A significant challenge in the development of drugs to treat central nervous system (CNS) disorders is to attain sufficient delivery of antibodies across blood-brain barriers (BBB). Since not all antibodies can pass through BBB, it is crucial to understand antibody exposure in the CNS quantitatively to construct drug characteristics and identify proper dosing regimens. We focus on a minimal physiologically-based pharmacokinetic (mPBPK) model of the brain for antibody therapeutics. This model is the reduced form of an existing multi-species platform brain PBPK model. The original PBPK model consists of 100  differential equations while the mPBPK model contains 16 differential equations which improves the speed of simulations. The model includes thirty-six parameters and their values are obtained from the original brain PBPK model. In this presentation, we will discuss the use of a sensitivity-based parameter subset selection algorithm to determine those parameters which are identifiable in the sense that they can be uniquely determined by data. We illustrate this for ascending human doses. Issues to be discussed include the computation of sensitivities using sensitivity equations and complex-step approximations. We will also discuss qualitative verification techniques as well as quantitative techniques based on energy statistics.

Zoom link: https://ncsu.zoom.us/j/98139608681?pwd=Z3RnRGRYL3JCRnlNT0NFdmZLRzRtdz09