Radiation in Combination with Immune Checkpoint Blockade and DNA Damage Response Inhibitors in Mice: Dosage Optimization in MC38 Syngeneic Tumors via Modelling and Simulation.

Clinical trials assessing the impact of radiotherapy (RT) in combination with DNA damage response pathway inhibitors (DDRis) and/or immune checkpoint blockade are currently ongoing. However, current methods for optimizing dosage and schedule are limited. A mathematical model was developed to capture the impacts of RT in combination with DDRi and/or anti-PD-L1 [immune checkpoint inhibitor (ICI)] on tumor immune interactions in the MC38 syngeneic tumor model. The model was fitted to datasets that assessed the impact of RT in combination with the DNA protein kinase inhibitor (DNAPKi) AZD7648. The model was further fitted to datasets from studies that were used to assess both RT/ICI combinations as well as RT/ICI combinations followed by concurrent administration of the poly ADP ribose polymerase inhibitor (PARPi) olaparib. Nonlinear mixed-effects modeling was performed followed by internal validation with visual predictive checks (VPC). Simulations of alternative dosage regimens and scheduling were performed to identify optimal candidate dosage regimens of RT/DNAPKi and RT/PARPi/ICI. Model fits and VPCs confirmed a successful internal validation for both datasets and demonstrated very small differences in the median, lower, and upper percentile values of tumor diameters between RT/ICI and RT/PARPi/ICI, which indicated that the triple combination of RT/PARPi/ICI at the given dosage and schedule does not provide additional benefit compared with ICI in combination with RT. Simulation of alternative dosage regimens indicated that lowering the dosage of ICI to between 2 and 4 mg/kg could induce similar benefits to the full dosage regimen, which could be of translational benefit. SIGNIFICANCE STATEMENT: This work provides a mixed-effects model framework to quantify the effects of combination radiotherapy/DNA damage response pathway inhibitors/immune checkpoint inhibitors in preclinical tumor models and identify optimal dosage regimens, which could be of translational benefit.

Evaluation of area under the concentration curve adjusted by the terminal-phase as a metric to reduce the impact of variability in bioequivalence testing.

AIM: To quantify the utility of a terminal-phase adjusted area under the concentration curve method in increasing the probability of a correct and conclusive outcome of a bioequivalence (BE) trial for highly variable drugs when clearance (CL) varies more than the volume of distribution (V). METHODS: Data from a large population of subjects were generated with variability in CL and V, and used to simulate a two-period, two-sequence crossover BE trial. The 90% confidence interval for formulation comparison was determined following BE assessment using the area under the concentration curve (AUC) ratio test, and the proposed terminal-phase adjusted AUC ratio test. An outcome of bioequivalent, nonbioequivalent or inconclusive was then assigned according to predefined BE limits. RESULTS: When CL varied more than V, the proposed approach enhanced the probability of correctly assigning bioequivalent or nonbioequivalent and reduced the risk of an inconclusive trial. For a hypothetical drug with between-subject variability of 35% for CL and 10% for V, when the true test-reference ratio of bioavailability was 1.15, a crossover study of n = 14 subjects analysed by the proposed method would have 80% or 20% probability of claiming bioequivalent or nonbioequivalent, compared to 22%, 46% or 32% probability of claiming bioequivalent, nonbioequivalent or inconclusive using the standard AUC ratio test. CONCLUSIONS: The terminal-phase adjusted AUC ratio test represents a simple and readily applicable approach to enhance the BE assessment of drug products when CL varies more than V.

Dose individualisation in oncology using chemotherapy-induced neutropenia: Example of docetaxel in non-small cell lung cancer patients.

