Abstract 4263: A semi-mechanistic pharmacokinetic-pharmacodynamic (PK-PD) model of savolitinib (AZD6094/HMPL-504), a novel MET inhibitor, to explore extent and duration of target inhibition required for optimal efficacy in the EBC-1 mouse xenograft model

MET is a transmembrane tyrosine kinase receptor that is deregulated across multiple cancer types. Savolitinib is a selective small molecule inhibitor of MET being co-developed with Hutchison MediPharma for papillary renal cell carcinoma (PRCC) and non-small cell lung cancer (NSCLC). In preclinical xenograft models, savolitinib demonstrates rapid and extensive inhibition of phosphorylated-MET (pMET) with an EC 50 of 0.35 ng/ml (CI 95% - 0.25 to 0.5), and anti-tumor activity in MET amplified models (RCC-43b & RCC-47 for PRCC; MKN-45, SNU-5 & Hs746T for gastric; EBC-1 for NSCLC). To evaluate the anti-tumor activity of savolitinib under a range of schedules, the EBC-1 xenografts were treated with savolitinib at 30 mg/kg daily (76% tumor growth inhibition (TGI)), every other day (43% TGI), or 4 on/3 off (39% TGI). Intermittent dosing was also explored at 100 mg/kg every other day (80% TGI), 4 on/3 off (67% TGI) and 2 on/5 off (46% TGI). To assess the effect of prolonged pMET inhibition, twice daily dosing at 15 mg/kg (83 % TGI) and 30 mg/kg (94 % TGI) and co-dosing savolitinib at 15 mg/kg (61 % regression) with the cytochrome P450 inhibitor, 1-aminobenzotriazole (ABT) were also tested. A population PK-pMET-TGI model was developed where savolitinib plasma concentration drives pMET inhibition, which in turn drives inhibition of tumor growth. The modelling objectives were to (1) determine whether a generalizable model can be applied to EBC-1, and data from other cell-lines investigated and (2) Determine the extent and duration of pMET inhibition that delivers optimal efficacy. An E max model of pMET inhibition driving TGI was applied and an E max was estimated as a factor of intrinsic tumor growth rate, where a value > 1 results in a shrinking tumor. It was found that a single E max estimate of 3.2 (CI 95% - 2.8 to 3.6) could be applied to EBC-1 and models previously investigated. The pMET inhibition needed for 50% of maximum effect was estimated to be > 90 % confirming that tumor regression is achieved through continuous high levels of pMET inhibition. Estimating drug effect relative to tumor growth rate offers a novel way in which to apply a mathematical model with parameters of drug effect that are shared across different xenograft models. The developed model offers the potential to be translated to predict the expected tumor growth inhibition in humans under varying dose regimens by accounting for mouse-to-man differences in PK and tumor growth dynamics. Citation Format: Rhys D. Jones, Michael Grondine, Alexandra Borodovsky, Maryann San Martin, Michelle DuPont, Celina D9Cruz, Alwin Schuller, Ryan Henry, Evan Barry, Klas Petersson, Tarjinder Sahota, Ghada F. Ahmed. A semi-mechanistic pharmacokinetic-pharmacodynamic (PK-PD) model of savolitinib (AZD6094/HMPL-504), a novel MET inhibitor, to explore extent and duration of target inhibition required for optimal efficacy in the EBC-1 mouse xenograft model [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 4263.