Beyond the basics: A deep dive into parameter estimation for advanced PBPK and QSP models.

Physiologically-based pharmacokinetic (PBPK) models and quantitative systems pharmacology (QSP) models have contributed to drug development strategies. The parameters of these models are commonly estimated by capturing observed values using the nonlinear least-squares method. Software packages for PBPK and QSP modeling provide a range of parameter estimation algorithms. To choose the most appropriate method, modelers need to understand the basic concept of each approach. This review provides a general introduction to the key points of parameter estimation with a focus on the PBPK and QSP models, and the respective parameter estimation algorithms. The latter part assesses the performance of five parameter estimation algorithms - the quasi-Newton method, Nelder-Mead method, genetic algorithm, particle swarm optimization, and Cluster Gauss-Newton method - using three examples of PBPK and QSP modeling. The assessment revealed that some parameter estimation results were significantly influenced by the initial values. Moreover, the choice of algorithms demonstrating good estimation results heavily depends on factors such as model structure and the parameters to be estimated. To obtain credible parameter estimation results, it is advisable to conduct multiple rounds of parameter estimation under different conditions, employing various estimation algorithms.

Practical QSP application from the preclinical phase to enhance the probability of clinical success: Insights from case studies in oncology.

Quantitative Systems Pharmacology (QSP) has emerged as a promising modeling and simulation (M&S) approach in drug development, with potential to improve clinical success rates. While conventional M&S has significantly contributed to quantitative understanding in late preclinical and clinical phases, it falls short in explaining unexpected phenomena and testing hypotheses in the early research phase. QSP presents a solution to these limitations. To harness the full potential of QSP in early preclinical stages, preclinical modelers who are familiar with conventional M&S need to update their understanding of the differences between conventional M&S and QSP. This review focuses on QSP applications during the preclinical stage, citing case examples and sharing our experiences in oncology. We emphasize the critical role of QSP in increasing the probability of success for clinical proof of concept (PoC) when applied from the early preclinical stage. Enhancing the quality of both hypotheses and QSP models from early preclinical stage is of critical importance. Once a QSP model achieves credibility, it facilitates predictions of clinical responses and potential biomarkers. We propose that sequential QSP applications from preclinical stages can improve success rates of clinical PoC, and emphasize the importance of refining both hypotheses and QSP models throughout the process.

Physiologically-based pharmacokinetic modeling for investigating the effect of simeprevir on concomitant drugs and an endogenous biomarker of OATP1B.

The orally available anti-hepatitis C virus (HCV) drug simeprevir exhibits nonlinear pharmacokinetics at the clinical doses due to saturation of cytochrome P450 (CYP) 3A4 metabolism and organic anion transporting peptide (OATP) 1B mediated hepatic uptake. Additionally, simeprevir increases exposures of concomitant drugs by CYP3A4 and OATP1B inhibition. The objective of this study was to develop physiologically-based pharmacokinetic (PBPK) models that could describe drug-drug interactions (DDIs) of simeprevir with concomitant drugs via CYP3A4 and OATP1B inhibition, and also to capture the effects on coproporphyrin-I (CP-I), an endogenous biomarker of OATP1B. PBPK modeling estimated unbound simeprevir inhibitory constant (Ki ) of 2.89 µM against CYP3A4 in the DDI results between simeprevir and midazolam in healthy volunteers. Then, we analyzed the DDIs between simeprevir and atorvastatin, a dual substrate of CYP3A4 and OATP1B, in healthy volunteers, and unbound Ki against OATP1B was estimated to be 0.00347 µM. Finally, we analyzed the increase in the blood level of CP-I by simeprevir to verify the Ki,OATP1B . Because CP-I was measured in subjects with HCV with various hepatic fibrosis state, Monte Carlo simulation was performed to involve the decreases in expression levels of hepatic CYP3A4 and OATP1B and their interindividual variabilities. The PBPK modeling coupled with Monte Carlo simulation using the Ki,OATP1B value obtained from atorvastatin study reasonably recovered the observed relationship between CP-I and simeprevir blood levels. In conclusion, the simeprevir PBPK model developed in this study can quantitatively describe the increase in exposures of concomitant drugs and an endogenous biomarker via inhibition of CYP3A4 and OATP1B.