AIMS: Chemotherapy-induced neutropenia has been associated with an increase in overall survival in non-small cell lung cancer patients. Therefore, neutrophil counts could be an interesting biomarker for drug efficacy as well as linked directly to toxicity. For drugs where neutropenia is dose limiting, neutrophil counts might be used for monitoring drug effect and for dosing optimisation. METHODS: The relationship between drug effect on the first cycle neutrophil counts and patient survival was explored in a Phase III clinical trial where patients with non-small cell lung cancer were treated with docetaxel. Once the association has been established, dosing optimisation was performed for patients with severe toxicities (neutropenia) without compromising drug efficacy (overall survival). RESULTS: After taking into account baseline prognostic factors, such as Eastern Cooperative Oncology Group performance status, smoking status, liver metastasis, tumour burden, neutrophil counts and albumin levels, a model-predicted drug effect on the first cycle neutrophil counts was strongly associated with patient survival (P = .005). Utilising this relationship in a dose optimisation algorithm, 194 out of 366 patients would have benefited from a dose reduction after the first cycle of docetaxel. The simulated 1-year survival probabilities associated with the original dose and the individualised dose were not different. CONCLUSION: The strong relationship between drug effect on the first cycle neutrophil counts and patient survival suggests that this variable could be used to individualise dosing, possibly without needing pharmacokinetic samples. The algorithm highlights that doses could be reduced in case of severe haematological toxicities without compromising drug efficacy.

Intravitreal Pharmacokinetic Study of the Antiangiogenic Glycoprotein Opticin.

Opticin is an endogenous vitreous glycoprotein that may have therapeutic potential as it has been shown that supranormal concentrations suppress preretinal neovascularization. Herein we investigated the pharmacokinetics of opticin following intravitreal injection in rabbits. To measure simultaneously concentrations of human and rabbit opticin, a selected reaction monitoring mass spectrometry assay was developed. The mean concentration of endogenous rabbit opticin in 7 uninjected eyes was measured and found to be 19.2 nM or 0.62 µg/mL. When the vitreous was separated by centrifugation into a supernatant and collagen-containing pellet, 94% of the rabbit opticin was in the supernatant. Intravitreal injection of human opticin (40 µg) into both eyes of rabbits was followed by enucleation at 5, 24, and 72 h and 7, 14, and 28 days postinjection (n = 6 at each time point) and measurement of vitreous human and rabbit opticin concentrations in the supernatant and collagen-containing pellet following centrifugation. The volume of distribution of human opticin was calculated to be 3.31 mL, and the vitreous half-life was 4.2 days. Assuming that rabbit and human opticin are cleared from rabbit vitreous at the same rate, opticin is secreted into the vitreous at a rate of 0.14 µg/day. We conclude that intravitreally injected opticin has a vitreous half-life that is similar to currently available antiangiogenic therapeutics. While opticin was first identified bound to vitreous collagen fibrils, here we demonstrate that >90% of endogenous opticin is not bound to collagen. Endogenous opticin is secreted by the nonpigmented ciliary epithelium into the rabbit vitreous at a remarkably high rate, and the turnover in vitreous is approximately 15% per day.

A study of the dosage and duration for levobupivacaine infusion by the caudal-epidural route in infants aged 3-6 months.

BACKGROUND: The local anesthetic, levobupivacaine, is the safer enantiomer of racemic bupivacaine. Present protocols for levobupivacaine are based on studies and pharmacokinetic modeling with racemic bupivacaine. AIMS: The aim is to investigate total serum levobupivacaine concentrations after a caudalepidural loading dose followed by a maintenance infusion over 48 hours in infants aged 3-6 months. METHODS: The clinical trial was conducted in eight infants aged 3-6 months, undergoing bladder exstrophy repair. Pharmacokinetic modeling allowed optimization of clinical sampling to measure total levobupivacaine and α1 -acid glycoprotein and prediction of the effect of α1 -acid glycoprotein on levobupivacaine plasma protein binding. RESULTS: The observed median total levobupivacaine serum concentration was 0.30 mg/L (range: 0.20-0.70 mg/L) at 1 hour after the loading dose of 2 mg/kg. The median total levobupivacaine concentration after 47 hours of infusion, at 0.2 mg/kg/h, was 1.21 mg/L (0.07-1.85 mg/L). Concentrations of α1 -acid glycoprotein were found to rise throughout the study period. Pharmacokinetic modeling suggested that unbound levobupivacaine quickly reached steady state at a concentration of approximately 0.03 mg/L. CONCLUSION: The study allows the development of a pharmacokinetic model, combining levobupivacaine and α1 -acid glycoprotein data. Modeling indicates that unbound levobupivacaine quickly reaches steady state once the infusion is started. Simulations suggest that it may be possible to continue the infusion beyond 48 hours.

Variance based global sensitivity analysis of physiologically based pharmacokinetic absorption models for BCS I-IV drugs.

