Chimeric mice with humanized liver: Application in drug metabolism and pharmacokinetics studies for drug discovery

Predicting human drug metabolism and pharmacokinetics (PK) is key to drug discovery. In particular, it is important to predict human PK, metabolite profiles and drug-drug interactions (DDIs). Various methods have been used for such predictions, including in vitro metabolic studies using human biological samples, such as hepatic microsomes and hepatocytes, and in vivo studies using experimental animals. However, prediction studies using these methods are often inconclusive due to discrepancies between in vitro and in vivo results, and interspecies differences in drug metabolism. Further, the prediction methods have changed from qualitative to quantitative to solve these issues. Chimeric mice with humanized liver have been developed, in which mouse liver cells are mostly replaced with human hepatocytes. Since human drug metabolizing enzymes are expressed in the liver of these mice, they are regarded as suitable models for mimicking the drug metabolism and PK observed in humans; therefore, these mice are useful for predicting human drug metabolism and PK. In this review, we discuss the current state, issues, and future directions of predicting human drug metabolism and PK using chimeric mice with humanized liver in drug discovery.     

Practical Anticipation of Human Efficacious Doses and Pharmacokinetics Using In Vitro and Preclinical In Vivo Data

Accurate predictions of human pharmacokinetic and pharmacodynamic (PK/PD) profiles are critical in early drug development, as safe, efficacious, and “developable” dosing regimens of promising compounds have to be identified. While advantages of successful integration of preclinical PK/PD data in the “anticipation” of human doses (AHD) have been recognized, pharmaceutical scientists have faced difficulties with practical implementation, especially for PK/PD profile projections of compounds with challenging absorption, distribution, metabolism, excretion and formulation properties. In this article, practical projection approaches for formulation-dependent human PK/PD parameters and profiles of Biopharmaceutics Classification System classes I-IV drugs based on preclinical data are described. Case examples for “AHD” demonstrate the utility of preclinical and clinical PK/PD modeling for formulation risk identification, lead candidate differentiation, and prediction of clinical outcome. The application of allometric scaling methods and physiologically based pharmacokinetic approaches for clearance or volume of distribution projections is described using GastroPlus™. Methods to enhance prediction confidence such as in vitro–in vivo extrapolations in clearance predictions using in vitro microsomal data are discussed. Examples for integration of clinical PK/PD and formulation data from frontrunner compounds via “reverse pharmacology strategies” that minimize uncertainty with PK/PD predictions are included. The use of integrated softwares such as GastroPlus™ in combination with established PK projection methods allow the projection of formulation-dependent preclinical and human PK/PD profiles required for compound differentiation and development risk assessments.Key words: formulation, human dose prediction, modeling, PBPK, PK/PD     

Dose Selection Based on Physiologically Based Pharmacokinetic (PBPK) Approaches

Physiologically based pharmacokinetic (PBPK) models are built using differential equations to describe the physiology/anatomy of different biological systems. Readily available in vitro and in vivo preclinical data can be incorporated into these models to not only estimate pharmacokinetic (PK) parameters and plasma concentration–time profiles, but also to gain mechanistic insight into compound properties. They provide a mechanistic framework to understand and extrapolate PK and dose across in vitro and in vivo systems and across different species, populations and disease states. Using small molecule and large molecule examples from the literature and our own company, we have shown how PBPK techniques can be utilised for human PK and dose prediction. Such approaches have the potential to increase efficiency, reduce the need for animal studies, replace clinical trials and increase PK understanding. Given the mechanistic nature of these models, the future use of PBPK modelling in drug discovery and development is promising, however some limitations need to be addressed to realise its application and utility more broadly.     

