|dc.description.abstract||In the latest years, multiple types of computational models have been used extensively in drug development, with a massive growth of uses of physiologically-based pharmacokinetic (PBPK) modeling in fields associated with drugs and natural chemicals. PBPK models are used for systemic and tissue exposure. Combined with the pharmacodynamic models (PD), PBPK models can predict drug-effect over time in many disease states and populations. To attain a mechanistic description of the effect of a drug in biological systems, PBPK models correlated drug-specific data with the biology and physiology at the organism level, permitting a deductive simulation of drug concentration–time profiles. In this dissertation, we studied the effect of transporters and metabolism on the pharmacokinetics (PK) and PD of 2 drugs: caffeine (specifically in pregnant populations), and granisetron using PBPK modeling.
About 80% of pregnant women consume caffeine orally on a daily basis. Many reports indicated that consumption of >200 mg caffeine during pregnancy could increase the likelihood of miscarriage. Thus, we developed and validated a pregnancy PBPK/PD model for caffeine to examine the association between maternal caffeine consumption during pregnancy, and caffeine plasma levels at doses between 70 mg and 300 mg. The developed model was used to predict changes in caffeine concentrations across the 3 trimesters, and to predict associated changes in caffeine PD parameters. The model successfully predicted the effect of decreased cytochrome P450 (CYP1A2) activity on caffeine plasma levels and predicted the increased levels of caffeine in the fetoplacental compartment (FPC). Increased caffeine levels in maternal blood were accompanied by greater inhibition of phosphodiesterase enzyme, higher cyclic adenosine monophosphate, and a greater increase of epinephrine levels, which could increase the risk of pregnancy loss. The application of the developed PBPK model to predict PD effect could provide a useful tool to help define potential cut-offs for caffeine intake in various stages of pregnancy. Our future directions for this project are to use this project to correlate the amount of caffeine intake during pregnancy to the percent of miscarriage on pregnant subject.
Chemotherapy-induced nausea and vomiting (CINV) is one of the most devastating side effects that affect a patient’s quality of life. Granisetron is effective in many cases; however, about 20-30% of patients remain to show unsatisfactory responses, which could be due to the development of drug resistance caused by anti-cancer drugs. The purpose of this project is to explain the observed variability in granisetron efficacy. We started by identifying the effect of P-glycoprotein (P-gp) and lysosomal entrapment on granisetron permeability and plasma profiles following oral dosing. Our predicted results, assessed by in vitro experiments, demonstrated that changes in P-gp function, as well as lysosomal pH, alters granisetron permeability and plasma concentrations. Granisetron is cleared mainly by hepatic metabolism, with less than 20% excreted unchanged in the urine. Granisetron is metabolized primarily by CYP 1A1 and CYP 3A5. Thus, in the second part of the study, we developed a PKPD model to validate and predict the effect of genetic variations in CYP1A1 and CYP3A5 on granisetron levels in plasma and brain and predict the effect of these genetic variations on the occupancy of 5-hydroxytryptamine (5-HT3) receptors. Our results showed that granisetron is a P-gp substrate and is usually effluxed out of the cells, thus limiting its absorption. Also, due to its physicochemical properties, granisetron gets entrapped inside the lysosomes, limiting its passage through cell lines. Furthermore, genetic polymorphism plays an important role in receptor occupancy, where we concluded that subjects with CYP1A1 single nucleotide polymorphism (SNP) (and extensive metabolizer) would lower plasma and brain levels thus decreasing receptor occupancy, while subjects with the CYP3A5 SNP (a poor metabolizer), would have an increase in granisetron levels, thus increasing receptor occupancy for better CINV control.
In conclusion, PBPK models can be invaluable support in drug development. Predicting potential drug effects in several populations is some of the major topics where PBPK approaches have presented significant advances in recent years. Our future directions will be to use PBPK modeling to build models and predict drug effect and doses in several other important populations like the pediatric population, where drug dosing is problematic. Another important direction is the utilization of PBPK modeling in simulating several drug-drug interactions (DDI), especially in geriatric population where most of the patients are on polypharmacy and the possibility of DDI is high.||en_US