UC Santa Barbara

Pharmacokinetic models, i.e., the models describing distribution and elimination kinetics of drugs within the body, have become an integral part of any drug discovery and development process. Unfortunately, the difficulty of measuring drug concentration levels in-vivo has restricted the pharmacokinetic modeling field to working with sparse datasets. The recent development of continuous molecular monitoring platforms such as electrochemical aptamer-based (EAB) sensors presents an unprecedented opportunity to overcome this obstacle by providing fast in-vivo drug concentration measurements.

In this work, I present our work on how such high-temporal resolution datasets can help us specify individualized pharmacokinetic models and the implications of having such high-precision models. We first developed a system identification-based approach to test physiological assumptions about drug transport kinetics that are almost universally accepted but rarely tested. Because the precision of the identification process is an important factor in testing these assumptions, we further developed a way to design experiments to maximize the information content of the measured data to improve the precision of model identification.

The availability of individualized compartmental pharmacokinetic models also offers a chance to improve clinical practices. In particular, we first discuss how such models can improve the sample collection process by developing a way to estimate plasma concentration levels based on easier-to-measure subcutaneous interstitial fluid (ISF) measurements. These individualized models can also help in the treatment of a patient by enabling feedback-controlled drug delivery. Particularly, we design an adaptive feedback control mechanism to adjust the drug intake rate to keep the measured drug concentration levels at a targeted value to improve the therapeutic impact while avoiding the toxicity of the drug.