Revisiting Nonlinear Bosentan Pharmacokinetics by Physiologically Based Pharmacokinetic Modeling: Target Binding, Albeit Not a Major Contributor to Nonlinearity, Can Offer Prediction of Target Occupancy.

Bosentan is a high-affinity antagonist of endothelin receptors and one of the earliest examples for target-mediated drug disposition [a type of nonlinear pharmacokinetics (PKs) caused by saturable target binding]. The previous physiologically based PK (PBPK) modeling indicated that the nonlinear PKs of bosentan was explainable by considering saturable hepatic uptake. However, it remained unexamined to what extent the saturable target binding contributes to the nonlinear PKs of bosentan. Here, we developed a PBPK model incorporating saturable target binding and hepatic uptake and analyzed the clinical bosentan PK data using the cluster Gauss-Newton method (CGNM). The PBPK model without target binding fell short in capturing the bosentan concentrations below 100 nM, based on the PK profiles and the goodness-of-fit plot. Both global and local identifiability analyses (using the CGNM and Fisher information matrix, respectively) informed that the target binding parameters were identifiable only if the observations from the lowest dose (10 mg) were included. By analyzing blood PK profiles alone, the PBPK model with target binding yielded practically identifiable target binding parameters and predicted the maximum target occupancies of 0.6-0.8 at clinical bosentan doses. Our results indicate that target binding, albeit not a major contributor to the nonlinear bosentan PKs, may offer a prediction of target occupancy from blood PK profiles alone and potential guidance on achieving optimal efficacy outcomes, under the condition when the high-affinity drug target is responsible for the efficacy of interest and when the dose ranges cover varying degrees of target binding. SIGNIFICANCE STATEMENT: By incorporating saturable target binding, our physiologically based pharmacokinetic (PBPK) model predicted in vivo target occupancy of bosentan based only on the blood concentration-time profiles obtained from a wide range of doses. Our analysis highlights the potential utility of PBPK models that incorporate target binding in predicting target occupancy in vivo.

A Simple Decision Tree Suited for Identification of Early Oral Drug Candidates With Likely Pharmacokinetic Nonlinearity by Intestinal CYP3A Saturation.

To identify oral drugs that likely display nonlinear pharmacokinetics due to saturable metabolism by intestinal CYP3A, our previous report using CYP3A substrate drugs proposed an approach using thresholds for the linear index number (LIN3A = dose/Km; Km, Michaelis-Menten constant for CYP3A) and the intestinal availability (FaFg). Here, we aimed to extend the validity of the previous approach using both CYP3A substrate and non-substrate drugs and to devise a decision tree suited for early drug candidates using in vitro metabolic intrinsic clearance (CLint, vitro) instead of FaFg. Out of 152 oral drugs (including 136 drugs approved in Japan, US or both), type I nonlinearity (in which systemic drug exposure increases in a more than dose-proportional manner) was noted with 82 drugs (54%), among which 58 drugs were identified as CYP3A substrates based on public information. Based on practical feasibility, 41 drugs were selected from CYP3A substrates and subjected to in-house metabolic assessment. The results were used to determine the thresholds for CLint, vitro (0.45 µL/min/pmol CYP3A4) and LIN3A (1.0 L). For four drugs incorrectly predicted, potential mechanisms were looked up. Overall, our proposed decision tree may aid in the identification of early drug candidates with intestinal CYP3A-derived type I nonlinearity.

Improved Prediction of the Drug-Drug Interactions of Pemafibrate Caused by Cyclosporine A and Rifampicin via PBPK Modeling: Consideration of the Albumin-Mediated Hepatic Uptake of Pemafibrate and Inhibition Constants With Preincubation Against OATP1B.