Regulatory agencies have a strong interest in sensitivity analysis for the evaluation of physiologically-based pharmacokinetic (PBPK) models used in pharmaceutical research and drug development and regulatory submissions. One of the applications of PBPK is the prediction of fraction absorbed and bioavailability for drugs following oral administration. In this context, we performed a variance based global sensitivity analysis (GSA) on in-house PBPK models for drug absorption, with the aim of identifying key parameters that influence the predictions of the fraction absorbed and the bioavailability for neutral, acidic and basic compounds. This analysis was done for four different classes of drugs, defined according to the Biopharmaceutics Classification System, differentiating compounds by permeability and solubility. For class I compounds (highly permeable, highly soluble), the parameters that mainly influence the fraction absorbed are related to the formulation properties, for class II compounds (highly permeable, lowly soluble) to the dissolution process, for class III (lowly permeable, highly soluble) to both absorption process and formulation properties and for class IV (lowly permeable, lowly soluble) to both absorption and dissolution processes. Considering the bioavailability, the results are similar to those for the fraction absorbed, with the addition that parameters related to gut wall and liver clearance influence as well the predictions. This work aimed to give a demonstration of the GSA methodology and highlight its importance in improving our understanding of PBPK absorption models and in guiding the choice of parameters that can safely be assumed, estimated or require data generation to allow informed model prediction.

Population pharmacokinetics of the MEK inhibitor selumetinib and its active N-desmethyl metabolite: data from 10 phase I trials.

AIMS: The aims of the study were to characterize the pharmacokinetics (PK) of selumetinib (AZD6244; ARRY-142886), a mitogen-activated protein kinase kinase (MEK) 1/2 inhibitor in clinical development for various indications, and its N-desmethyl metabolite in healthy volunteers, and evaluate clinically important covariates. METHODS: A pooled-population PK analysis was performed using a nonlinear mixed-effects approach with plasma concentration data from 346 subjects who received single oral doses of selumetinib 20-75 mg across 10 phase I studies. Absolute bioavailability was determined using intravenous [14 C] selumetinib. RESULTS: A two-compartment linear model with sequential zero-first order absorption and a lag time for the zero-order process was described for selumetinib PK. N-desmethyl metabolite disposition was described by a single compartment with linear elimination, without back transformation. The parent-only and joint models generally described pooled data adequately. For the median subject, not taking interacting drugs, estimates for clearance (CL) and central volume of distribution (V2) for selumetinib in the final joint model were 12.7 l h-1 and 35.6 l, respectively. Food effects, comedication with itraconazole [a cytochrome P450 (CYP) 3A4 inhibitor], fluconazole (a CYP2C19 inhibitor) and rifampicin (a CYP3A4 inducer) and formulation effects were incorporated into the base model a priori. Race and hepatic function were also influential in the PK model. Additional covariates affecting selumetinib disposition identified from covariate analysis were age on V2, bilirubin on CL, and weight on CL and V2. CONCLUSIONS: Analysis confirmed previous clinical pharmacology study findings of drug-drug interactions and food effects, with additional covariates that influence selumetinib and N-desmethyl selumetinib PK identified. Dose modifications based on these additional covariates were not considered necessary.

Establishment and validation of microsampling techniques in wild rodents for ecotoxicological research.

Compounds and products in the biocide and plant protection sector can only be registered after formal risk assessment to ensure safety for users and the environment. In bird and mammal risk assessment, this is routinely done using generic focal species as models, which are of particular exposure risk. Such a species is the common vole (Microtus arvalis) due to its high food intake relative to the low body weight. For wild species, biological samples, data and hence realistic exposure estimations are particularly difficult to obtain. In recent years, advances have been made in the techniques related to serial microsampling of laboratory mice and rats that allow for a reduction in sampling volumes. Similar progress in wild species sampling is missing. This study presents a proof of concept to dose wild rodents with relevant compounds and to draw serial, low volume blood samples suitable for state-of-the art toxicokinetic analyses. For the first time, the jugular vein of common voles was used to administer compounds (two frequently used fungicidal components). This procedure and the following microsampling of blood (2 × 10 µl six times within 24 hours) from the lateral tail vein did not affect body weight and mortality of voles. Samples were sufficient to detect dissipation patterns of the compounds from blood in toxicokinetic analysis. These results suggest that microsampling can be well translated from laboratory mice to wild rodent species and help to obtain realistic exposure estimates in wild rodents for ecotoxicological studies as well as to promote the 3R concept in studies with wild rodent species.

