Anticipation of drug-drug interactions (DDIs) in the paediatric population are merely based on data generated in adults. Hence decision on avoiding certain combinations or attempts to adjust and manage the doses under combination-therapy are mainly speculative from the knowledge of what occurs in adults. However, due to developmental changes in elimination pathways from birth to adolescents, the assumption of DDIs being similar in adults and children might not be correct. This thesis firstly identifies and quantitatively compares the reported DDIs in paediatric and adult populations through a systematic literature review of DDIs reported in paediatric subjects. The study highlights the clear paucity of the data in children younger than 2 years. Therefore, the logical approach to test this hypothesis has been through modelling and simulation and incorporation of the biological knowledge on ontogeny of various enzymes and other elimination routes. The magnitude of any metabolic DDI depends on fractional importance of inhibited pathway which may not necessarily be the same in young children when compared to adults. To show this disparity between rate of ontogeny for metabolic pathways, the ontogeny pattern of CYP enzymes and renal function were analysed systematically. Bootstrap methodology was used to account for variability, and to define the age range over which a statistical difference is likely between each pair of specific pathways. A number of DDIs were simulated for virtual compounds to highlight the possibility that the magnitude of DDI can be influenced by age. Depending on the extent of contribution of metabolic pathways, neonates could be more sensitive to DDI than adults in certain scenarios or vice versa. Thus, extrapolation from adult DDI data may not be applicable across paediatric age groups. The uncertainty around the ontogeny functions based on in vitro information led us to carry out comprehensive performance verification for in vivo data on probe substrates of CYP1A2, -2C9 and 3A4 and assess the predictions of clearance (CL) by monitoring AUC. Although the evaluation showed that in most cases predictions were within two fold of observed data in adult and paediatric studies, the outcome suggests that the current ontogeny profiles result in under-prediction of CL values compared to clinical studies in infants and children and there is a need for better ontogeny models. Therefore, we derived novel ontogeny functions for CYP1A2 and CYP3A based on caffeine-theophylline and midazolam in vivo data. Age related CL data for caffeine, theophylline and midazolam were reconstructed back to intrinsic CL values per milligram of microsomal protein and best fit ontogeny models for CYP1A2 and CYP3A were derived from these data. The function for CYP1A2 describes an increase in relative intrinsic metabolic CL from birth to 3 years followed by a decrease to adult values. The function for CYP3A4 describes a continuous rise in relative intrinsic metabolic CL, reaching the adult value at about 2 years of age. The new models were validated by showing improved predictions of the systemic CL of ropivacaine (major CYP1A2 substrate; minor CYP3A4 substrate) and alfentanil (major CYP3A4 substrate) compared to those using a previous ontogeny function based on in vitro data. When implementing enzyme ontogeny functions it is important to consider potential confounding factors related to disease, anaesthesia and surgery that may affect the prediction of net in vivo CL. Finally, we demonstrated the application of paediatric physiologically-based pharmacokinetic (p-PBPK) models for calculation of sample size in paediatric clinical pharmacokinetic (PK) studies in a methodology suggested by Wang et al., based on desired precision for a PK parameter of interest. We obtained estimates of variability for CL, volume of distribution and area under the plasma concentration-time curve for 5 different drugs from (i) adult and paediatric classic clinical PK studies, and (ii) p-PBPK combined with in vitro-in vivo extrapolation. The estimates were applied to the sample size calculation proposal methodology for non-compartmental analysis. There were clear and drug dependent differences in calculated sample size based on various estimates of variability and overall, there was no consistent discrepancy in the sample size calculated according to the source of variability used for sample size calculations. The results are discussed in terms of their potential impact on the clinical PK studies in children. In general, considering the sensitivity of paediatric clinical PK studies and paucity of data in this group of patients, the use of p-PBPK models may offer an interim solution to uncovering age bands with potential higher vulnerability to DDI. However, these models require further refinements and testing before widely used in clinical practice with confidence.

