© 1999 AlphaMed Press
Clinical Applications of Dendritic Cell Cancer VaccinesUniversity of Pittsburgh Cancer Institute, Pittsburgh, Pennsylvania USA Correspondence: Joseph Baar, M.D., Ph.D., University of Pittsburgh Cancer Institute, Division of Hematology/Oncology, 200 Lothrop Street, Room MUH N-755, Pittsburgh, Pennsylvania 15213-2582, USA. Telephone: 412-648-6507; Fax: 412-648-6579: e-mail: BAARJ{at}MSX.UPMC.EDU
It has been more than 100 years since Dr. William B. Coley reported the regression of tumors in some patients who had been injected with bacterial extracts [1]. Since this early exercise in cancer vaccination, clinicians and scientists have attempted to harness the immune system to mediate the rejection of tumors in vivo. Because of many important recent discoveries in immunology and tumor cell biology, we now have the exciting opportunity to explore the therapeutic potentials of novel cancer vaccines.
It can be demonstrated in some transplantation challenge models that irradiated tumor cells express tumor antigens which are presented on the cell surface by major histocompatibility (MHC) class I or class II molecules and are capable of eliciting tumor-specific immune responses in syngeneic animals. Immunized animals reject subsequent challenges of the immunizing tumor and may occasionally reject unrelated syngeneic tumors expressing common antigens. In these models, tumor cell rejection is mediated by CD8+ cytotoxic T lymphocytes (CTL) which recognize a complex made up of an MHC class I heavy chain allele, a ß2-microglobulin light chain, and an 8-12 amino acid-long, tumor-derived peptide [2-4]. CD4+ helper T lymphocytes, in contrast, recognize MHC class II molecules presenting a 10 to 25 amino acid-long peptide [5]. Antigen-presenting cells (APCs) are critical for the antigen-specific priming of T cells, and dendritic cells (DCs) are the most potent stimulatory APCs. DCs constitute a heterogeneous population of cells [6], and although the precise ontogeny of DCs remains to be elucidated, they can differentiate from both bone marrow (BM) and peripheral blood precursors [7]. DCs exhibit at least five important characteristics for the generation of T-cell-mediated anti-tumor immunity ( Table 1) [8, 9].
DCs are readily obtained in large numbers from peripheral blood or CD34+ bone marrow progenitors expanded in vitro in medium containing various combinations of cytokines, including GM-CSF, interleukin 4 (IL-4), c-kit ligand, and tumor necrosis factor (TNF- ). Furthermore, of major importance in the design of DC-based cancer vaccines, it has been demonstrated that functional DCs can be generated in cancer patients from CD34+ BM progenitors mobilized in vivo with high-dose cyclophosphamide, IL-3, and G-CSF [10]. Also, administration by s.c. injection to mice of a newly described cytokine, flt-3 ligand, has been shown to expand both lymphoid and myeloid DCs in vivo [11]. The phenotype of DCs varies according to the cytokines added to the culture media. For example, media containing IL-4 and GM-CSF [12, 13] appear to stimulate the outgrowth of "immature" DCs, which have high phagocytic capacity for antigen but low expression of costimulatory molecules such as CD80 and CD86, or other molecules such as MHC class I and class II, CD83 and p55. The main role of "immature" DCs appears to be the uptake and processing of antigen in tissues. Addition of TNF- or monocyte-conditioned medium to DC cultures results in the outgrowth of more "mature" DCs, characterized by high expression of MHC class I, class II, and costimulatory molecules, but with little or no phagocytic capacity. The main role of "mature" DCs appears to be the priming of T cells in lymphoid tissues. Immature and mature DCs form an evolutionary biological continuum rather than two distinct populations of DCs: the uptake of antigen by immature DCs rapidly leads to phenotypic changes in DCs associated with maturation.
The application of autologous DCs to cancer therapy has received much interest [14-16]. Animal models have clearly documented the ability of syngeneic BM-derived DCs prepulsed with tumor-derived peptide epitopes or genetically engineered to express immunostimulatory cytokines to serve as effective immunogens of antitumor CTLs in vivo [17-22]. A number of tumor-associated antigens have been identified as potential immunogens in DC-based vaccination strategies. Such tumor rejection antigens derive, for example, from oncogenes (ras) [23], overexpressed genes (HER-2/neu) [24, 25], embryonic genes (MAGE, BAGE, GAGE) [26-28], normal differentiation genes (MART-1/ Melan-A, gp100, tyrosinase) [29-32], viral genes (HPV) [33], tumor-suppressor genes (p53) [34], B-cell idiotypes [35], and other tumor-associated proteins (PSMA, MUC1) [36, 37]. The major advantage of such peptide-based immunization resides in the fine specificity of the CTL reactivity raised against the tumor epitope of interest. However, there are two important limitations to peptide-based vaccination strategies. First, the majority of identified tumor peptides are presented in an HLA-restricted fashion, and therefore such vaccines are available only to patients with specific HLA haplotypes. Second, because of tumor heterogeneity, not all tumor cells may express the target epitope of interest. Given these limitations, other sources of tumor antigen are being sought for DC-based vaccines. These include tumor-derived RNA [38], tumor-derived apoptotic bodies [39], and tumor lysates [40, 41]. Using these strategies, there is no need to identify a priori the relevant tumor antigen(s) for immunization. Furthermore, because whole-cell products are being used, there is the possibility of immunizing against more than one tumor-derived antigen. Finally, because autologous tumor cells serve as the source of antigen, such vaccines are available to all patients, irrespective of HLA type.