Pemafibrate (PMF) is highly albumin-bound (>99.8%) and a substrate for hepatic uptake transporters (OATP1B) and CYP enzymes. Here, we developed a PBPK model of PMF to capture drug-drug interactions (DDI) incurred by cyclosporine (CsA) and rifampicin (RIF), the two OATP1B inhibitors. Initial PBPK modeling of PMF utilized in vitro hepatic uptake clearance (PSinf) obtained in the absence of albumin, but failed in capturing the blood PMF pharmacokinetic (PK) profiles. Based on the results that in vitro PSinf of unbound PMF was enhanced in the presence of albumin, we applied the albumin-facilitated dissociation model and the resulting PSinf parameters improved the prediction of the blood PMF PK profiles. In refining our PBPK model toward improved prediction of the observed DDI data (PMF co-administered with single dosing of CsA or RIF; PMF following multiple RIF dosing), we adjusted the previously obtained in vivo OATP1B inhibition constants (Ki,OATP1B) of CsA or RIF for pitavastatin by correcting for substrate-dependency. We also incorporated the induction of OATP1B and CYP enzymes after multiple RIF dosing. Sensitivity analysis informed that the higher gastrointestinal absorption rate constant could further improve capturing the observed DDI data, suggesting the possible inhibition of intestinal ABC transporter(s) by CsA or RIF.

Evaluation of Hepatic Uptake of OATP1B Substrates by Short Term-Cultured Plated Human Hepatocytes: Comparison With Isolated Suspended Hepatocytes.

Hepatic uptake clearance has been measured in suspended human hepatocytes (SHH). Plated human hepatocytes (PHH) after short-term culturing are increasingly employed to study hepatic transport driven mainly by its higher throughput. To know pros/cons of both systems, the hepatic uptake clearances of several organic anion transporting polypeptide 1B substrates were compared between PHH and SHH by determining the initial uptake velocities or through dynamic model-based analyses. For cerivastatin, pitavastatin and rosuvastatin, initial uptake clearances (PSinf) obtained using PHH were comparable to those using SHH, while cell-to-medium concentration (C/M) ratios were 2.7- to 5.4-fold higher. For pravastatin and dehydropravastatin, hydrophilic compounds with low uptake/cellular binding, their PSinf and C/M ratio in PHH were 1.8- to 3.2-fold lower than those in SHH. These hydrophilic substrates are more prone to wash-off during the uptake study using PHH, which may explain the apparently lower uptake than SHH. The C/M ratios obtained using PHH were stable over an extended time, making PHH suitable to estimate the C/M ratios and hepatocyte-to-medium unbound concentration ratios (Kp,uu). In conclusion, PHH is useful in evaluating hepatic uptake/efflux clearances and Kp,uu of OATP1B substrates in a high-throughput manner, however, a caution is warranted for hydrophilic drugs with low uptake/cellular binding.

Physiologically-Based Pharmacokinetic Modeling Analysis for Quantitative Prediction of Renal Transporter-Mediated Interactions Between Metformin and Cimetidine.

Metformin is an important antidiabetic drug and often used as a probe for drug-drug interactions (DDIs) mediated by renal transporters. Despite evidence supporting the inhibition of multidrug and toxin extrusion proteins as the likely DDI mechanism, the previously reported physiologically-based pharmacokinetic (PBPK) model required the substantial lowering of the inhibition constant values of cimetidine for multidrug and toxin extrusion proteins from those obtained in vitro to capture the clinical DDI data between metformin and cimetidine.1 We constructed new PBPK models in which the transporter-mediated uptake of metformin is driven by a constant membrane potential. Our models successfully captured the clinical DDI data using in vitro inhibition constant values and supported the inhibition of multidrug and toxin extrusion proteins by cimetidine as the DDI mechanism upon sensitivity analysis and data fitting. Our refined PBPK models may facilitate prediction approaches for DDI involving metformin using in vitro inhibition constant values.

Strategies to improve the prediction accuracy of hepatic intrinsic clearance of three antidiabetic drugs: Application of the extended clearance concept and consideration of the effect of albumin on CYP2C metabolism and OATP1B-mediated hepatic uptake.

