596. Effect of Chemokine/Cytokine Delivered by Nanoparticles on Tumor Migration of Neuroblastoma Targeting Chimeric Antigen Receptor T Cells

Abstract

INTRODUCTION: Neuroblastoma is the most common extracranial solid tumor in children and morbidity and mortality of high risk disease remains significant. The ganglioside GD2 is overexpressed on neuroblastoma and T cells expressing chimeric antigen receptors (CARs) specific for GD2 (GD2.CAR T cells) have been shown to induce complete tumor remissions without toxicity in a clinical trial. However, insufficient migration of CAR T cells to solid tumors remains a major problem. XCL1 (lymphotactin) is a chemokine that attracts T cells and natural killer cells, however, like other chemokines, diffuses quickly when injected at the tumor site. Calcium Alginate Nanoparticles (CANs) have been used clinically to safely deliver vaccines and chemokines. We are now evaluating the use of CANs to deliver chemokines to the tumor microenvironment and improve tumor homing of GD2.CAR T cells. METHODS/RESULTS: In our in vitro experiments, XCL1 was released more gradually, with a change from 798 pg/µL detected after 1 hour to 1233 pg/µL at 50 hours when loaded with CANs compared to a decrease from 1434 pg/ µL to 398 pg/ µL at 50 hours without CANs as measured by ELISA and was able to attract CAR T cells. For in vivo experiments, T cells obtained from fresh peripheral blood mononuclear cells were activated on CD3 and CD28 coated plates and transduced 3 days later with firefly luciferase and GD2. CD28-OX40z or GD2. 41BBz retroviral supernatants with 70-95% transduction rate. Immunodeficient NSG mice were subcutaneously injected with GD2 expressing LAN-1 tumor cells and divided into different treatment groups which received IL-2 intraperitoneally twice weekly (n=5). Mice that received CANs loaded with XCL1 showed a 6.6-fold increase in CAR T cell signal on day 10 while peritumorally injected XCL1 alone or empty nanoparticles had no such impact on T cell migration. We subsequently incorporated IL-15 in our experiments based on its reported enhancement of T cell expansion in vivo. Without other exogenous cytokines, mice receiving the combination of IL-15 and XCL1 with CANs on days 0, 7, 12, and 19 showed remarkable improvement in T cell signal at the tumor site: 8.19 and 4.62-fold increase compared to groups treated with CANs only and XCL1 loaded with CANs on day 7, respectively. As early as 6 days after T cell injection and the first CANs injections, mice in the control CANs had increased tumor burden, nearly double that of IL-15/XCL1/CANs group with intermediate tumor burden in the CANs/XCL1 group. The combination of both chemokine and cytokine delivered by CANs also had a positive impact on survival, which was lengthened to >45 days compared to 26 days in the CANs only control group. CONCLUSION: Calcium Alginate Nanoparticles loaded with a combination of both XCL1 and IL-15 achieved superior T cell expansion, tumor control and better survival outcome compared to controls. These findings indicate that CANs are a promising delivery vehicle for chemokines in combination with adoptive T cell therapies. In the future we will attempt to optimize dosing and plan to transduce T cells with selected phenotypes to further enhance T cell persistence.