The evidence that DCs can mediate the in vivo rejection of established tumors in murine models and the relative ease with which it is possible to generate large numbers of DCs in vitro has made them feasible components in human cancer vaccine protocols. In a report from Hsu et al. [35], autologous DCs were pulsed ex vivo with tumor-specific idiotype protein to stimulate host antitumor immunity in four patients with follicular non-Hodgkin's B-cell lymphoma. Patients received a series of three or four infusions of antigen-pulsed DCs followed by s.c. injections of soluble antigen two weeks later. Antitumor cellular immune responses were detected in all four patients, and three of the four patients had either partial or total regression of detectable disease. Nestle et al. [40] have reported on the vaccination of 16 melanoma patients with peptide- or tumor lysate-pulsed DCs. Keyhole limpet hemocyanin was added as a CD4 helper antigen and immunological tracer molecule. DC vaccination induced a positive peptide-specific delayed-type hypersensitivity (DTH) response in 11 patients. Five of 16 patients demonstrated objective responses to the DC vaccine (two complete responses, three partial responses) with regression of metastases in various organs (skin, soft tissue, lung, and pancreas). In a report from Chakraborty et al. [41], patients with malignant melanoma were immunized with a tumor cell lysate-loaded autologous APC-based vaccine. Seventeen patients (11 with metastatic disease) were immunized intradermally with the vaccine in a phased dose escalation (105-107 cells/injection) monthly for four months. One patient had a partial regression of an s.c. nodule. Nine patients had a DTH response at the vaccine site. Five of nine vaccine-infiltrating lymphocyte (VIL) specimens were predominantly CD8+. Antigen-specific CD8+ T-cell responses were detected in three of the five CD8+ VIL specimens. In a report by Mukherji et al. [42], patients with malignant melanoma who were HLA-A1+ and whose tumors expressed the MAGE-1 gene were immunized with a vaccine consisting of a MAGE-1 nonapeptide (EADPTGHSY) pulsed onto autologous APCs. Vaccination induced autologous melanoma-reactive and peptide-specific CTL responses. In particular, the frequency of circulating autologous melanoma-reactive CTL precursors was increased. In a follow-up study, Hu et al. [43] showed that in vitro stimulation of the post-immunization peripheral blood lymphocytes with the MAGE-1 nonapeptide loaded onto APCs resulted in a significant expansion of peptide-specific and autologous melanoma-reactive CTL responses. Salgaller et al. [44] have reported on DC-based immunotherapy of prostate cancer. The authors performed a phase 1 clinical trial assessing the administration of autologous DCs pulsed with an HLA-A0201-specific prostate-specific membrane antigen (PSMA) for the treatment of 51 men with hormone-refractory prostate cancer. Participants were divided into five groups receiving four or five infusions of peptides alone (PSM-P1 or PSM-P2; groups 1 and 2, respectively), autologous DC (group 3), or DC pulsed with PSM-P1 or P2 (groups 4 and 5, respectively). Immune reactivity against PSM-P2 was detected in HLAA2+ patients infused with DC pulsed with PSM-P1 or -P2 (groups 4 and 5). An average decrease in PSA was observed only in group 5. Seven men had partial responses. Wen et al. [45] have reported on idiotypic (Id) protein-pulsed DCs in the immunotherapy of a patient with multiple myeloma (MM). DCs were pulsed with autologous Id protein and administered to a patient with advanced-stage refractory MM. Id-specific immune responses were detected, as characterized by T-cell-proliferative responses and the production of anti-Id antibodies. A T-cell line generated after vaccination lysed autologous Id-pulsed targets and recognized fresh autologous MM cells. All of these studies have demonstrated that the administration of the various DC vaccines is safe with little or no toxicity.
DC-based cancer vaccines offer the potential for an effective, non-toxic, and outpatient-based approach to cancer therapy. As depicted in Figure 1, future-generation clinical trials will undoubtedly incorporate DCs pulsed with tumor epitopes derived from newly identified tumor-associated peptides, RNA, lysates, or apoptotic bodies. These tumor-derived compounds would be processed by DCs for presentation in the context of both MHC class I and class II molecules for the priming of T cells. DCs might also be genetically modified with cDNAs encoding, for example, immunostimulatory cytokines such as IL-12 to augment the generation of effective anti-tumor CTL responses, or with full-length cDNAs encoding TAAs. While the majority of DC-based clinical trials is being undertaken in patients with melanoma, such trials will eventually serve as a basis for future DC vaccine trials incorporating patients with more common malignancies such as breast, colon, lung, and prostate cancer. As a result of the major advances in cancer immunobiology, we now find ourselves at the threshold of a novel, exciting, and hopefully rewarding approach to the treatment of cancer. Undoubtedly, Dr. Coley would be amazed and gratified by the progress achieved in the immunotherapy of cancer.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||