Quantitative Prediction of Human Pharmacokinetics for mAbs Exhibiting Target-Mediated Disposition

Prediction of human pharmacokinetics (PK) can be challenging for monoclonal antibodies (mAbs) exhibiting target-mediated drug disposition (TMDD). In this study, we performed a quantitative analysis of a diverse set of six mAbs exhibiting TMDD to explore translational rules that can be utilized to predict human PK. A TMDD model with rapid-binding approximation was utilized to fit PK and PD (i.e., free and/or total target levels) data, and average absolute fold error (AAFE) was calculated for each model parameter. Based on the comparative analysis, translational rules were developed and applied to a test antibody not included in the original analysis. AAFE of less than two-fold was observed between monkey and human for baseline target levels (R₀), body-weight (BW) normalized central elimination rate (K_el/BW^−0.25) and central volume (V_c/BW^1.0). AAFE of less than three-fold was estimated for the binding affinity constant (K_D). The other four parameters, i.e., complex turnover rate (K_int), target turnover rate (K_deg), central to peripheral distribution rate constant (K_pt) and peripheral to central rate constant (K_tp) were poorly correlated between monkey and human. The projected human PK of test antibody based on the translation rules was in good agreement with the observed nonlinear PK. In conclusion, we recommend a TMDD model-based prediction approach that integrates in vitro human biomeasures and in vivo preclinical data using translation rules developed in this study.

Electronic supplementary material

The online version of this article (doi:10.1208/s12248-014-9690-8) contains supplementary material, which is available to authorized users.KEY WORDS: ADME of biologics, human translation, monoclonal antibodies, PK/PD modeling, TMDD     

Benchmarking of Human Dose Prediction for Inhaled Medicines from Preclinical <Emphasis Type="Italic">In Vivo</Emphasis> Data

Purpose

A scientifically robust prediction of human dose is important in determining whether to progress a candidate drug into clinical development. A particular challenge for inhaled medicines is that unbound drug concentrations at the pharmacological target site cannot be easily measured or predicted. In the absence of such data, alternative empirical methods can be useful. This work is a post hoc analysis based on preclinical in vivo pharmacokinetic/pharmacodynamic (PK/PD) data with the aim to evaluate such approaches and provide guidance on clinically effective dose prediction for inhaled medicines.

Methods

Five empirically based methodologies were applied on a diverse set of marketed inhaled therapeutics (inhaled corticosteroids and bronchodilators). The approaches include scaling of dose based on body weight or body surface area and variants of PK/PD approaches aiming to predict the therapeutic dose based on having efficacious concentrations of drug in the lung over the dosing interval.

Results

The most robust predictions of dose were made by body weight adjustment (90% within 3-fold) and by a specific PK/PD approach aiming for an average predicted 75% effect level during the dosing interval (80% within 3-fold). Scaling of dose based on body surface area consistently under predicted the therapeutic dose.

Conclusions

Preclinical in vivo data and empirical scaling to man can be used as a baseline method for clinical dose predictions of inhaled medicines. The development of more sophisticated translational models utilizing free drug concentration and target engagement data is a desirable build.

     

Predicting human exposure of active drug after oral prodrug administration,using a joined in vitro/in silico–in vivo extrapolation and physiologically-based pharmacokinetic modeling approach

Introduction: Predicting the pharmacokinetics (PK) of prodrugs and their corresponding active drugs is challenging, as there are many variables to consider. Prodrug conversion characteristics in different tissues are generally measured, but integrating these variables to a PK profile is not a common practice. In this paper, a joined in vitro/in silico–in vivo extrapolation (IVIVE) and physiologically-based pharmacokinetic (PBPK) modeling approach is presented to predict active drug exposure in human after oral prodrug administration. Methods: Physico-chemical and in vitro assays as well as in silico predictions were proposed to characterize key pharmacokinetic properties (e.g. clearance, volume of distribution, conversion rates) of three marketed prodrugs. These data were used to parameterize a PBPK model for simulating human PK profiles of the active drugs after prodrug administration, which were compared to literature data by evaluating the accuracy and uncertainty of the predictions. Results: For mycophenate mofetil and midodrine the PK of their active moieties could be adequately predicted. The assumptions of the PBPK–IVIVE approach were valid, i.e. being hepatically cleared, converted in the gut lumen, blood and liver and not metabolized in the gut wall. However, the observed profiles after oral bambuterol administration clearly fell outside the prediction interval as the PBPK model failed to predict the observed bioavailability. Discussion: Adding quantitative information about prodrug conversion in the gut, liver and blood to a PBPK model for the absorption, distribution, metabolism and excretion (ADME) properties of prodrugs and their active moieties resulted, retrospectively, in reasonable predictions of the human PK when the ADME properties are well understood. Also in a prospective compound selection process, this integrative approach can improve decision making on prodrug candidates by putting relative differences in prodrug conversion of a large number of candidates into the perspective of their human PK profile, before conducting any in vivo experiments.     