The antidiabetic drugs glibenclamide, repaglinide, and nateglinide are well-known substrates for hepatic uptake transporters of the organic anion transporting polypeptide (OATP) family and metabolizing enzymes of the cytochrome P450 (CYP) 2C subfamily. The systemic exposure of these drugs varies substantially among individuals, impacted by genetic polymorphisms of transporters and metabolizing enzymes as well as drug-drug interactions. The use of the conventional in vitro-in vivo extrapolation (IVIVE) method was found to underestimate their hepatic intrinsic clearance (CLint,all); the clinically observed CLint,all values were ≥10-fold higher than the predicted values from in vitro data. In order to improve the accuracy in predicting CLint,all of these drugs, the following modifications were implemented; i) the extended clearance concept was applied during IVIVE processes, ii) albumin was added to metabolic assays using human liver microsomes (to minimize the impact of intrinsic inhibitors on kinetic parameters for CYP2C-mediated metabolism) and to hepatic uptake assays (to accommodate the enhanced hepatic uptake observed with albumin-bound drugs), and iii) differing rates of efflux and influx via diffusion were used. The IVIVE method with these modifications yielded the predicted CLint,all values from in vitro data in closer agreement with the CLint,all values observed in vivo; the fold differences between the predicted and observed CLint,all values reduced from 13-15 to 5.9-6.7. Our current approach offers an improvement in the prediction of CLint,all and further investigations are warranted to enhance the prediction accuracy of IVIVE.

PBPK Modeling of Coproporphyrin I as an Endogenous Biomarker for Drug Interactions Involving Inhibition of Hepatic OATP1B1 and OATP1B3.

The aim of the present study was to establish a physiologically based pharmacokinetic (PBPK) model for coproporphyrin I (CP-I), a biomarker supporting the prediction of drug-drug interactions (DDIs) involving hepatic organic anion transporting polypeptide 1B (OATP1B), using clinical DDI data with an OATP1B inhibitor rifampicin (300 and 600 mg, orally). The in vivo inhibition constants of rifampicin used as initial input parameters for OATP1Bs (Ki,u,OATP1Bs ) and multidrug resistance-associated protein two-mediated biliary excretion were estimated as 0.23 and 0.87 µM, respectively, from previous reports. Sensitivity analysis demonstrated that the Ki,u,OATP1Bs and biosynthesis rate of CP-I affected the magnitude of the interaction. Ki,u,OATP1Bs values optimized by nonlinear least-squares fitting were ~0.5-fold of the initial value. It was determined that the blood concentration-time profiles of four statins were well-predicted using corrected individual Ki,u,OATP1B values (ratio of in vitro Ki,u(statin) /in vitro Ki,u(CP-I) ). In conclusion, PBPK modeling of CP-I supports dynamic prediction of OATP1B-mediated DDIs.

In Silico Prediction of Major Clearance Pathways of Drugs among 9 Routes with Two-Step Support Vector Machines.

PURPOSE: The clearance pathways of drugs are critical elements for understanding the pharmacokinetics of drugs. We previously developed in silico systems to predict the five clearance pathway using a rectangular method and a support vector machine (SVM). In this study, we improved our classification system by increasing the number of clearance pathways available for our prediction (CYP1A2, CYP2C8, CYP2C19, and UDP-glucuronosyl transferases (UGTs)) and by accepting multiple major pathways. METHODS: Using the four default descriptors (charge, molecular weight, logD at pH 7.0, and unbound fraction in plasma), three kinds of SVM-based predictors based on traditional single-step approach or two-step focusing approaches with subset or partition clustering were developed. The two-step approach with subset clustering resulted in the highest prediction performance. The feature-selection of additional descriptors based on a greedy algorithm was employed to further improve the predictability. RESULTS: The prediction accuracy for each pathway was increased to more than 0.83 with the exception of CYP2C19 and UGTs pathways, whose accuracies were below 0.7. Prediction performance of CYP1A2, CYP3A4 and renal excretion pathways were found to be acceptable using external dataset. CONCLUSIONS: We successfully constructed a novel SVM-based predictor for the multiple major clearance pathways based on chemical structures.

Application of PBPK Modeling and Virtual Clinical Study Approaches to Predict the Outcomes of CYP2D6 Genotype-Guided Dosing of Tamoxifen.

