Abstract Epigenetic modulators are increasing in prominence as potential cancer therapies. These drugs achieve their effectiveness by inducing transcriptional changes that can inhibit cancer progression. Here, we focus on a potent and selective covalent small molecule inhibitor of LSD1 (RO7051750/ORY-1001), a lysine-specific histone demethylase enzyme. LSD1 inhibition causes epigenetic reprogramming of cancer cells by downregulating pro-proliferative neuroendocrine genes, inducing a pro-differentiation, cytostatic effect. Furthermore, the durability of this response is enhanced by covalent drug binding and the resulting dependence on target turnover. To explore the impact of dose and scheduling on the pharmacodynamic profiles underlying treatment efficacy, we developed a predictive model of drug effect, which incorporates experimentally-derived pharmacokinetic and pharmacodynamic data. This mechanistic mathematical model was calibrated based on in vitro cell-type specific kinetic data, which possesses high dimensionality across time points and dose ranges. The model describes drug-induced target engagement of LSD1, which acts as a transcriptional regulator of neuroendocrine gene, GRP. GRP levels drive reversible epigenetic “switching” between a proliferating (denoted P) and a quiescent (denoted Q) cell population (low GRP: P to Q; high GRP: Q to P), capturing the cytostatic effects of the drug. The model was trained with in vitro measurements of target engagement, GRP mRNA levels, and cell growth inhibition (including drug-free cell growth) across various time points and dosages from a small cell lung cancer cell line (NCI-H510A). Importantly, GRP mRNA and cell viability dynamics were also captured upon drug withdrawal, which enables the modeling of drug effect durability. The resulting model quantitatively describes the relationship between drug dose, target engagement, biomarker and cell growth dynamics in vitro. To scale to the in vivo setting, we integrated the in vitro-trained pharmacodynamic model with a pharmacokinetic model describing drug distribution and clearance in mouse. This hybrid model accurately predicted in vivo tumor growth dynamics in the NCI-H510A mouse xenograft across a range of doses and schedules. We used the model to explore the target engagement and biomarker profiles underlying treatment efficacy, both in vitro and in vivo. In conclusion, by training our model on key cell-type specific in vitro kinetic data we were able to accurately predict anti-tumor growth effects in vivo. This quantitative modeling framework highlights the biological profiles underlying efficacy for this latest generation of target-selective epigenetic modifying drugs. Citation Format: Mehdi Bouhaddou, Li J. Yu, Serena Lunardi, Spyros K. Stamatelos, Fiona Mack, James M. Gallo, Marc R. Birtwistle, Antje-Christine Walz. Predicting in vivo efficacy from in vitro data: Quantitative systems pharmacology modeling for an epigenetic modifier drug in cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 2796.
Read full abstract