Medulloblastoma is a common malignant brain tumor of childhood. Human epidermal growth factor receptor 2 (HER2) is expressed by 40% of medulloblastomas and is a risk factor for poor outcome with current aggressive multimodal therapy. In contrast to breast cancer, HER2 is expressed only at low levels in medulloblastomas, rendering monoclonal antibodies ineffective. We determined if T cells grafted with a HER2-specific chimeric antigen receptor (CAR; HER2-specific T cells) recognized and killed HER2-positive medulloblastomas. Ex vivo, stimulation of HER2-specific T cells with HER2-positive medulloblastomas resulted in T-cell proliferation and secretion of IFN-; and interleukin 2 (IL-2) in a HER2-dependent manner. HER2-specific T cells killed autologous HER2-positive primary medulloblastoma cells and medulloblastoma cell lines in cytotoxicity assays, whereas HER2-negative tumor cells were not killed. No functional difference was observed between HER2-specific T cells generated from medulloblastoma patients and healthy donors. In vivo, the adoptive transfer of HER2-specific T cells resulted in sustained regression of established medulloblastomas in an orthotopic, xenogenic severe combined immunodeficiency model. In contrast, delivery of nontransduced T cells did not change the tumor growth pattern. Adoptive transfer of HER2-specific T cells may represent a promising immunotherapeutic approach for medulloblastoma. [Cancer Res 2007;67(12):5957-64]

Mature dendritic cells (DCs) are potent antigen presenting cells (APCs) that have been used in vaccine studies and adoptive immunotherapy protocols. For many clinical studies DCs are derived from monocytes in the presence of cytokines, which are expensive and often unavailable for clinical use. Here we describe a cytokine independent method for the differentiation of monocytes into APCs for the reactivation of antigen-specific memory T cells from both healthy donors and cancer patients. Contact activation of monocytes resulted in secretion of proinflammatory cytokines, such as IL-8, and increased cell surface expression of costimulatory molecules. To determine if activated monocytes (actMo) like DC can reactivate antigen-specific CTL, they were transduced with adenoviral vectors encoding the subdominant Epstein Barr virus antigens, latent membrane proteins (LMP) 1 and 2, which are expressed in Epstein Barr virus-positive malignancies. Stimulation of peripheral blood mononuclear cells with LMP1- and LMP2-expressing actMo activated LMP1- and LMP2-specific T cells, which could be further expanded with LMP1 or LMP2 expressing lymphoblastoid cell lines. The use of actMo as APCs simplifies the production/manufacture of antigen-specific T cells for clinical trials.

Dendritic cells (DCs) are essential for initiating and directing antigen-specific T-cell responses. Genetic modification of DC is under study for cancer immunotherapy, vaccine development, and antigen-targeted immunosuppression. Adenovirus (Ad) type 5 (Ad5)-mediated gene transfer to mouse bone marrow DCs and human monocyte-derived DCs is inefficient because neither express the cognate high-affinity Ads receptor. We show that co-precipitating adenoviral vectors with calcium phosphate (CaPi) increased gene expression (2000-fold) and transduction efficiency (50-fold) in mouse DC, primarily owing to receptor-independent viral uptake. Moreover, Ad5:CaPi-treated DCs were activated to express the maturation surface molecules CD40 and CD86, and to secrete proinflammatory cytokines tumor necrosis factor-alpha and interleukin 6. However, neither DC transduction nor maturation was dependent on viral protein interactions with cell surface integrin. Ad5:CaPi also transduced human DC more efficiently than Ad5 alone, similar to a genetically modified vector (Ad5f35) targeted to the CD46 receptor. As such, this approach combines the efficiency of adenoviral-mediated endosomal escape and nuclear trafficking with the receptor independence of nonviral gene delivery. Importantly, CaPi co-precipitation could be used to functionally modify DC to activate and expand cytomegalovirus-specific memory cytotoxic T lymphocytes. This study identifies a simple technique to improve the efficacy of current Ad5 gene transfer, in support of clinical adoptive immunotherapy.

No abstract

9008 Background: The long-term objective of this project is to develop an innovative HER2-targeted immunotherapeutic approach for medulloblastoma, the most common malignant brain tumor of childhood. HER2 is expressed in 40% of medulloblastomas and at present less than one third of patients with HER2-positive tumors are cured by conventional therapies. The aim of this study was to determine if T cells grafted with a HER2-specific chimeric antigen receptor (CAR) recognize and kill HER2-positive medulloblastomas. Methods: Mitogen-activated T cells from healthy donors and medulloblastoma patients were transduced with a retroviral vector encoding a HER2-specific CAR with a ζ-signaling domain (HER2T-cells). We analyzed the ability of HER2T-cells to 1) proliferate, 2) produce immunostimulatory cytokines (IFN-γ and IL-2), and 3) kill HER2-positive targets in cytoxicity assays upon exposure to HER2-positive primary medulloblastoma cells and cell lines. The in vivo function was tested in an orthotopic murine xenograft model of human medulloblastoma, which allows for serial imaging by bioluminescence. Results: HER2T-cells killed both HER2-positive primary medulloblastoma cells and cell lines in cytotoxicity assays, whereas HER2-negative tumor cells were not killed. Stimulation of HER2T-cells resulted in T-cell proliferation and secretion of IFN-γ and IL-2 in a HER2-dependent manner. No functional difference was observed between cells generated from medulloblastoma patients receiving dexamethasone and healthy donors. In vivo, the adoptive transfer of HER2T-cells resulted in sustained regression of established medulloblastomas in an orthotopic murine xenograft model as judged by bioluminescence imaging. In contrast, delivery of non-transduced T cells did not change tumor growth in comparison to untreated tumors. Conclusions: This study shows for the first time that HER2 is a target antigen for the immunotherapy of medulloblastoma. HER2-redirected T-cells not only recognized and killed HER2-positive medulloblastomas ex vivo, but also induced regression of experimental medulloblastoma in vivo. Hence, adoptive transfer of HER2-redirected T-cells may represent a promising immunotherapeutic approach for medulloblastoma. No significant financial relationships to disclose.