The Tamoxifen Response by CYP2D6 Genotype-based Treatment-1 (TARGET-1) study (n = 180) was conducted from 2012-2017 in Japan to determine the efficacy of tamoxifen dosing guided by cytochrome P450 2D6 (CYP2D6) genotypes. To predict its outcomes prior to completion, we constructed the comprehensive physiologically based pharmacokinetic (PBPK) models of tamoxifen and its metabolites and performed virtual TARGET-1 studies. Our analyses indicated that the expected probability to achieve the end point (demonstrating the superior efficacy of the escalated tamoxifen dose over the standard dose in patients carrying CYP2D6 variants) was 0.469 on average. As the population size of this virtual clinical study (VCS) increased, the expected probability was substantially increased (0.674 for n = 260). Our analyses also informed that the probability to achieve the end point in the TARGET-1 study was negatively impacted by a large variability in endoxifen levels. Our current efforts demonstrate the promising utility of the PBPK modeling and VCS approaches in prospectively designing effective clinical trials.

Quantitative Analysis of Complex Drug-Drug Interactions between Cerivastatin and Metabolism/Transport Inhibitors Using Physiologically Based Pharmacokinetic Modeling.

Cerivastatin (CER) was withdrawn from the world market because of lethal rhabdomyolysis. Coadministrations of CER and cyclosporine A (CsA) or gemfibrozil (GEM) have been reported to increase the CER blood concentration. CsA is an inhibitor of organic anion transporting polypeptide (OATP)1B1 and CYP3A4, and GEM and its glucuronide (GEM-glu) inhibit OATP1B1 and CYP2C8. The purpose of this study was to describe the transporter-/enzyme-mediated drug-drug interactions (DDIs) of CER with CsA or GEM based on unified physiologically based pharmacokinetic (PBPK) models and to investigate whether the DDIs can be quantitatively analyzed by a bottom-up approach. Initially, the PBPK models for CER and GEM/GEM-glu were constructed based on the previously reported standard protocols. Next, the drug-dependent parameters were optimized by Cluster Newton Method. Thus, described concentration-time profiles for CER and GEM/GEM-glu agreed well with the clinically observed data. The DDIs were then simulated using the established PBPK models with previously obtained in vitro inhibition constants of CsA or GEM/GEM-glu against the OATP1B1 and cytochrome P450s. DDIs with the inhibitors were underestimated compared with observed data using the geometric means of reported values. To search for better described parameters within the range of in vitro values, sensitivity analyses were performed for DDIs of CER. Using the in vitro parameter sets selected by sensitivity analyses, these DDIs were well reproduced, indicating that the present PBPK models were able to describe adequately the clinical DDIs based on a bottom-up approach. The approaches in this study would be applicable to the prediction of other DDIs involving both transporters and metabolic enzymes.

A Clinical Quantitative Evaluation of Hepatobiliary Transport of [11C]Dehydropravastatin in Humans Using Positron Emission Tomography.

Various positron emission tomography (PET) probes have been developed to assess in vivo activities in humans of drug transporters, which aid in the prediction of pharmacokinetic properties of drugs and the impact of drug-drug interactions. We developed a new PET probe, sodium (3R, 5R)-3, 5-dihydroxy-7-((1S, 2S, 6S, 8S)-6-hydroxy-2-methyl-8- ((1-[11C]-(E)-2-methyl-but-2-enoyl) oxy) -1, 2, 6, 7, 8, 8a-hexahydronaphthalen-1-yl) heptanoate ([11C]DPV), and demonstrated its usefulness for the quantitative investigation of Oatps (gene symbol SLCO) and Mrp2 (gene symbol ABCC2) in rats. To further analyze the species differences and verify the pharmacokinetic parameters in humans, serial PET scanning of the abdominal region with [11C]DPV was performed in six healthy volunteers with and without an OATP1Bs and MRP2 inhibitor, rifampicin (600 mg, oral), in a crossover fashion. After intravenous injection, [11C]DPV rapidly distributed to the liver and kidney followed by secretion into the bile and urine. Rifampicin significantly reduced the liver distribution of [11C]DPV 3-fold, resulting in a 7.5-fold reduced amount of excretion into the bile and the delayed elimination of [11C]DPV from the blood circulation. The hepatic uptake clearance (CLuptake, liver) and canalicular efflux clearance (CLint, bile) of [11C]DPV (544 ± 204 and 10.2 ± 3.5 µl/min per gram liver, respectively) in humans were lower than the previously reported corresponding parameters in rats (1800 and 298 µl/min per gram liver, respectively) (Shingaki et al., 2013). Furthermore, rifampicin treatment significantly reduced CLuptake, liver and CLint, bile by 58% and 44%, respectively. These results suggest that PET imaging with [11C]DPV is an effective tool for quantitatively characterizing the OATP1Bs and MRP2 functions in the human hepatobiliary transport system.