Using DTA and DTAARRAY Variables and Programming in WinNonlin ASCII Models to Streamline User-Defined Calculation and Data Analysis

In pharmacokinetic (PK) analysis, there are many occasions where user-defined calculations need to be performed before or after the primary PK modeling/analysis. Conventionally, these calculations are often executed outside of the primary PK analysis by pre- or post-processing data from multiple sources, manually entering formulas and multiple additional set-ups. Such analysis approaches increase the risk of generating data defects and can employ software that is not fully compliant. We propose a method of leveraging DTA and DTAARRAY variables plus simple programming techniques in an ASCII model to automate these user-defined calculations in WinNonlin and eliminate the need for manual handling of data outside of the primary analysis. We demonstrated the application of this strategy through three case study examples. In case 1 (post-processing data), DTA variables were used to calculate three user-defined parameters in the primary PK model. In case 2 (pre-processing data), a baseline correction decision tree was programmed into the PK model to account for both the endogenous baseline level as well as the presence of residual drug. In case 3, DTAARRAY variables were used to perform a looping operation to calculate the difference factor (F1) and the similarity factor (F2) in support of in vitro bioequivalence evaluations.

Electronic supplementary material

The online version of this article (doi:10.1208/s12248-014-9711-7) contains supplementary material, which is available to authorized users.Key words: ASCII model, decision tree, DTA, DTAARRAY, programming, user-defined calculation, WinNonlin     

Chimeric mice transplanted with human hepatocytes as a model for prediction of human drug metabolism and pharmacokinetics

Preclinical studies in animal models are used routinely during drug development, but species differences of pharmacokinetics (PK) between animals and humans have to be taken into account in interpreting the results. Human hepatocytes are also widely used to examine metabolic activities mediated by cytochrome P450 (P450) and other enzymes, but such in vitro metabolic studies also have limitations. Recently, chimeric mice with humanized liver (h‐chimeric mice), generated by transplantation of human donor hepatocytes, have been developed as a model for the prediction of metabolism and PK in humans, using both in vitro and in vivo approaches. The expression of human‐specific metabolic enzymes and metabolic activities was confirmed in humanized liver of h‐chimeric mice with high replacement ratios, and several reports indicate that the profiles of P450 and non‐P450 metabolism in these mice adequately reflect those in humans. Further, the combined use of h‐chimeric mice and r‐chimeric mice, in which endogenous hepatocytes are replaced with rat hepatocytes, is a promising approach for evaluation of species differences in drug metabolism. Recent work has shown that data obtained in h‐chimeric mice enable the semi‐quantitative prediction of not only metabolites, but also PK parameters, such as hepatic clearance, of drug candidates in humans, although some limitations remain because of differences in the metabolic activities, hepatic blood flow and liver structure between humans and mice. In addition, fresh h‐hepatocytes can be isolated reproducibly from h‐chimeric mice for metabolic studies. Copyright © 2013 John Wiley & Sons, Ltd.     

