Abstract 3362: Miniaturized PBPK model improves pharmacodynamic characterization and physiological interpretability for compounds with profound hysteresis in tumor.

Proceedings: AACR 104th Annual Meeting 2013; Apr 6-10, 2013; Washington, DC Background: Significant hysteresis between plasma concentration and target inhibition at the effect site (e.g., tumor) is a frequent observation, commonly described mathematically by connecting the central (i.e., plasma) compartment to an ‘effect compartment’ by a ‘link’ which causes the concentration in the latter to be delayed relative to the plasma. The result is a direct response between effect-compartment concentration and target inhibition. A significant drawback is that the effect compartment cannot be observed (making it impossible to validate) and has no physiological interpretation (rendering communication with other disciplines difficult). We develop a novel approach that is more physiologically meaningful, provides more-precise model parameter estimates, and gives insight into the physico-chemical factors limiting distribution into the tumor. Method: We orally administered single doses of several compounds (including Crizotinib, AZD3463, and others) targeting ALK to mice bearing tumors derived from the DEL and H3122 non-small-cell lung cancer line, at several dose levels. At 6, 24, and 48 hours post-dose, we measured the plasma and tumor concentrations of each compound and associated target inhibition (phosphorylated ALK, pALK) in the tumor. pALK inhibition shows a direct response not to plasma, but to tumor concentration, indicating that the delay is distributional in nature. We constructed a miniature physiologically-based pharmacokinetic (mPBPK) model consisting of a central compartment and a tumor of fixed physiological volume. pALK inhibition was modeled as a direct Emax response to tumor concentration. For each compound, we simultaneously fitted the mPBPK model to the naive-pooled plasma and tumor concentrations, as well as pALK, using all available dose levels. Beyond the standard PK and PD parameters (Emax, E, IC50) we also fitted the tumor partition constant Kp, and tumor blood flow rate Qt. For comparison, we fitted a standard effect-compartment (‘link’) model to the plasma concentrations and pALK levels to the same data. Results: For each compound, we computed unbound EC50 for both effect-compartment and mPBPK models. We found that while the point estimates largely agree, the mPBPK model delivers more-precise estimates (typically 50% lower CV%). We attribute this to its use of additional data (tumor concentration) to constrain the model, which more than compensates for the additional parameters in the mPBPK model. We find that there is broad consistency in estimates of tumor flow rate Qt across the compounds studied, indicating that distribution from plasma to site of action is limited by blood flow, rather than by permeability. Additionally, we found that the greater physiological interpretability of the mPBPK model enhances cross-functional communication within project teams. Citation Format: Francis D. Gibbons, Dan Widzowski, Minhui Shen, Jane Cheng, Lisa Drew, Jamal C. Saeh, Douglas Ferguson. Miniaturized PBPK model improves pharmacodynamic characterization and physiological interpretability for compounds with profound hysteresis in tumor. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 3362. doi:10.1158/1538-7445.AM2013-3362