Abstract 4868: A preclinical PK/PD model based on a mouse glioblastoma survival model for AZD1390, a novel, brain-penetrant ATM kinase inhibitor, to predict the inhibition of DNA damage response induced by radiation and the human efficacious dose

Abstract

Abstract AZD1390 is a novel, highly selective, brain-penetrant, potent inhibitor of the Ataxia Telangiectasia Mutated (ATM) kinase. ATM is activated in response to double-strand breaks (DSBs) and coordinates cellular responses to ionising radiation and other insults. Radiation is the mainstay of treatment for patients with brain tumors, and glioma cells are exquisitely sensitive to ATM inhibition. Therefore, AZD1390 is in clinical development in combination with radiation therapy for the treatment of patients with Glioblastoma Multiforme (GBM) and brain metastases from solid tumours (NCT03423628). [1] The aim of this work was to develop a translational PK/PD-efficacy model for AZD1390 that would enable the project team to assess the extent/duration of inhibition of target engagement required and help to predict the optimal dose and regimen of AZD1390 to treat patients with brain malignancies in combination with radiation. Time-course data of pharmacokinetics (PK), pharmacodynamics (PD) and tumor growth inhibition from mouse NCI-H2228 brain tumor orthotopic model studies were used to build a PK-PD/Efficacy or survival model. ATM activation is induced by IR treatment, becoming transiently phosphorylated within minutes of IR exposure and dissipates over a 24-hour period. This model quantitatively and dynamically integrates AZD1390 brain PK to the rate and extent of inhibition of phosphorylation of ATM, a proximal PD marker of ATM kinase activation caused by radiation induced DNA damage in the tumor and rate of induction of cell death (determined by bioluminescence signalling as a measure of tumor growth). In parallel, we also modelled cell cycle profiles in GBM cell lines to complement the in-vivo modelling work. The pATM time-course relative to IR radiation (2 Gy) was modelled using AZD1390 concentrations in the brain tumor and the corresponding pATM inhibition, which was then linked to efficacy by estimating tumor cell kill rate via exponential tumor growth model. The free brain concentration of AZD1390 that resulted in half-maximal inhibition (EC50) of pATM after radiation was estimated to be 0.8 nM (range 0.4 to 1.6 nM). Significant tumor regression in orthotopic tumor model was observed at doses >5 mg/kg. An average pATM inhibition over 24h in the mouse GBM survival model was in the range of 44% (5 mg/kg QD) to 88% (20 mg/kg BD). The % overall survival at 5 mg/kg QD and 20 mg/kg BD dose were 56% and 100%, respectively. Conclusion: The translation of mouse survival data to clinical schedules is unknown, thus efficacious dose was anchored to pATM inhibition required to deliver significant tumor regression.