Physiologically-based pharmacokinetic (PBPK) modelling is traditionally employed to predict drug concentration-time profiles in plasma and tissues using information from physiology/biology, in vitro experiments and in silico predictions. Model-based analysis of population pharmacokinetic (PK) data is rarely performed in such a mechanistic framework, as empirical compartmental models are mainly utilised for this purpose. However, the combination of traditional PBPK methodologies with parameter estimation techniques and non-linear mixed effects modelling is an approach with progressively increasing impact due to the significant advantages it offers. Therefore, the general aim of this thesis is to illustrate, explore and thus further facilitate the application of physiologically-based pharmacokinetic models in the context of population data analysis. In order to pursue this aim, this work firstly particularly focuses on the population pharmacokinetics of simvastatin (SV) and its active metabolite, simvastatin acid (SVA). The complex simvastatin pharmacokinetics and their clinical significance, due to the association with simvastatin-induced myopathy, provide an excellent case to illustrate the advantages of a mechanistically sound population model. In the current work, both conventional and physiologically-based population models were developed using clinical PK data for SV and SVA. Specifically, the developed model-based approaches successfully quantified the impact of demographics, genetic polymorphisms and drug-drug interactions (DDIs) on the SV/SVA pharmacokinetics. Therefore, they can be of significant application either in the clinic or during drug development in order to assess myopathy and DDI risk. Secondly, in this work the following advantages offered by integrated population PBPK modelling were clearly illustrated through specific applications: 1) prediction of drug concentrations at the tissue level, 2) ability to extrapolate outside the studied population and/or conditions and 3) ability to guide the design (sample size) of prospective clinical studies. Finally, in the current work, further methodological aspects related to the application of this integrated population PBPK modelling approach were explored. Of specific focus was the parameter estimation process aided by prior distributions and the derivation of the latter from different in vitro/in silico sources. In addition, specific methodology is illustrated in this work that allows the incorporation of stochastic population variability in the structural parameters of such models without neglecting the underlying physiological constraints.

Une étape critique au cours du développement de médicaments est la mesure ou la prédiction des concentrations du médicament dans un tissu cible, qui peuvent ensuite être liées à des mesures de leur efficacité ou leur toxicité. Les concentrations de médicaments ne pouvant être mesurées dans le cerveau humain, ils doivent être simulés ou prédits en utilisant des approches alternatives de modélisation. L'objectif de cette thèse est de développer in silico des approches de prédiction combinant à la fois des données précliniques in vitro et in vivo dans un modèle physiologique structuré, avec une stratégie translationnelle afin de permettre la simulation de concentrations totales et libres des médicaments dans le cerveau humain. Des modèles pharmacocinétiques physiologiques (PBPK) ont été développés dans cette thèse et évalués pour des médicaments de référence déjà sur le marché, et pour un médicament en cours de développement clinique dans l'industrie pharmaceutique. Ces modèles ont été développés à partir de stratégies de type « Bottom-up » sur la base de données in vitro pour la prédiction de la distribution des médicaments dans le cerveau et comparées à des méthodes de type «top-down » en utilisant des données in vivo. Cette thèse est une thèse sur article construite à partir de 5 articles scientifiques qui sont soit publiés soit en cours de soumission. Le premier article est une revue de la littérature publiée dans le Journal de l'AAPS. Cette revue discute les modèles PBPK disponibles actuellement et a permis d’élaborer des hypothèses de travail dans cette thèse afin de proposer des améliorations de ces modèles. Le deuxième article un article de recherche original publié dans Molecular Pharmaceutics. Cet article vise à démontrer l'élaboration d'une approche cohérente de modélisation pharmacocinétique chez le rat qui peut s’adapter en fonction de la quantité et de la qualité des données obtenues in vivo au cours du développement des médicaments. Un arbre de décision a été construit pour faciliter le paramétrage et la structure appropriée du modèle en fonction des données disponibles. Le troisième article est un article de recherche original publié dans Journal of Pharmaceutical Sciences. Cette étude porte sur le développement