Background NM32-2668 is a fragment-based multispecific antibody therapeutic¹ that has been designed to activate T-cells (via CD3) in the presence of tumour antigen receptor tyrosine kinase-like orphan receptor 1 (ROR1). The objective of this work was to build a mathematical model to establish a PKPD relationship using both in vitro and in vivo data for NM32-2668.

Methods Data were collected from in vitro studies measuring CD4/8 activation and cytotoxicity from a panel of cell lines and patient samples to increasing concentrations of NM32-2668, and from in vivo tumour growth inhibition (TGI) data from a humanised mouse model in one cell line with two different donors with increasing doses of NM32-2668. Nonlinear mixed-effects models were used to assess the variability in cytotoxicity as a function of drug concentration and immune system activation. The translatability of in vitro potency values for immune system activation was assessed by linking PK to in vitro data and to TGI in vivo data.

Results The combination of ROR1 expression and CD8 activation fully explained the variance in cytotoxicity across all in vitro data. The estimated in vitro potency for CD8 activation could successfully be used to provide a link between PK and TGI in vivo.

Conclusions A PK-PD-Efficacy model based on the in vitro data was established showing that the cytotoxicity response was strongly correlated to ROR1 expression and CD8 activation. Building on this in vitro model, we developed an in vivo PK-TGI model that can link immune system activation to TGI.

Reference

1. Egan TJ, Diem D, Weldon R, Neumann Y, Meyer S, Urech DM. Novel multispecific heterodimeric antibody format allowing modular assembly of variable domain fragments. MABS. 2017;9(1):68–84