GPC2 CAR T Cell Targeted Delivery of BiKEs in High-Risk Neuroblastoma
Children with high-risk neuroblastoma continue to have poor outcomes despite maximal intensification of multimodal treatments. Neuroblastoma tumors continue to express lineage-specific cell surface molecules that can be exploited using immune-based therapies, as exemplified by the unprecedented clinical success of anti-GD2 monoclonal antibodies. However, anti-GD2 immunotherapy is usually associated with unacceptable toxicities such as difficult-to-treat pain, as GD2 is also expressed on healthy tissues such as peripheral nerves. Thus, the identification of additional cell surface targets is required to developed novel immunotherapies with improved efficacy and safety profiles. With this in mind, our laboratory applied an integrated discovery pipeline to identify cell surface molecules differentially expressed in neuroblastoma tumors compared to normal tissues, identifying glypican-2 (GPC2) as a new candidate immunotherapeutic target. Our recent efforts have focused on developing GPC2-targeted therapies including antibody drug conjugates and chimeric antigen receptor (CAR) T cells. GPC2 CAR T cells show strong initial anti-tumor activity in preclinical studies using neuroblastoma patient-derived xenograft models. However, most tumors relapse several weeks after T cell infusion which correlates with GPC2 down-regulation. In addition, the clinical efficacy of GPC2 CAR T cells might also be limited by the presence of immunosuppressive cells within the tumor microenvironment, as observed in neuroblastoma clinical trials using GD2-targeted CAR T cells. Thus, re-engineering GPC2 CAR constructs that overcome these mechanisms of immune evasion will be critical to achieve improved and sustained clinical anti-tumor activity.
In this project we will develop next-generation GPC2 CAR T cells that secrete bispecific natural killer (NK) engagers (BiKEs) targeting GD2 on tumor cells and CD16a on NK cells. We expect that these BiKEs will generate new tumor-NK synapses that will boost killing of neuroblastoma cells in synergy with GPC2 CAR T cell mediated tumor killing. Our newly designed CAR.BiKE approach will 1) Overcome immunotherapeutic target modulation occurring under the pressure of first-generation CAR T cells by targeting two antigens simultaneously, and 2) Bypass neuroblastoma immunosuppression by increasing tumor infiltration of NK cells. In addition, our approach will likely reduce common toxicities related to GD2-based therapies, as BiKEs will be locally secreted at the tumor site but not in healthy normal tissues. To address our hypothesis, we will first study whether BiKEs are correctly secreted by GPC2 CAR T cells. Second, we will study whether secreted BiKEs activate the killing capacity of NK cells utilizing a panel of neuroblastoma cell lines that express different amounts of GPC2 and GD2. After the completion of these studies, we will use immunodeficient mice implanted with human neuroblastoma tumors and human NK cells to evaluate the ability of the CAR.BiKE treatment to increase tumor penetration of NK cells, induce tumor regressions and increase mice survival without affecting their healthy tissues. These preclinical studies will provide proof of concept data for further clinical translation. Altogether, we expect to develop next-generation CAR T cells with improved efficacy and safety that will circumvent critical mechanisms of CAR T cell resistance in neuroblastoma.