d'un modèle PBPK pour la prédiction de la pénétration cérébrale des médicaments, dans lequel son transfert à travers la BHE a été traité de façon mécanistique en séparant les paramètres régissant la quantité (perméabilité) à travers la BHE de sa liaison dans le tissu cérébral. Une stratégie de type vitro - vivo en fonction de la perméabilité des médicaments à travers les monocouches cellulaires in vitro a été proposé afin d'extrapoler la composante de transport actif du composé à l’aide de facteurs d’extrapolation (RAF). Deux autres articles sont en cours d’écriture ou soumis. Ces articles viennent compléter les approches de PBPK pour les médicaments du SNC décrites dans les deux autres articles originaux. Une dernière partie de la thèse constitue la discussion qui met très clairement en évidence l'importance du choix d’une approche de modélisation appropriée ou mieux encore la combinaison des approches fondées sur les connaissances physiologiques, les données expérimentales et les applications prévues dans le développement du médicament. L'avantage du paramétrage mécanistique dans ces modèles PBPK est qu’il améliore leur prédictivité et la simulation de différences inter-espèces. Cette thèse a considérablement contribué à démontrer la nécessité d’associer des données in vitro à des données in vivo dans la structuration des modèles PBPK qui se révèlent alors comme des outils précieux pour la prédiction de la pharmacocinétique cérébrale chez l'homme. / A critical step during drug development is the measurement or the prediction of drug concentrations in the target tissue, which can then be linked to measures of drug efficacy or toxicity. Drug concentrations cannot be directly measured in the human brain, and must be simulated or predicted using alternative modeling approaches. The objective of this thesis is to develop in silico approaches to predict BBB penetration of drugs, combining in vitro and in vivo preclinical data in a physiologically structured model, with a translational strategy to allow the simulation of total and free drug concentrations in the human brain. Physiologically based pharmacokinetic (PBPK) models were developed and evaluated for reference molecules already on the market, as well as for a drug currently under clinical development within the pharmaceutical industry. These models were developed based on both ‘bottom-up’ (model parameter values predicted from in vitro data) and ‘top-down’ (model parameters estimated from in vivo data) strategies. This thesis is comprised of 5 scientific papers which are either published or submitted to peer-reviewed journals. The first article is a review of the literature, published in the AAPS journal. This review discusses the currently published PBPK models available for the mechanistic prediction of BBB penetration of drugs, and proposes a strategy for in vitro-in vivo (IVIVE) extrapolation. The second article is an original research article published in Molecular Pharmaceutics. This article aims to show the development of a coherent pharmacokinetic modeling approach in the rat which can be adapted based on the quantity and quality of data obtained in vivo during the development of new drugs. A decision tree was constructed to enable the appropriate parameterization and model structure based on the available data. The third article is an original research article published in Journal of Pharmaceutical Sciences. This article was based on the development of a PBPK model for the mechanistic prediction of BBB penetration of drugs, in which the active and passive components of permeability were considered separately, as well as the intra-brain tissue binding parameters. An in vitro-in vivo strategy was proposed to extrapolate the active transport component using a relative activity factor (RAF) to account for in vitro-in vivo differences in transporter activity and/or abundance. Two additional articles are either submitted or under preparation. These articles extend the PBPK approaches described in the previous two published original research articles. The final part of this thesis consists of a discussion which emphasizes clearly the importance of the appropriate choice of modeling approach, or even better, a combination of approaches based on physiological knowledge, experimental data and knowledge gathered during the course of drug development. The advantage of mechanistic parameterization of PBPK models is the improved ability for inter-species extrapolation for the subsequent simulation of free or total drug concentrations within the human brain. This thesis has considerably contributed to this rapidly evolving field of CNS drug research and development, showing the importance of combining in vitro and in vivo data within a physiologically based model structure, thus providing a valuable tool for the quantitative prediction of the penetration of drugs in the human brain.