Modelling of atorvastatin pharmacokinetics and the identification of the effect of a BCRP polymorphism in the Japanese population.

AIM: Ethnicity plays a modulating role in atorvastatin pharmacokinetics (PK), with Asian patients reported to have higher exposure compared with Caucasians. Therefore, it is difficult to safely extrapolate atorvastatin PK data and models across ethnic groups. This work aims to develop a population PK model for atorvastatin and its pharmacologically active metabolites specifically for the Japanese population. Subsequently, it aimed to identify genetic polymorphisms affecting atorvastatin PK in this population. METHODS: Atorvastatin acid (ATA) and ortho-hydroxy-atorvastatin acid (o-OH-ATA) plasma concentrations, clinical/demographic characteristics and genotypes for 18 (3, 3, 1, 1, 7, 2 and 1 in the ABCB1, ABCG2, CYP3A4, CYP3A5, SLCO1B1, SLCO2B1 and PPARA genes, respectively) genetic polymorphisms were collected from 27 Japanese individuals (taking 10 mg atorvastatin once daily) and analysed using a population PK modelling approach. RESULTS: The population PK model developed (one-compartment for ATA linked through metabolite formation to an additional compartment describing the disposition of o-OH-ATA) accurately described the observed data and the associated population variability. Our analysis suggested that patients carrying one variant allele for the rs2622604 polymorphism (ABCG2) show a 55% (95% confidence interval: 16-131%) increase in atorvastatin oral bioavailability relative to the value in individuals without the variant allele. CONCLUSION: The current work reports the identification in the Japanese population of a BCRP polymorphism, not previously associated with the PK of any statin, that markedly increases ATA and o-OH-ATA exposure. The model developed may be of clinical importance to guide dosing recommendations tailored specifically for the Japanese.

What do we mean by identifiability in mixed effects models?

We discuss the question of model identifiability within the context of nonlinear mixed effects models. Although there has been extensive research in the area of fixed effects models, much less attention has been paid to random effects models. In this context we distinguish between theoretical identifiability, in which different parameter values lead to non-identical probability distributions, structural identifiability which concerns the algebraic properties of the structural model, and practical identifiability, whereby the model may be theoretically identifiable but the design of the experiment may make parameter estimation difficult and imprecise. We explore a number of pharmacokinetic models which are known to be non-identifiable at an individual level but can become identifiable at the population level if a number of specific assumptions on the probabilistic model hold. Essentially if the probabilistic models are different, even though the structural models are non-identifiable, then they will lead to different likelihoods. The findings are supported through simulations.

Physiologically based pharmacokinetic model for 6-mercpatopurine: exploring the role of genetic polymorphism in TPMT enzyme activity.

AIMS: To extend the physiologically based pharmacokinetic (PBPK) model developed for 6-mercaptopurine to account for intracellular metabolism and to explore the role of genetic polymorphism in the TPMT enzyme on the pharmacokinetics of 6-mercaptopurine. METHODS: The developed PBPK model was extended for 6-mercaptopurine to account for intracellular metabolism and genetic polymorphism in TPMT activity. System and drug specific parameters were obtained from the literature or estimated using plasma or intracellular red blood cell concentrations of 6-mercaptopurine and its metabolites. Age-dependent changes in parameters were implemented for scaling, and variability was also introduced for simulation. The model was validated using published data. RESULTS: The model was extended successfully. Parameter estimation and model predictions were satisfactory. Prediction of intracellular red blood cell concentrations of 6-thioguanine nucleotide for different TPMT phenotypes (in a clinical study that compared conventional and individualized dosing) showed results that were consistent with observed values and reported incidence of haematopoietic toxicity. Following conventional dosing, the predicted mean concentrations for homozygous and heterozygous variants, respectively, were about 10 times and two times the levels for wild-type. However, following individualized dosing, the mean concentration was around the same level for the three phenotypes despite different doses. CONCLUSIONS: The developed PBPK model has been extended for 6-mercaptopurine and can be used to predict plasma 6-mercaptopurine and tissue concentration of 6-mercaptopurine, 6-thioguanine nucleotide and 6-methylmercaptopurine ribonucleotide in adults and children. Predictions of reported data from clinical studies showed satisfactory results. The model may help to improve 6-mercaptopurine dosing, achieve better clinical outcome and reduce toxicity.

