Description

Comparison of GD2 versus O-acetylated GD2 Specific CAR T cell immunotherapy in Preclinical Neuroblastoma Model Systems

Kelly-Spratt, KS., Kalinoski, H., Yokoyama, J., Baldwin, M., Chang, C., and Jensen, MJ.

Seattle Children’s Research Institute, Seattle, WA 98101

Ganglioside GD2 is highly expressed in neuroectoderm-derived tumors including neuroblastoma, osteosarcoma and Ewing’s sarcoma in children. Anti-GD2 antibodies have proven anti-neuroblastoma activity and are now a routinely used adjuvant to chemo/radiation therapy. One untoward feature of anti-GD2 antibody therapy is severe acute pain due to GD2 expression on peripheral nerve fibers. A tumor selective modification of GD2 by 9-o-acetylation on the terminal sialic acid can be detected on the cell surface of neuroblastomas, astrocytomas and schwannomas. Antibodies specific to o-acetylated GD2 do not bind to peripheral nerves, suggesting o-acetylated GD2 may be a more specific target for CAR T cell therapy. Here we report the design and optimization of single chain CARs directed by a humanized scFv to GD2 and a murine scFv specific for anti-o-acetylated GD2. In vitro analysis of GD2 and o-acetylated GD2 CARs revealed spacer length and single chain variable fragment (scFv; Vh and Vl) orientation are important factors that influence the cytolytic activity and effector cytokine production of CAR expressing T cells. In vitro chromium and cytokine release assays demonstrated the the human anti-GD2 scFv CAR is significantly more potent than the murine o-acetylated GD2-specific scFv CAR in eliminating tumor cells and producing cytokines. In vivo mouse models demonstrate intracranial GD2 CAR T cell adoptive transfer is efficient in eradicating the intracranial human SK-N-DZ neuroblastoma tumor xenografts, resulting in full clearance of tumor by 30 days post-injection. In comparison, the o-acetylated GD2-specific CAR T cells were capable of inhibiting the intracranial growth of the neuroblastoma cells, but only partially reduced total tumor cell burden by 30 days post-injection. Interestingly, the o-acetylated GD2 CAR T cells continued to persist long term in the mice and were able to clear the SK-N-DZ neuroblastoma cells by 120 days post-injection. Throughout the in vivo study, there were no observable neurotoxicities in mice treated with either the GD2 or the o-acetylated GD2 CAR T cells. Overall, GD2 may be a promising target for CAR T cell immunotherapy, and understanding the role of o-acetylated GD2 in neuroblastoma may clarify the validity of o-acetylated GD2 as a candidate target for T cell immunotherapy.