Drug discovery information integration: virtual humans for pharmacokinetics

The ‘virtual human’ refers to simulation models based on explicit mathematical models of mammalian physiology. When applied to pharmacokinetics (PK), the virtual human embodies models that can be parameterised for different species, different individuals and populations of individuals, and that allow simulation of PK from measured and/or predicted in vitro properties. The models are independent of the properties of any specific drug, and can be used for the prediction of drug behaviour in specific human individuals and in pre-clinical species. The hope is that these models will allow the prediction of PK throughout the drug discovery and development process, enabling effective targeting of synthetic chemistry efforts to compounds with the required therapeutic effect, careful evaluation of multiple candidates for development, assessment of the potential for drug-drug interactions, evaluation of multiple formulations, efficient clinical trial design and other benefits, including reduction of animal usage in drug discovery. However, we believe that the biggest impact on drug discovery productivity would accrue from achieving two clear objectives: an estimate of human in vivo activity on every compound synthesised and a reliable prioritisation, based on predictions of human in vivo activity, for the next round of synthesis. It is realistic to believe that the benefits of delivering these objectives could, through shortening the time to decision and increasing the chance of success, lead to a five- to ten-fold increase in discovery output. It is also now possible to demonstrate that a combination of virtual human simulation software and industrial-scale absorption, distribution, metabolism and elimination (ADME) testing can deliver both of these objectives.     

Comparison of information on the pharmacokinetic interactions of Ca antagonists in the package inserts from three countries (Japan,USA and UK)

Objective: Ca antagonists are one of the most popular classes of drugs used to treat hypertension and angina. These drugs may interact with either CYP3A4 or MDR-1 substrates, with the degree of interaction differing with each drug. We carried out a literature search to examine and compare the extent to which crucial pharmacokinetic (PK) information is included in package inserts (PIs) in Japan, USA and the UK. Methods: A MEDLINE search from 1966 to November 2004 was undertaken with the aim of identifying studies on clinical PK drug interactions between seven Ca antagonists that are available in three countries and three CYP3A4 inhibitors (erythromycin, itraconazole and cimetidine), a CYP3A4 inhibitory food, grapefruit juice (GFJ) and the MDR-1 substrate, digoxin. The current PIs for Ca antagonists were obtained from the website of the regulatory authorities or the electronic Medicines Compendium. Results: Of all possible combinations of seven Ca antagonists with three CYP3A4 inhibitor drugs, drug interaction information was available in the literature on nine combinations: Seven of these were listed in the USA PIs, two in the UK PIs, and none in the Japanese PIs. Interaction studies with GFJ were reported for every Ca antagonist; PIs in the USA provided quantitative data for four of these interactions, whereas UK PIs provided quantitative data for only one of the interactions and Japanese PIs provided no quantitative information. The PK data of co-medication of digoxin with Ca antagonists have been reported for every Ca antagonists. The USA PIs provided quantitative data for five Ca antagonists, whereas the UK PIs provided quantitative data for three Ca antagonists and Japanese PIs provided no quantitative data. Conclusion: The literature search revealed that PIs in the USA provided a great deal of quantitative information on PK interactions between Ca antagonists and other drugs or GFJ. In contrast, PIs in the UK and Japan did not provide sufficient information. We conclude that crucial quantitative information on these drug interactions should be incorporated in PIs, especially in Japan and the UK, as a means of assisting healthcare providers.     

Exploratory translational modeling approach in drug development to predict human brain pharmacokinetics and pharmacologically relevant clinical doses

The central nervous system (CNS) pharmacokinetics (PK) of drugs that have pharmacological targets in the brain are not often understood during drug development, and this gap in knowledge is a limitation in providing a quantitative framework for translating nonclinical pharmacologic data to the clinical patient population. A focus of translational sciences is to improve the efficiency of clinical trial design via a more judicious selection of clinical doses on the basis of nonclinical data. We hypothesize that this can be achieved for CNS-acting drugs based on knowledge of CNS PK and brain target engagement obtained in nonclinical studies. Translating CNS PK models from rat to human can allow for the prediction of human brain PK and the human dose-brain exposure relationship, which can provide insight on the clinical dose(s) having potential brain activity and target engagement. In this study, we explored the potential utility of this translational approach using rat brain microdialysis and PK modeling techniques to predict human brain extracellular fluid PK of atomoxetine and duloxetine. The results show that this translational approach merits consideration as a means to support the clinical development of CNS-mediated drug candidates by enhancing the ability to predict pharmacologically relevant doses in humans in the absence of or in association with other biomarker approaches.     