Combining the 'bottom up' and 'top down' approaches in pharmacokinetic modelling: fitting PBPK models to observed clinical data.

Pharmacokinetic models range from being entirely exploratory and empirical, to semi-mechanistic and ultimately complex physiologically based pharmacokinetic (PBPK) models. This choice is conditional on the modelling purpose as well as the amount and quality of the available data. The main advantage of PBPK models is that they can be used to extrapolate outside the studied population and experimental conditions. The trade-off for this advantage is a complex system of differential equations with a considerable number of model parameters. When these parameters cannot be informed from in vitro or in silico experiments they are usually optimized with respect to observed clinical data. Parameter estimation in complex models is a challenging task associated with many methodological issues which are discussed here with specific recommendations. Concepts such as structural and practical identifiability are described with regards to PBPK modelling and the value of experimental design and sensitivity analyses is sketched out. Parameter estimation approaches are discussed, while we also highlight the importance of not neglecting the covariance structure between model parameters and the uncertainty and population variability that is associated with them. Finally the possibility of using model order reduction techniques and minimal semi-mechanistic models that retain the physiological-mechanistic nature only in the parts of the model which are relevant to the desired modelling purpose is emphasized. Careful attention to all the above issues allows us to integrate successfully information from in vitro or in silico experiments together with information deriving from observed clinical data and develop mechanistically sound models with clinical relevance.

A physiologically based pharmacokinetic model for clobazam and stiripentol in adults and children.

PURPOSE: To develop a physiologically based pharmacokinetic model in adults and children for clobazam, its active metabolite norclobazam and stiripentol and to account for significant clinical interaction that has been reported when clobazam and stiripentol are co-administered. METHODS: A PBPK model with ten compartments was developed. An in vitro-in vivo extrapolation technique was used to scale clearance in children for clobazam and norclobazam and clearance parameters for stiripentol were obtained from fitting. Other drug and system parameters were obtained from the literature. RESULTS: The tissue/blood partition coefficients adequately predict observed volume of distribution for clobazam and stiripentol. In a clinical study in children where clobazam was administered alone and co-administered with stiripentol, the predicted and observed minimum concentration at steady state (mean and 95% confidence interval) during clobazam monotherapy were 0.19 (0.05-0.49 mg/L) and 0.20 (0.17-0.23 mg/L), respectively, and predicted and observed norclobazam concentrations were 0.49 (0.16-1.38 mg/L) and 0.95 (0.91-0.99 mg/L), respectively. From an interaction study with stiripentol the predicted stiripentol concentration was 10.12 (2.51-39.36 mg/L) and the observed concentration was 10.0 (8.3-11.7 mg/L); the predicted clobazam concentration was 0.29 (0.07-1.05 mg/L) and the observed concentration was 0.31 (0.24-0.38 mg/L); and the predicted norclobazam concentration was 2.30 (0.45-5.53 mg/L) and the observed concentration was 4.32 (3.77-4.87 mg/L). CONCLUSIONS: The PBPK model adequately described observed data and the extent of interaction between clobazam/norclobazam and stiripentol.

Model-based evaluation of the impact of formulation and food intake on the complex oral absorption of mavoglurant in healthy subjects.

PURPOSE: To compare the pharmacokinetics of intravenous (IV), oral immediate-release (IR) and oral modified-release (MR) formulations of mavoglurant in healthy subjects, and to assess the food effect on the MR formulation's input characteristics. METHODS: Plasma concentration-time data from two clinical studies in healthy volunteers were pooled and analysed using NONMEM®. Drug entry into the systemic circulation was modelled using a sum of inverse Gaussian (IG) functions as an input rate function, which was estimated specifically for each formulation and food state. RESULTS: Mavoglurant pharmacokinetics was best described by a two-compartment model with a sum of either two or three IG functions as input function. The mean absolute bioavailability from the MR formulation (0.387) was less than from the IR formulation (0.436). The MR formulation pharmacokinetics were significantly impacted by food: bioavailability was higher (0.508) and the input process was shorter (complete in approximately 36 versus 12 h for the fasted and fed states, respectively). CONCLUSIONS: Modelling and simulation of mavoglurant pharmacokinetics indicate that the MR formulation might provide a slightly lower steady-state concentration range with lower peaks (possibly better drug tolerance) than the IR formulation, and that the MR formulation's input properties strongly depend on the food conditions at drug administration.