Evaluation of Alteration in Hepatic and Intestinal BCRP Function In Vivo from ABCG2 c.421C>A Polymorphism Based on PBPK Analysis of Rosuvastatin.

Polymorphism c.421C>A in the ABCG2 gene is thought to reduce the activity of breast cancer resistance protein (BCRP), a xenobiotic transporter, although it is not clear which organ(s) contributes to the polymorphism-associated pharmacokinetic change. The aim of the present study was to estimate quantitatively the influence of c.421C>A on intestinal and hepatic BCRP activity using a physiologically based pharmacokinetic (PBPK) model of rosuvastatin developed from clinical data and several in vitro studies. Simultaneous fitting of clinical data for orally and intravenously administered rosuvastatin, obtained in human subjects without genotype information, was first performed with the PBPK model to estimate intrinsic clearance for hepatic elementary process. The fraction of BCRP activity in 421CA and 421AA (fca and faa values, respectively) with respect to that in 421CC subjects was then estimated based on extended clearance concepts and simultaneous fitting to oral administration data for the three genotypes (421CC, 421CA, and 421AA). On the assumption that c.421C>A affects both intestinal and hepatic BCRP, clinical data in each genotype were well reproduced by the model, and the estimated terminal half-life was compatible with the observed values. The assumption that c.421C>A affects only either intestinal or hepatic BCRP gave poorer agreement with observed values. The faa values obtained on the former assumption were 0.48-0.54. Thus, PBPK model analysis enabled quantitative evaluation of alteration in BCRP activity owing to c.421C>A, and BCRP activity in 421AA was estimated as half that in 421CC.

Physiologically Based Pharmacokinetic Modeling of Bosentan Identifies the Saturable Hepatic Uptake As a Major Contributor to Its Nonlinear Pharmacokinetics.

Bosentan is a substrate of hepatic uptake transporter organic anion-transporting polypeptides (OATPs), and undergoes extensive hepatic metabolism by cytochrome P450 (P450), namely, CYP3A4 and CYP2C9. Several clinical investigations have reported a nonlinear relationship between bosentan doses and its systemic exposure, which likely involves the saturation of OATP-mediated uptake, P450-mediated metabolism, or both in the liver. Yet, the underlying causes for the nonlinear bosentan pharmacokinetics are not fully delineated. To address this, we performed physiologically based pharmacokinetic (PBPK) modeling analyses for bosentan after its intravenous administration at different doses. As a bottom-up approach, PBPK modeling analyses were performed using in vitro kinetic parameters, other relevant parameters, and scaling factors. As top-down approaches, three different types of PBPK models that incorporate the saturation of hepatic uptake, metabolism, or both were compared. The prediction from the bottom-up approach (models 1 and 2) yielded blood bosentan concentration-time profiles and their systemic clearance values that were not in good agreement with the clinically observed data. From top-down approaches (models 3, 4, 5-1, and 5-2), the prediction accuracy was best only with the incorporation of the saturable hepatic uptake for bosentan. Taken together, the PBPK models for bosentan were successfully established, and the comparison of different PBPK models identified the saturation of the hepatic uptake process as a major contributing factor for the nonlinear pharmacokinetics of bosentan.

Comprehensive PBPK Model of Rifampicin for Quantitative Prediction of Complex Drug-Drug Interactions: CYP3A/2C9 Induction and OATP Inhibition Effects.