A Mechanistic Pharmacokinetic Model Elucidating the Disposition of Trastuzumab Emtansine (T-DM1), an Antibody–Drug Conjugate (ADC) for Treatment of Metastatic Breast Cancer

Trastuzumab emtansine (T-DM1) is an antibody–drug conjugate (ADC) therapeutic for treatment of human epidermal growth factor receptor 2 (HER2)-positive cancers. The T-DM1 dose product contains a mixture of drug-to-antibody ratio (DAR) moieties whereby the small molecule DM1 is chemically conjugated to trastuzumab antibody. The pharmacokinetics (PK) underlying this system and other ADCs are complex and have not been elucidated. Accordingly, we have developed two PK modeling approaches from preclinical data to conceptualize and understand T-DM1 PK, to quantify rates of DM1 deconjugation, and to elucidate the link between trastuzumab, T-DM1, and DAR measurements. Preclinical data included PK studies in rats (n = 34) and cynomolgus monkeys (n = 18) at doses ranging from 0.3 to 30 mg/kg and in vitro plasma stability. T-DM1 and total trastuzumab (TT) plasma concentrations were measured by enzyme-linked immunosorbent assay. Individual DAR moieties were measured by affinity capture liquid chromatography-mass spectrophotometry. Two PK modeling approaches were developed for T-DM1 using NONMEM 7.2 software: a mechanistic model fit simultaneously to TT and DAR concentrations and a reduced model fit simultaneously to TT and T-DM1 concentrations. DAR moieties were well described with a three-compartmental model and DM1 deconjugation in the central compartment. DM1 deconjugated fastest from the more highly loaded trastuzumab molecules (i.e., DAR moieties that are ≥3 DM1 per trastuzumab). T-DM1 clearance (CL) was 2-fold faster than TT CL due to deconjugation. The two modeling approaches provide flexibility based on available analytical measurements for T-DM1 and a framework for designing ADC studies and PK–pharmacodynamic modeling of ADC efficacy- and toxicity-related endpoints.KEY WORDS: antibody–drug conjugate, deconjugation, population pharmacokinetic model, T-DM1, trastuzumab emtansine     

Integration of In Silico Pharmacokinetic Modeling Approaches Into In Vitro Dissolution Profiles to Predict Bioavailability of a Poorly Soluble Compound

The objective of present study is to develop pharmacokinetic (PK) prediction methods using in silico PK model for oral immediate release drug products (i.e. solution, suspension, and amorphous solid dispersion). A poorly water soluble compound with low bioavailability in rat was used (CS-758 as a model compound). A constructed in silico PK model contained an advance compartmental absorption and transit model. For solution, the in silico PK model reproduced an observed rat plasma concentration (Cp)-time profile. In addition, an in vitro dissolution method was developed to predict a rat Cp-time profile for suspension. As a result, the in silico PK model could predict the observed one by using dissolution profiles as the input. Furthermore, a dissolution profile of amorphous solid dispersion was applied to verify the in silico PK model. A result indicated the simulated rat Cp-time profile was significantly comparable to the observed one. This study demonstrated that the integration of an in silico PK model into dissolution profiles can predict rat Cp-time profiles for suspension and amorphous solid dispersion. These results suggest that the integration of in silico PK modeling approaches into dissolution profiles can contribute to the formulation screening for poorly soluble compounds by predicting PK behaviors.     

