INTRACELLULAR UNBOUND CONCENTRATIONS OF ATORVASTATIN AND BOSENTAN: PREDICTION USING MODELING AND SIMULATION, AND EFFECT OF METABOLISM AND TRANSPORT

Abstract

Accurate prediction of target activity of a drug and rational design of dosing regimen requires knowledge of concentration-time course of the drug at the target. In vitro in vivo correlation (IVIVC) successfully predicts activity and pharmacokinetics of some drugs but is unsuccessful with many others due to poor permeability, transporter activity and use of plasma drug concentrations for determination of PK parameters. According to the free drug hypothesis, at steady state, the unbound drug concentration on either side of a biomembrane is equal. In this case, unbound plasma drug concentration acts as a good surrogate for unbound cell concentrations. However, presence of transporters coupled with poor membrane permeability result in violation of the free drug hypothesis. This results in failure of IVIVC and subsequent discrepancies in the prediction of target activity of pharmacokinetic predictions. Since it is the unbound drug that is capable of exerting the pharmacodynamic effect and available for intracellular metabolizing and transport machinery, knowledge of the unbound concentration inside the cell is very important. Experimental measurement of intracellular unbound concentration is very difficult due to the small size of the cell and complex cellular disposition resulting from activity of metabolizing enzymes, transporters, target binding and organelle binding within the cell. The present study, therefore, aims at predicting the intracellular unbound concentrations using modeling and simulation approach. Liver perfusion experiments were conducted in male Sprague Dawley rats with uptake transporter substrates atorvastatin and bosentan, in presence and absence of inhibitors of active uptake and metabolism, to study tissue distribution of these drugs in presence of uptake transport and metabolism. The outflow perfusate data thus obtained were used as input for the explicit membrane model for liver to predict the unbound intracellular concentrations of atorvastatin and bosentan. Similarly, in vivo pharmacokinetic experiments were also conducted in rats in presence and absence of inhibitors of active uptake and metabolism. The data obtained were used as input for hybrid compartmental models to predict unbound concentrations of these drugs upon intravenous dosing. Modeling exercises were also conducted to study the differential impact of inhibition of active uptake on plasma versus unbound cell concentrations. The effect of uptake transport on the induction potential of bosentan was studied in sandwich cultured rat hepatocytes and in in vivo studies in rats. Inhibition of active uptake in the liver perfusion studies increased the outflow perfusate concentrations, decreased the amount recovered in the bile for atorvastatin and bosentan, and decreased the liver concentrations for atorvastatin. The liver concentrations for bosentan with inhibition of active uptake were not different than the control group. Inhibition of active uptake in the in vivo studies also decreased the systemic clearance of atorvastatin and bosentan. Inhibition of metabolism decreased the systemic clearance of bosentan. It was observed that the perpetrators for metabolism and transport used for this project were not specific for the pathway of interest. Active uptake appeared to be of major significance for disposition of atorvastatin. The model predicted unbound concentrations of atorvastatin at the end of 50 min perfusion were about 7-fold higher in presence of active uptake than in absence of active uptake. On the other hand, inhibition of metabolism resulted in 1.26 fold increase in unbound atorvastatin concentrations inside the cell. Modeling the in vivo data indicated that atorvastatin disposition was not affected until 90% inhibition of active uptake clearance was achieved. However, any further inhibition of active uptake clearance had a largely increased the exposure of this drug. The predicted unbound intracellular bosentan concentrations in presence of active uptake were only marginally higher than in the absence of active uptake, possibly due to inhibition of apical efflux of this drug by the uptake inhibitor, rifampin, used in this study. The modeling exercise showed that in the in vivo studies, BOS disposition was sensitive to intrinsic uptake clearance until 99% inhibition was achieved. However, any further inhibition resulted in minimal change in the exposure of this drug. The differential sensitivity of atorvastatin and bosentan exposure for active uptake clearance was thought to be due to the different diffusional clearance for these drugs. For both atorvastatin and bosentan, simulations indicated that any extent of inhibition of the active uptake clearance did not affect the cell exposure of these drugs. In vitro induction of bosentan could not be characterized in sandwich cultured rat hepatocytes. Bosentan appeared to be a weak inducer of cyp3a mediated metabolism in rats. In summary, the impact of uptake transport and metabolism on the systemic and intracellular disposition of atorvastatin and bosentan was studied. Liver perfusion and in vivo pharmacokinetic studies along with explicit membrane models were successfully used to predict unbound cell concentrations of atorvastatin and bosentan.