Development and Application of a Mechanistic Pharmacokinetic Model for Simvastatin and its Active Metabolite Simvastatin Acid Using an Integrated Population PBPK Approach.

PURPOSE: To develop a population physiologically-based pharmacokinetic (PBPK) model for simvastatin (SV) and its active metabolite, simvastatin acid (SVA), that allows extrapolation and prediction of their concentration profiles in liver (efficacy) and muscle (toxicity). METHODS: SV/SVA plasma concentrations (34 healthy volunteers) were simultaneously analysed with NONMEM 7.2. The implemented mechanistic model has a complex compartmental structure allowing inter-conversion between SV and SVA in different tissues. Prior information for model parameters was extracted from different sources to construct appropriate prior distributions that support parameter estimation. The model was employed to provide predictions regarding the effects of a range of clinically important conditions on the SV and SVA disposition. RESULTS: The developed model offered a very good description of the available plasma SV/SVA data. It was also able to describe previously observed effects of an OATP1B1 polymorphism (c.521 T > C) and a range of drug-drug interactions (CYP inhibition) on SV/SVA plasma concentrations. The predicted SV/SVA liver and muscle tissue concentrations were in agreement with the clinically observed efficacy and toxicity outcomes of the investigated conditions. CONCLUSIONS: A mechanistically sound SV/SVA population model with clinical applications (e.g., assessment of drug-drug interaction and myopathy risk) was developed, illustrating the advantages of an integrated population PBPK approach.

Application of a Bayesian approach to physiological modelling of mavoglurant population pharmacokinetics.

Mavoglurant (MVG) is an antagonist at the metabotropic glutamate receptor-5 currently under clinical development at Novartis Pharma AG for the treatment of central nervous system diseases. The aim of this study was to develop and optimise a population whole-body physiologically-based pharmacokinetic (WBPBPK) model for MVG, to predict the impact of drug-drug interaction (DDI) and age on its pharmacokinetics. In a first step, the model was fitted to intravenous (IV) data from a clinical study in adults using a Bayesian approach. In a second step, the optimised model was used together with a mechanistic absorption model for exploratory Monte Carlo simulations. The ability of the model to predict MVG pharmacokinetics when orally co-administered with ketoconazole in adults or administered alone in 3-11 year-old children was evaluated using data from three other clinical studies. The population model provided a good description of both the median trend and variability in MVG plasma pharmacokinetics following IV administration in adults. The Bayesian approach offered a continuous flow of information from pre-clinical to clinical studies. Prediction of the DDI with ketoconazole was consistent with the results of a non-compartmental analysis of the clinical data (threefold increase in systemic exposure). Scaling of the WBPBPK model allowed reasonable extrapolation of MVG pharmacokinetics from adults to children. The model can be used to predict plasma and brain (target site) concentration-time profiles following oral administration of various immediate-release formulations of MVG alone or when co-administered with other drugs, in adults as well as in children.

Incorporation of stochastic variability in mechanistic population pharmacokinetic models: handling the physiological constraints using normal transformations.

The utilisation of physiologically-based pharmacokinetic models for the analysis of population data is an approach with progressively increasing impact. However, as we move from empirical to complex mechanistic model structures, incorporation of stochastic variability in model parameters can be challenging due to the physiological constraints that may arise. Here, we investigated the most common types of constraints faced in mechanistic pharmacokinetic modelling and explored techniques for handling them during a population data analysis. An efficient way to impose stochastic variability on the parameters of interest without neglecting the underlying physiological constraints is through the assumption that they follow a distribution with support and properties matching the underlying physiology. It was found that two distributions that arise through transformations of the normal, the logit-normal generalisation and the logistic-normal, are excellent for such an application as not only they can satisfy the physiological constraints but also offer high flexibility during characterisation of the parameters' distribution. The statistical properties and practical advantages/disadvantages of these distributions for such an application were clearly displayed in the context of different modelling examples. Finally, a simulation study clearly illustrated the practical gains of the utilisation of the described techniques, as omission of population variability in physiological systems parameters leads to a biased/misplaced stochastic model with mechanistically incorrect variance structure. The current methodological work aims to facilitate the use of mechanistic/physiologically-based models for the analysis of population pharmacokinetic clinical data.