Ertapenem is a carbapenem used to treat a wide range of bacterial infections. What sets ertapenem apart from other carbapenems is its longer half-life which implies it need only be administered once daily. We developed a physiologically-based pharmacokinetic model for the distribution of ertapenem within the body. In the model, parameters such as human body weight and height, age, organ volumes, blood flow rates, and partition coefficients of particular tissues are used to examine the absorption, distribution, metabolism, and excretion of ertapenem. The total and free blood concentrations found were then compared to experimental data. We then examined the sensitivity of the total concentration in the blood to body weight, body height, and age. This analysis allows the possibility of the model being used as a basis for understanding how differing health conditions might alter the concentration of ertapenem in the body and hence dosage may need to be adjusted.

Assessment of human exposure to environmental chemicals is inherently subject to uncertainty and variability. There are data gaps concerning the inventory, source, duration, and intensity of exposure as well as knowledge gaps regarding pharmacokinetics in general. These gaps result in uncertainties in exposure assessment. The uncertainties compound further with variabilities due to population variations regarding stage of life, life style, and susceptibility, etc. Use of physiologically-based pharmacokinetic (PBPK) models promises to reduce the uncertainties and enhance extrapolation between species, between routes, from high to low dose, and from acute to chronic exposure. However, fitting PBPK models is challenging because of a large number of biochemical and physiological parameters to be estimated. Many of these model parameters are non-identifiable in that their estimates cannot be uniquely determined using statistical criteria. In practice some parameters are fixed in value and some determined through mathematical calibration or computer simulation. These estimated values are subject to substantial uncertainties. The first part of this paper illustrates the use of iteratively-reweighted-nonlinear-least-squares for fitting pharmacokinetic (PK) models, highlighting some common difficulties in obtaining statistical estimates of non-identifiable parameters and use bootstrap confidence interval to quantify uncertainties. Statistical estimation of parameters in physiologically based pharmacokinetic (PBPK) models is a relatively new area of research. Over the past decade or so PBPK models have become important and valuable tools in risk assessment as these models are used to describe the absorption, distribution, metabolism, and excretion of xenobiotics in a biological system such as the human or rat. Because these models incorporate information on biological processes, they are well equipped to describe the kinetic behaviors of chemicals and are useful for extrapolation across dose routes, between species, from high-to-low-doses, and across exposure scenarios. A PBPK model has been developed based on published models in the literature to describe the absorption, distribution, metabolism, and excretion of Dioxin and dioxin like compounds (DLCs) in the rat. Data from the National Toxicology Program (NTP) two year experiment TR-526 is used to illustrate model fitting and statistical estimation of the parameters. Integrating statistical methods into risk assessments is the most efficient way to characterize the variation in parameter values. In this dissertation a Markov Chain Monte Carlo (MCMC) method is used to estimate select parameters of the system and to describe the variation of the select parameters.

Background: Elucidating the role of intestinal drug transporter function in drug development is crucial, as transporter proteins can impact on drug absorption, efficacy and adverse events. In Vitro-In Vivo Extrapolation linked to Physiologically-Based Pharmacokinetic (IVIVE-PBPK) models aim to predict the in vivo impact of transporters from in vitro cell–based transport data and expression-based scaling factors. Currently, these models depend on relative measurements of transporter expression i.e., mRNA or immunoblotting. There is a critical need for physiologically relevant measures of transporter protein abundance to populate these biological frameworks. Objectives: The key objectives were to develop and validate a targeted proteomics workflow to quantify transporter protein abundances in human enterocytes and Caco-2 cells with a QconCAT technique. A cross-laboratory comparison on matched samples was also performed to assess between-laboratory bias in abundance determination. Together with abundance data from each laboratory, BCRP and P-gp transporter activities from Caco-2 cells were used to identify function-abundance relationships, to facilitate the potential development of abundance-function scaling factors. Results: Development of a differential centrifugation technique to obtain plasma membranes was undertaken using MDCK-II and Caco-2 cells. The plasma membrane fraction showed little enrichment from the preceding total membrane fraction and was contaminated with endoplasmic reticulum, as assessed by marker enzyme activities. There were also no differences in Na/K-ATPase, BCRP and P-gp abundances between plasma and total membrane fractions in Caco-2 cells. This may be due to losses of protein from the target membrane fraction, thus, a theoretical framework combining protein assay (BCA) and transporter abundance determinations was proposed. Pilot data on the generation of recovery correction factors using Villin and Na/K-ATPase abundances, to account for protein losses is also presented. The abundances of 6 transporters in jejunal enterocyte membranes (n=3), including the key efflux proteins BCRP (2.56±0.82 fmol/μg), P-gp (1.89±1.07 fmol/μg) and MRP2 (0.59±0.246 fmol/μg) were determined with precision. In addition, peptide losses during protein digestion stages were accounted for in abundance determinations. A cross laboratory comparison of transporter abundances from intestinal (n=4) and Caco-2 cells (n=7) measured in our laboratory and Bertin Pharma (BPh), showed that P-gp abundances were highly correlated (rs=0.72), yet BPh abundances were systematically lower than determined in our laboratory (2.0±2.08 versus. 4.8±3.51 fmol/μg, respectively). No differences or correlations were found for Na/K-ATPase and BCRP abundances between laboratories. A jejunal-Caco-2 cell relative expression factor (REF) for each protein for both laboratories was generated. The P-gp REF was similar for BPh and our laboratory (0.37 vs. 0.4, respectively) however, for BCRP there was a distinct difference (1.11 versus 2.22, respectively). These findings provide the first evidence that employing expression factors generated from abundances quantified in different laboratories may produce altered IVIVE-PBPK outcomes. Functional studies in Caco-2 cells using E-3-S and vinblastine as probes for BCRP and P-gp, respectively, show that protein abundance is more closely correlated to transporter activity than mRNA expression. In addition, it was only possible to verify that increasing P-gp abundances in Caco-2 cells were ranked alongside vinblastine intrinsic clearance, as there was little consistency when estimating Km between the different Caco-2 cell models expressing increasing P-gp abundances, which may be attributed to limited absorptive transport saturation. Thus, forming any conclusions with confidence on concentration dependent abundance-activity relationships was difficult. These data suggest the value of REF scaling factors based on protein abundances, but emphasises the need to generate these from both in vitro and in vivo samples, using the same proteomic workflow. Further work to verify abundance-function relationships is required. Conclusion: A targeted proteomic workflow has been developed allowing quantification of protein abundances for key drug transporters in human gut tissues and cell models. The study has highlighted important areas including losses of targeted proteins, contamination of plasma membrane fractions and standardisation between laboratories that need to be addressed before implementation of transporter abundances into PBPK models is undertaken. Nevertheless, the evidence for a close relationship between transporter abundance and function indicate the potential value of this data for generation of robust REF scaling factors.