Applications of minimal physiologically-based pharmacokinetic models

Conventional mammillary models are frequently used for pharmacokinetic (PK) analysis when only blood or plasma data are available. Such models depend on the quality of the drug disposition data and have vague biological features. An alternative minimal-physiologically-based PK (minimal-PBPK) modeling approach is proposed which inherits and lumps major physiologic attributes from whole-body PBPK models. The body and model are represented as actual blood and tissue (usually total body weight) volumes, fractions (f _d ) of cardiac output with Fick??s Law of Perfusion, tissue/blood partitioning (K _p ), and systemic or intrinsic clearance. Analyzing only blood or plasma concentrations versus time, the minimal-PBPK models parsimoniously generate physiologically-relevant PK parameters which are more easily interpreted than those from mammillary models. The minimal-PBPK models were applied to four types of therapeutic agents and conditions. The models well captured the human PK profiles of 22 selected beta-lactam antibiotics allowing comparison of fitted and calculated K _p values. Adding a classical hepatic compartment with hepatic blood flow allowed joint fitting of oral and intravenous (IV) data for four hepatic elimination drugs (dihydrocodeine, verapamil, repaglinide, midazolam) providing separate estimates of hepatic intrinsic clearance, non-hepatic clearance, and pre-hepatic bioavailability. The basic model was integrated with allometric scaling principles to simultaneously describe moxifloxacin PK in five species with common K _p and f _d values. A basic model assigning clearance to the tissue compartment well characterized plasma concentrations of six monoclonal antibodies in human subjects, providing good concordance of predictions with expected tissue kinetics. The proposed minimal-PBPK modeling approach offers an alternative and more rational basis for assessing PK than compartmental models.     

Modeling of Pharmacokinetic Systems Using Stochastic Deconvolution

In environments where complete mechanistic knowledge of the system dynamics is not available, a synergy of first-principle concepts, stochastic methods and statistical approaches can provide an efficient, accurate, and insightful strategy for model development. In this work, a system of ordinary differential equations describing system pharmacokinetics (PK) was coupled to a Wiener process for tracking the absorption rate coefficient, and was embedded in a nonlinear mixed effects population PK formalism. The procedure is referred to as “stochastic deconvolution” and it is proposed as a diagnostic tool to inform on a mapping function between the fraction of the drug absorbed and the fraction of the drug dissolved when applying one-stage methods to in vitro–in vivo correlation modeling. The goal of this work was to show that stochastic deconvolution can infer an a priori specified absorption profile given dense observational (simulated) data. The results demonstrate that the mathematical model is able to accurately reproduce the simulated data in scenarios where solution strategies for linear, time-invariant systems would assuredly fail. To this end, PK systems that are representative of Michaelis–Menten kinetics and enterohepatic circulation were investigated. Furthermore, the solution times are manageable using a modest computer hardware platform.     

A Simplified PBPK Modeling Approach for Prediction of Pharmacokinetics of Four Primarily Renally Excreted and CYP3A Metabolized Compounds During Pregnancy

During pregnancy, a drug’s pharmacokinetics may be altered and hence anticipation of potential systemic exposure changes is highly desirable. Physiologically based pharmacokinetics (PBPK) models have recently been used to influence clinical trial design or to facilitate regulatory interactions. Ideally, whole-body PBPK models can be used to predict a drug’s systemic exposure in pregnant women based on major physiological changes which can impact drug clearance (i.e., in the kidney and liver) and distribution (i.e., adipose and fetoplacental unit). We described a simple and readily implementable multitissue/organ whole-body PBPK model with key pregnancy-related physiological parameters to characterize the PK of reference drugs (metformin, digoxin, midazolam, and emtricitabine) in pregnant women compared with the PK in nonpregnant or postpartum (PP) women. Physiological data related to changes in maternal body weight, tissue volume, cardiac output, renal function, blood flows, and cytochrome P450 activity were collected from the literature and incorporated into the structural PBPK model that describes HV or PP women PK data. Subsequently, the changes in exposure (area under the curve (AUC) and maximum concentration (C _max)) in pregnant women were simulated. Model-simulated PK profiles were overall in agreement with observed data. The prediction fold error for C _max and AUC ratio (pregnant vs. nonpregnant) was less than 1.3-fold, indicating that the pregnant PBPK model is useful. The utilization of this simplified model in drug development may aid in designing clinical studies to identify potential exposure changes in pregnant women a priori for compounds which are mainly eliminated renally or metabolized by CYP3A4.