GD2-directed bispecific trifunctional antibody outperforms dinutuximab beta in a murine model for aggressive metastasized neuroblastoma

Abstract

Background Neuroblastoma is the most common extracranial solid tumor of childhood. Patients with high-risk disease undergo extremely aggressive therapy and nonetheless have cure rates below 50%. Treatment with the ch14.18 monoclonal antibody (dinutuximab beta), directed against the GD2 disialoganglioside, improved 5-year event-free survival in high-risk patients when administered in postconsolidation therapy and was recently implemented in standard therapy. Relapse still occurred in 57% of these patients, necessitating new therapeutic options. Bispecific trifunctional antibodies (trAbs) are IgG-like molecules directed against T cells and cancer surface antigens, redirecting T cells (via their CD3 specificity) and accessory immune cells (via their functioning Fc-fragment) toward tumor cells. We sought proof-of-concept for GD2/CD3-directed trAb efficacy against neuroblastoma.Methods We used two GD2-specific trAbs differing only in their CD3-binding specificity: EKTOMUN (GD2/human CD3) and SUREK (GD2/mouse Cd3). This allowed trAb evaluation in human and murine experimental settings. Tumor-blind trAb and the ch14.18 antibody were used as controls. A coculture model of human peripheral blood mononuclear cells (PBMCs) and neuroblastoma cell lines was established to evaluate trAb antitumor efficacy by assessing expression of T-cell surface markers for activation, proinflammatory cytokine release and cytotoxicity assays. Characteristics of tumor-infiltrating T cells and response of neuroblastoma metastases to SUREK treatment were investigated in a syngeneic immunocompetent neuroblastoma mouse model mimicking minimal residual disease.Results We show that EKTOMUN treatment caused effector cell activation and release of proinflammatory cytokines in coculture with neuroblastoma cell lines. Furthermore, EKTOMUN mediated GD2-dependent cytotoxic effects in human neuroblastoma cell lines in coculture with PBMCs, irrespective of the level of target antigen expression. This effect was dependent on the presence of accessory immune cells. Treatment with SUREK reduced the intratumor Cd4/Cd8 ratio and activated tumor infiltrating T cells in vivo. In a minimal residual disease model for neuroblastoma, we demonstrated that single-agent treatment with SUREK strongly reduced or eliminated neuroblastoma metastases in vivo. SUREK as well as EKTOMUN demonstrated superior tumor control compared with the anti-GD2 antibody, ch14.18.Conclusions Here we provide proof-of-concept for EKTOMUN preclinical efficacy against neuroblastoma, presenting this bispecific trAb as a promising new agent to fight neuroblastoma.