Factors influencing the Central Nervous System (CNS) distribution of the ATR inhibitor elimusertib (BAY1895344): Implications for the treatment of CNS tumors.

Abstract

Glioblastoma (GBM) is a disease of the whole brain, with infiltrative tumor cells protected by an intact BBB. GBM has a poor prognosis despite aggressive treatment, in part due to lack of adequate drug permeability at the BBB. Standard of care GBM therapies include radiation and cytotoxic chemotherapy that lead to DNA damage. Subsequent activation of DNA damage response (DDR) pathways can induce resistance. Various DDR inhibitors, targeting the key regulators of these pathways such as ataxia telangiectasia mutated and Rad3-related (ATR), are being explored as radio- and chemo-sensitizers. Elimusertib, a novel ATR kinase inhibitor, can prevent repair of damaged DNA, increasing efficacy of DNA damaging cytotoxic therapies. Robust synergy was observed in vitro when elimusertib was combined with the DNA-damaging agent, temozolomide, however, we did not observe improvement with this combination in in vivo efficacy studies in GBM orthotopic tumor-bearing mice. This in vitro - in vivo disconnect was explored to understand factors influencing CNS distribution of elimusertib and reasons for lack of efficacy. We observed that elimusertib is rapidly cleared from systemic circulation in mice and would not maintain adequate exposure in the CNS for efficacious combination therapy with temozolomide. CNS distribution of elimusertib is partially limited by P-gp efflux at the BBB, and high binding to CNS tissues leads to low levels of pharmacologically active (unbound) drug in the brain. Acknowledging the potential for inter-species differences in pharmacokinetics, these data suggest that clinical translation of elimusertib in combination with temozolomide for treatment of GBM may be limited. Significance Statement This study examined the disconnect between the in vitro synergy and in vivo efficacy of elimusertib/temozolomide combination therapy by exploring systemic and CNS distributional pharmacokinetics. Results indicate that the lack of improvement in in vivo efficacy in GBM PDX models could be attributed to inadequate exposure of pharmacologically active drug concentrations in the CNS. These observations can guide further exploration of elimusertib for the treatment of GBM, or other CNS tumors.