigen-transduced B16 melanoma cells. Cultured DC and stimulation of Toll- like receptors (TLR) with synthetic oligonucleotides, which activate the innate immune system, were required for the induction of an effective tumor immune response against lineage-specific differentiation antigens in a therapeutic experimental setting.
Efficacy of melanoma vaccines in clinical trials
Different strategies for therapeutic melanoma vaccination have been tested in clinical trials in the last 20 years. They can be grouped into vaccines on the basis of the following:
- Autologous tumor cells
- Allogeneic tumor cells
- Synthetic or recombinant molecules
The use of autologous tumor cells has the principal advantage that immune responses can potentially be stimulated against a broad spectrum of (unknown) tumor antigens. This includes the mutated proteins of the individual tumor. Both the generation of adequate amounts of autologous tumor cell lysates and the in vitro establishment of autologous melanoma cell lines require the acquisition of sufficient tumor material. This is only possible for patients with disseminated disease where tumor metastases can be easily excised. Clinical vaccine trials using autologous GM-CSF-transduced melanoma cells, cultured DC pulsed with autologous melanoma cell lysate, or purified heat-shock protein preparations have all shown that this strategy can successfully induce objective tumor regressions and long-lasting remissions in a small subset of patients. Unfortunately, most patients did not profit from the vaccine, but it needs to be considered that these trials were largely performed in patients with advanced disease.
The idea of inducing a tumor-specific immune response using a vaccine approach appears to be particularly attractive in patients with clinical stage III melanoma after complete operative elimination of all detectable locoregional metastases. Autologous melanoma is usually not available in this situation. Therefore, well-characterized allogeneic melanoma cell lines derived from other patients have been used for vaccine purposes. This approach is potentially able to induce immune responses against cancer-testis antigens and lineage-specific differentiation antigens. Clinical trials involving large numbers of patients have been performed over many years. The results of these trials, reported in the last couple of years, claimed that this type of melanoma vaccine can be as effective as adjuvant treatment with recombinant IFNα.
The molecular identification of melanoma antigens and the fact that many melanoma patients sponta- neously show tumor antigen-specific cellular immunity have also promoted the development of vaccine strategies with synthetic peptides and recombinant viral vectors. Peptide-based vaccines could be made increasingly efficient with the help of recombinant cytokines (such as GM-CSF, IFNα, or IL-12), cultured DC, or synthetic CpG oligonucleotides as a TLR9 agonist. Various recombinant viral vectors including adenovirus, vaccinia virus, or canarypox virus carrying cDNA for defined melanoma antigens were also successfully used in clinical trials. Optimized melanoma vaccine strategies could effectively induce or enhance specific immune responses against the MAGE family of cancer-testis antigens or against melanocyte differentiation antigens like tyrosinase, gp100, or MART/Melan-A in vivo. However, despite objectively measurable vaccine effects, only very few patients responded clinically with tumor regression and a longer-lasting remission.
Barriers for the successful implementation of therapeutic melanoma vaccines
Current research is directed towards understanding the reasons why melanoma vaccines are not as effective as expected in the clinical situation. Sensitive and reproducible procedures for the measurement of spontaneous and vaccine-induced immune responses against defined as well as unknown melanoma antigens were developed in the last decade. These include flow cytometric enumeration of specific CTL with recombinant MHC class I-peptide tetramers and evaluation of their ability to produce IFNg using intracellular cytokine staining or the Enzyme-Linked ImmunoSpot (ELISPOT) technique. The detection of melanoma-specific CTL present in very low numbers in blood or tumor tissue in vivo could be achieved by in vitro expansion in limiting dilution assays.
Insights into the molecular and cellular mechanisms underlying the control of immune reactions and the maintenance of peripheral tolerance in recent years have also advanced the development of vaccines. It has become clear that the induction of CD8 + CTL is tightly controlled by regulatory lymphocyte populations via specific cell surface molecules and transcription factors such as CTLA-4 and FoxP3, respectively. In mice, selective elimination of regulatory T cells or blockade of CTLA-4 is able to circumvent peripheral tolerance mechanisms and enhance the induction of melanoma-specific cellular immunity in vivo. Stimulation of TLR was also shown to transiently suppress regulatory control and enhance vaccine responses.
