(P 121) IMPROVING THE EFFICACY OF GD2 CHIMERIC ANTIGEN RECEPTOR T-CELLS USING HISTONE DEACETYLASE INHIBITORS

Objective: Chimeric Antigen Receptor (CAR) T-cell treatment has transformed the landscape of treatment for relapsed / refractory B-lineage hematologic malignancies. Treatment of solid tumors with CAR T-cells is more challenging, in part due to heterogeneous expression of CAR antigens on most solid tumors. Neuroblastoma has uniformly high GD2 expression and has been successfully targeted with GD2.CAR T-cells at our and other centers, but responses for osteosarcoma and other solid tumors with heterogeneous GD2 expression have been mixed. This lack of high and uniform GD2 expression is likely a major factor limiting the success of GD2.CAR T-cells for osteosarcoma and other sarcomas.

HDAC inhibitors (HDACi), an epigenetic modifier drug class, are known to increase GD2 expression on tumors in vitro. We hypothesize that these drugs can be used in combination with GD2.CAR T-cells to synergistically improve tumor killing, both due to increased GD2 expression and due to the anti-tumor effect of HDACi themselves. Limited data exists on the effect of HDACi on T-cell proliferation and function. We wanted to determine whether any HDACi could (i) boost GD2 expression while (ii) maintaining CAR T-cell function.

Methods: In an initial step we analyzed patient tumor tissue samples by immunohistochemical staining for GD2. To investigate the effects of HDACi on CAR T-cells, we incubated GD2.CAR T-cells with clinically relevant concentrations of different HDACi and measured their proliferation in vitro. We also evaluated the cytotoxicity of GD2.CAR T-cells against osteosarcoma cells, comparing cells pre-treated with an HDACi versus cells without pre-treatment. In vivo, we established subcutaneous osteosarcoma tumors in NSG mice, then treated them with varying doses of oral HDACi with and without GD2.CAR T-cells to measure tumor GD2 expression and tumor response.

Results: We analyzed 95 patient tumor tissue samples and found GD2 expression to be highly variable in osteosarcoma and other pediatric solid tumors, both by intensity and by proportion of positive cells. (Table 1) Next, we confirmed that HDACi can increase GD2 expression on osteosarcoma cell lines. (Figure 1) We determined that survival and expansion of GD2.CAR T-cells in vitro are limited by HDACi in a concentration-dependent manner with the exception of vorinostat. CAR T-cell expansion decreased by 80% or more by day 21 in 3 of 4 tested HDACi, whereas vorinostat only reduced growth by 20% at this same timepoint. In a xenograft model of osteosarcoma, the addition of vorinostat to a subtherapeutic dose of GD2.CAR T-cells resulted in a moderate decrease in tumor growth (mean tumor volume of 292 mm3 at day 29 compared to volume of 895 mm3, p< 0.05) when compared to CAR T-cells alone. (Figure 2) GD2.CAR T-cells persisted and expanded in vivo despite simultaneous and continued oral HDACi administration.

Conclusion: We have shown GD2 expression to be variable and heterogeneous in many solid tumors, thus the ability to overcome this could have wide applications. HDAC inhibitors upregulate GD2 expression on tumor cells in vitro and in vivo. Our preclinical studies suggest vorinostat may be an ideal candidate to use in combination with GD2.CAR T-cells in vivo due to an absence of negative effects on CAR T-cells.