Many methods exist in the literature for the prediction of pharmacokinetic parameters which describe drug distribution in rat and human, such as tissue-to-plasma partition coefficients (Kps) and volume of distribution (Vss). However, none of these methods make use of the in vivo information obtained at the early stages of the drug development process in the form of plasma concentration vs. time profiles. The overall aim of the presented study was to improve upon an existing Kp prediction method by making use of the distribution information contained within this experimental data. Chapter 2 shows that Kp values can be successfully obtained experimentally, but that this process is expensive and time-consuming. Chapter 3 compares six Kp prediction methods taken from the literature for their ability to predict the Kp values of 80 drugs. The Rodgers et al. model was found to be the most accurate, with over 77% of predictions within 3-fold of experimental values. This Chapter also discusses the Vss prediction ability of some of these methods, with the Poulin & Theil and Rodgers et al. models shown to be the most accurate predictors for rat Vss and human Vss respectively. Chapter 4 investigates the relationship between muscle Kp and the Kps of all other tissues, to show that experimental muscle Kp can be used as a surrogate from which all other non-adipose Kp values can be predicted. However, the predictions made using this method were shown to be less accurate than predictions made by the Rodgers et al. model for the same dataset of drugs. A relationship was identified between muscle Kp and tumour Kp in rat, suggesting a potential way to predict tumour Kp in the future. In Chapter 5, a novel method is developed whereby Kp predictions made by the Rodgers et al. model are updated using prior information obtained from the in vivo concentration-time profile. These updated values are then used within a physiologically-based pharmacokinetic (PBPK) model and are shown in Chapter 6 to generate improved predictions for other pharmacokinetic parameters such as Vss and clearance in both rat and human. 100% of human Vss predictions made by the most accurate of the novel methods presented here were within 3-fold of experimental values, compared to 68.8% of predictions made by the Rodgers et al. model. The work presented here has highlighted the need for a more accurate method for the prediction of Kp values, and has addressed this need by generating a model which improves upon the most accurate Kp prediction method currently found in the literature. This will lead to an increase in confidence in the use of predicted pharmacokinetic parameters within PBPK modelling.

No description available.

Modeling and simulation are becoming indispensable elements in drug development. Mechanistic modeling is susceptible to impacting several drug development and regulatory decisions through extrapolations beyond clinically researched scenarios by using its capacity to incorporate diverse data to provide a detailed and comprehensive understanding of a physiological and biological system. Physiologically based pharmacokinetic (PBPK) modeling is the fastest-growing Model-informed drug development subfield, and regulatory authorities currently demand or advocate its usage for specific applications. Shiny is based on the open-source software R, which is widely used in the statistics and data science communities, including finance, medicine, and pharmaceuticals. Shiny conveys complex models to non-technical audiences via interactive graphics and sensitivity analysis. The Shiny R package is a crucial development that opens up PBPK models developed in R to a wider audience. This project's main goal is to create a framework and, a user-friendly tool for PBPK model evaluations and performance comparisons during model development, customized for the Open System Pharmacology Suite. To develop the PBPK framework, a PBPK model was retrieved from the open system pharmacology repository on GitHub, and based on the model and parameter identification inputs, the shiny framework was coded in R.Principal diagnostic techniques such as visual predictive checks to match and correlate model-simulated concentration-time profiles with clinical data, the goodness of fit (GOF) analysis (e.g., residuals over time, residuals against predictions, etc.), and extensive precision and bias measures were used to assess and validate the performance robustness of the model. Various quantitative metrics such as root mean square error, mean absolute error, and concordance correlation coefficient were used to assess and evaluate the model performance. The PBPK-QSP shiny framework was developed, allowing users to do the model evaluation with minimal effort. Using frameworks such as Shiny can expedite and automate the PBPK procedure, saving a significant amount of money and time in the evaluation of model performance.

No description available.