Further important insights could be gained with a detailed analysis of cellular immune responses in blood and tumor tissue of individual melanoma patients, which showed remarkable clinical regressions following melanoma vaccination. Injections of autologous tumor cells together with cultured DC as an adjuvant could enhance existing spontaneous CTL responses against various melanoma antigens. Vacci- nations against the MAGE family of antigens were able to induce MAGE-specific CTL, which appeared to support the functional reactivation of preexisting CTL in the tumor tissue with other specificities. This observation demonstrated the importance of mechanisms regulating immune cell function within the tumor microenvironment, which is still not completely understood.
Tumor cells need to communicate with endothelial cells, fibroblasts, and immune cells to support progressive tumor growth, invasive spread, and angiogenesis. The interaction between tumor growth control, inflammatory responses, and cytotoxic immunity in the pathogenesis of cancer is actively studied by many groups in the field. For example, the production of cytokines such as IL-6 and TGFb in tumor tissue is able to stimulate cell proliferation and simultaneously suppress the function of CTL. Various mechanisms contribute to the escape of melanoma cells from recognition and destruction by the cellular immune system. For example, melanoma cells are able to impair the function of DC and T cells and can downregulate MHC class I-restricted processing and presentation of antigens. Novel experimental mouse models have been generated in the last few years where melanomas can be induced in the skin by UVand spontaneously metastasize in lymph nodes and visceral organs on the basis of defined genetic alterations in the germline. These genetically engineered mouse melanoma models, which portray the clinical situation in patients much more closely than the transplantation of melanoma cell lines, may help to understand the role of molecular and cellular mechanisms regulating immune cell function in the pathogenesis of melanoma and find ways to improve the therapeutic efficacy of melanoma vaccines.
Summary and future perspective
Various strategies for melanoma vaccination have been evaluated in clinical trials in the last 10–20 years. Several approaches, which had been rationally developed in experimental animal models, were clearly able to induce melanoma antigen-specific cellular and humoral immunity in patients. A favorable influence on the clinical course of the disease could only be achieved in a small minority of patients. Melanoma vaccines have disappointed the expectations. Most vaccine approaches have been evaluated in patients with advanced metastatic disease that had failed standard treatment options. Conceptually, melanoma vaccines should be much more effective in the adjuvant setting where the immune system can specifically detect and eradicate melanoma micrometastases in patients at high risk of recurrent disease. Melanoma vaccines may also become part of future treatment regimens for inoperable metastases which combine novel targeted antiproliferative, antiangiogenic, and immunostimulatory measures in order to attack tumors simultaneously from several sides. These approaches will have to aim at supporting the efficacy of melanoma-specific CTL in tumor tissue, for example, by creating the cytokine milieu associated with viral infections using appropriate synthetic oligonucleotides and by blocking regulatory mechanisms at the molecular or cellular level. Futuristic concepts include the adoptive transfer of ex vivo genetically engineered DC and CTL which are optimized for tumor cell destruction and may represent the “magic bullets” of the twentieth century.
Bhardwaj N (ed) (2007) Review series tumor immunology. J Clin Invest 117:1130–1212
Blattmann NJ, Greenberg P (2004) Cancer immunotherapy: a treatment for the masses. Science 305:200–205
Boon T, Coulie PG, Van den Eynde BJ, van der Bruggen P (2006) Human T cell responses against melanoma. Annu Rev Immunol 124:175–208
Parmiani G, Castelli C, Santinami M, Rivoltini L (2006) Melanoma immunology: past, present and future. Curr Opin Oncol 19:121–127
Rosenberg SA, Yang YC, Restifo NP (2004) Cancer immunotherapy: moving beyond current vaccines.Nat Med 10:909–915