Enhancing GPC2 CAR T-cell therapy against retinoblastoma
Retinoblastoma is the most common childhood cancer of the eye. With early detection and access to current optimal clinical management, retinoblastoma can be controlled in some children, although treatment-related toxicities remain including potential vision loss and development of secondary cancers. In cases where retinoblastoma has gone untreated or comes back after treatment, removal of the eye is still the only option available to prevent the cancer from spreading to the brain, where it is incurable. In recent years, a therapy called chimeric antigen receptor (CAR) T-cells, which engineers a patient’s own immune cells to fight cancer by enabling them to detect specific cell surface molecules, has achieved great success in the treatment of certain cancers. Our laboratory has developed CAR T-cells targeting a cell surface molecule called GPC2 that have progressed to patients with the pediatric solid tumor neuroblastoma. We have also shown that these GPC2 CAR T-cells are initially capable of controlling retinoblastoma growth, however our recent studies demonstrate that these eye tumors employ creative defense mechanisms to hide from or disable GPC2 CAR T-cells, limiting long-term treatment efficacy. As such, here we propose re-engineering GPC2 CAR T-cells with specific enhancements to overcome retinoblastoma immune evasion as a critical next step to achieving sustained antitumor activity.
Project Goal:
In this project we will develop next-generation GPC2 CAR T-cells enhanced to overcome two different mechanisms that retinoblastoma cells may employ to resist their tumor-killing activity. In our preliminary data, we have shown that some retinoblastomas escape GPC2 CAR T-cells by decreasing the amount of GPC2 displayed on their surface, thus enabling them to remain undetected from these potent immune cells. In our first approach, we propose to re-engineer GPC2 CAR T-cells with the ability to additionally detect a second retinoblastoma cell surface molecule called GD2, which can also be targeted with CAR T-cells and which we have demonstrated to remain at high levels on the retinoblastoma cell surface. Interestingly, not all retinoblastomas appear to down-regulate GPC2 in order to survive therapy, suggesting that other mechanisms may also be employed. We have shown that retinoblastomas secrete high levels of molecules that can inactivate T-cells, including one in particular called MIF. Thus, in our second approach, we propose to re-engineer GPC2 CAR T-cells with the ability to capture tumor-secreted MIF and convert it into an additional T-cell activating signal rather than one that disables them. The preclinical efficacy and safety studies that we propose here will provide proof-of-concept data propelling the clinical translation of enhanced GPC2 CAR T-cells for retinoblastoma and other pediatric tumors such as neuroblastoma that employ similar mechanisms of immune evasion.