This study aimed to construct a physiologically based pharmacokinetic (PBPK) model of rifampicin that can accurately and quantitatively predict complex drug-drug interactions (DDIs) involving its saturable hepatic uptake and auto-induction. Using in silico and in vitro parameters, and reported clinical pharmacokinetic data, rifampicin PBPK model was built and relevant parameters for saturable hepatic uptake and UDP-glucuronosyltransferase (UGT) auto-induction were optimized by fitting. The parameters for cytochrome P450 (CYP) 3A and CYP2C9 induction by rifampicin were similarly optimized using clinical DDI data with midazolam and tolbutamide as probe substrates, respectively. For validation, our current PBPK model was applied to simulate complex DDIs with glibenclamide (a substrate of CYP3A/2C9 and hepatic organic anion transporting polypeptides (OATPs)). Simulated results were in quite good accordance with the observed data. Altogether, our constructed PBPK model of rifampicin demonstrates the robustness and utility in quantitatively predicting CYP3A/2C9 induction-mediated and/or OATP inhibition-mediated DDIs with victim drugs.

Comparison of Methods for Estimating Unbound Intracellular-to-Medium Concentration Ratios in Rat and Human Hepatocytes Using Statins.

It is essential to estimate concentrations of unbound drugs inside the hepatocytes to predict hepatic clearance, efficacy, and toxicity of the drugs. The present study was undertaken to compare predictability of the unbound hepatocyte-to-medium concentration ratios (Kp,uu) by two methods based on the steady-state cell-to-medium total concentration ratios at 37°C and on ice (Kp,uu,ss) and based on their initial uptake rates (Kp,uu,V0). Poorly metabolized statins were used as test drugs because of their concentrative uptake via organic anion-transporting polypeptides. Kp,uu,ss values of these statins provided less interexperimental variation than the Kp,uu,V0 values, because only data at longer time are required for Kp,uu,ss Kp,uu,V0 values for pitavastatin, rosuvastatin, and pravastatin were 1.2- to 5.1-fold Kp,uu,ss in rat hepatocytes; Kp,uu,V0 values in human hepatocytes also tended to be larger than corresponding Kp,uu,ss To explain these discrepancies, theoretical values of Kp,uu,ss and Kp,uu,V0 were compared with true Kp,uu (Kp,uu,true), considering the inside-negative membrane potential and ionization of the drugs in hepatocytes and medium. Membrane potentials were approximately -30 mV in human hepatocytes at 37°C and almost abolished on ice. Theoretical equations considering the membrane potentials indicate that Kp,uu,ss values for the statins are 0.85- to 1.2-fold Kp,uu,true, whereas Kp,uu,V0 values are 2.2- to 3.1-fold Kp,uu,true, depending on the ratio of the passive permeability of the ionized to nonionized forms. In conclusion, Kp,uu,ss values of anions are similar to Kp,uu,true when the inside-negative membrane potential is considered. This suggests that Kp,uu,ss is preferable for estimating the concentration of unbound drugs inside the hepatocytes.

Analysis of Nonlinear Pharmacokinetics of a Highly Albumin-Bound Compound: Contribution of Albumin-Mediated Hepatic Uptake Mechanism.

The cause of nonlinear pharmacokinetics (PK) (more than dose-proportional increase in exposure) of a urea derivative under development (compound A: anionic compound [pKa: 4.4]; LogP: 6.5; and plasma protein binding: 99.95%) observed in a clinical trial was investigated. Compound A was metabolized by CYP3A4, UGT1A1, and UGT1A3 with unbound Km of 3.3-17.8 µmol/L. OATP1B3-mediated uptake of compound A determined in the presence of human serum albumin (HSA) showed that unbound Km and Vmax decreased with increased HSA concentration. A greater decrease in unbound Km than in Vmax resulted in increased uptake clearance (Vmax/unbound Km) with increased HSA concentration, the so-called albumin-mediated uptake. At 2% HSA concentration, unbound Km was 0.00657 µmol/L. A physiologically based PK model assuming saturable hepatic uptake nearly replicated clinical PK of compound A. Unbound Km for hepatic uptake estimated from the model was 0.000767 µmol/L, lower than the in vitro unbound Km at 2% HSA concentration, whereas decreased Km with increased concentration of HSA in vitro indicated lower Km at physiological HSA concentration (4%-5%). In addition, unbound Km values for metabolizing enzymes were much higher than unbound Km for OATP1B3, indicating that the nonlinear PK of compound A is primarily attributed to saturated OATP1B3-mediated hepatic uptake of compound A.