Handling missing data in a duloxetine population pharmacokinetic/pharmacodynamic model - imputation methods and selection models.

PURPOSE: In pharmacokinetic (PK)/pharmacodynamic (PD) modelling and simulations (M&S), omitting dropouts can cause inaccuracies in parameter estimation and clinical trial simulations (CTS). This study examines the impact of different imputation methods for missing data on the interpretation of model results, as well as develops a selection model (where dropout and efficacy are jointly modelled) for use in CTS. METHODS: Missing data were imputed using single and multiple imputation and pattern mixtures methods for a previously reported duloxetine PK/PD model. The probability of dropout was described in the selection model and CTS was conducted with a hypothetical drug to examine the impact of dropout on trial results. RESULTS: The study completion rate was 75% and dropouts were not random. Model parameters obtained with different imputation methods were mostly within 40% (range 0 to 63%) compared to the model without dropouts. CTS showed 0.3 points lower median pain scores and 3% lower coefficient of variation over the 12-week simulations when dropout was included. CONCLUSIONS: Missing data had little impact on the original population PK/PD analyses. Sensitivity analyses for dropouts should be conducted in M&S exercises. The utility of selection models in CTS was explored via a hypothetical case study.

The use of ROC analysis for the qualitative prediction of human oral bioavailability from animal data.

PURPOSE: To develop and evaluate a tool for the qualitative prediction of human oral bioavailability (Fhuman) from animal oral bioavailability (Fanimal) data employing ROC analysis and to identify the optimal thresholds for such predictions. METHODS: A dataset of 184 compounds with known Fhuman and Fanimal in at least one species (mouse, rat, dog and non-human primates (NHP)) was employed. A binary classification model for Fhuman was built by setting a threshold for high/low Fhuman at 50%. The thresholds for high/low Fanimal were varied from 0 to 100 to generate the ROC curves. Optimal thresholds were derived from 'cost analysis' and the outcomes with respect to false negative and false positive predictions were analyzed against the BDDCS class distributions. RESULTS: We successfully built ROC curves for the combined dataset and per individual species. Optimal Fanimal thresholds were found to be 67% (mouse), 22% (rat), 58% (dog), 35% (NHP) and 47% (combined dataset). No significant trends were observed when sub-categorizing the outcomes by the BDDCS. CONCLUSIONS: Fanimal can predict high/low Fhuman with adequate sensitivity and specificity. This methodology and associated thresholds can be employed as part of decisions related to planning necessary studies during development of new drug candidates and lead selection.

Physiologically based pharmacokinetic modelling of methotrexate and 6-mercaptopurine in adults and children. Part 1: methotrexate.

Methotrexate is an antimetabolite and antifolate drug that is widely used in the treatment of malignancies and auto-immune disorders. In childhood acute lymphoblastic leukaemia, methotrexate is often combined with 6-mercaptopurine and both of them have been shown to be very effective for maintenance of remission. Large variability in the pharmacokinetics of methotrexate has led to increasing use of therapeutic drug monitoring in its clinical use to identify patients with high risk of toxicity and optimise clinical outcome. A physiologically based pharmacokinetic model was developed for methotrexate for oral and intravenous dosing and adults and paediatric use. The model has compartments for stomach, gut lumen, enterocyte, gut tissue, spleen, liver vascular, liver tissue, gall bladder, systemic plasma, red blood cells, kidney vascular, kidney tissue, skin, bone marrow, thymus, muscle and rest of body. A mechanistic model was also developed for the kidney to account for renal clearance of methotrexate via filtration and secretion. Variability on system and drug specific parameters was incorporated in the model to reflect observed clinical data and assuming the same pathways in adults and children, age-dependent changes in body size, organ volumes and plasma flows, the model was scaled to children. The model was developed successfully for adults and parameters such as net secretion clearance, biliary transit time and red blood cell distribution and binding parameters were estimated from published adult profiles. A relationship between fraction absorbed and dose using reported mean bioavailability data in the literature was also established. The model also incorporates non-linear binding in some tissues that has been described in the literature. Predictions using this model provide adequate description of observed plasma concentration data in adults and children. The model can be used to predict plasma and tissue concentrations of methotrexate following intravenous and oral dosing in adults and children and therefore help to improve clinical outcome.