The use of clinical irrelevance criteria in covariate model building with application to dofetilide pharmacokinetic data

Tutorial on model selection and validation of model input into precision dosing software for model-informed precision dosing

There has been rising interest in using model-informed precision dosing to provide personalized medicine to patients at the bedside. This methodology utilizes population pharmacokinetic models, measured drug concentrations from individual patients, pharmacodynamic biomarkers, and Bayesian estimation to estimate pharmacokinetic parameters and predict concentration-time profiles in individual patients. Using these individualized parameter estimates and simulated drug exposure, dosing recommendations can be generated to maximize target attainment to improve beneficial effect and minimize toxicity. However, the accuracy of the output from this evaluation is highly dependent on the population pharmacokinetic model selected. This tutorial provides a comprehensive approach to evaluating, selecting, and validating a model for input and implementation into a model-informed precision dosing program. A step-by-step outline to validate successful implementation into a precision dosing tool is described using the clinical software platforms Edsim++ and MwPharm++ as examples.

Population Pharmacokinetics of Tralokinumab in Adult Subjects With Moderate to Severe Atopic Dermatitis

Tralokinumab is the first biologic therapy for moderate-to-severe atopic dermatitis (AD) that specifically neutralizes interleukin-13 activity, a key driver of AD signs and symptoms. Tralokinumab is a human immunoglobulin G4 monoclonal antibody administered subcutaneously every 2 weeks (with possibility of maintenance dosing every 4 weeks). This population pharmacokinetic (PK) analysis aimed to identify sources of PK variability and relevant predictors of tralokinumab exposure in adults with moderate to severe AD. Nonlinear mixed-effect modeling, including covariate analysis, was used on a data set including 2561 subjects (AD, asthma, healthy) from 10 clinical trials. A 2-compartment model with first-order absorption and elimination adequately described the tralokinumab PK. Body weight was identified as a relevant predictor of tralokinumab exposure; other covariates including age, sex, race, ethnicity, disease type, AD severity, and renal and hepatic impairment were not. For body weight, the difference in exposure between the upper- and lower-weight quartiles in patients with AD was <2-fold, supporting the appropriateness of flat dosing (300 mg). Given the reduced exposure associated with higher body weight, coupled with the reduced exposure provided by dosing every 4 weeks, it is uncertain whether higher-weight patients will achieve sufficient exposure to maintain efficacy if dosed every 4 weeks instead of the standard every 2 weeks.

Population Pharmacokinetics of Vericiguat in Patients with Heart Failure with Reduced Ejection Fraction: An Integrated Analysis

Vericiguat, a novel stimulator of soluble guanylate cyclase (sGC), is indicated for the treatment of patients following a hospitalization for heart failure or need for outpatient IV diuretics, with symptomatic chronic heart failure and ejection fraction less than 45%. Pharmacokinetic (PK) data from the phase II trial SOCRATES-REDUCED (SOluble guanylate Cyclase stimulatoR in heArT failurE Study) and the phase III trial VICTORIA (Vericiguat Global Study in Patients with Heart Failure with Reduced Ejection Fraction) were used to characterize vericiguat PK. A total of 8,092 concentration records from 2,321 participants (362 from SOCRATES-REDUCED and 1,959 from VICTORIA) were utilized for the development of the population PK (PPK) model. The final PK model was a one-compartment model with first-order absorption and linear elimination. Baseline body weight and time-varying body weight were identified as statistically significant covariates affecting apparent clearance (CL/F) and volume of distribution (V_c ), respectively. Age, sex, race, bilirubin, estimated glomerular filtration rate (eGFR), and albumin did not affect vericiguat PK. Baseline disease-related factors, such as left ventricular ejection fraction, New York Heart Association (NYHA) class, and N-terminal pro B-type natriuretic peptide (NT-proBNP), also did not influence vericiguat PK. Since vericiguat is a titrated drug, the impact of vericiguat PK on the titration to and maintenance of the target dose in VICTORIA was assessed. The distribution of steady-state doses in VICTORIA was similar across CL/F quartiles, suggesting that the ability to reach and maintain dosing at the target 10 mg dose was not related to vericiguat exposure.

Population pharmacokinetics of azathioprine active metabolite in patients with inflammatory bowel disease and dosage regimens optimisation

Azathioprine is a first-line drug used to maintain the remission of inflammatory bowel disease (IBD). As a prodrug, azathioprine is metabolised to produce active 6-thioguanine nucleotides (6-TGN). There are large individual variations in the pharmacokinetics/pharmacodynamics of 6-TGN in patients with IBD. Here, we aimed to develop a model to quantitatively investigate factors that affect 6-TGN pharmacokinetics to formulate a dosage guideline for azathioprine. Data were collected prospectively from 100 adult patients with IBD who were receiving azathioprine. Patients were genotyped for two single-nucleotide polymorphisms (TPMT*3C c.719A > G and NUDT15 c.415C > T). Using high-performance liquid chromatography, we measured 156 steady-state trough concentrations of 6-TGN within the range 0.09 to 1.16 mg/L (ie 133-1733 pmol per 8 × 10⁸ RBC). The covariates analysed included sex, age, body-weight, laboratory tests and concomitant medications. A population pharmacokinetic model was established using "non-linear mixed-effects modelling" software and the "first-order conditional estimation method with interaction." Body-weight, TPMT*3C polymorphisms and co-therapy with mesalazine were found to be important factors influencing the clearance of 6-TGN. A dosage guideline for azathioprine was developed based on the PPK model that enables individualised azathioprine dosing in adult patients with different body-weights, TPMT*3C genotypes and co-administration with mesalazine.

Evaluating Lean Liver Volume as a Potential Scaler for In Vitro-In Vivo Extrapolation of Drug Clearance in Obesity Using the Model Drug Antipyrine

BACKGROUND

In vitro-in vivo extrapolation (IVIVE) of hepatic drug clearance (CL) involves the scaling of hepatic intrinsic clearance (CL_int,uH) by functional liver size, which is approximated by total liver volume (LV) as per the convention. However, in most overweight and obese patients, LV includes abnormal liver fat, which is not thought to contribute to drug elimination, thus overestimating drug CL. Therefore, lean liver volume (LLV) might be a more appropriate scaler of CL_int,uH.

OBJECTIVE

The objective of this work was to assess the application of LLV in CL extrapolation in overweight and obese patients (BMI >25 kg/m2) using a model drug antipyrine.

METHODS

Recently, a model to predict LLV from patient sex, weight, and height was developed and evaluated. In order to assess the LLV model's use in IVIVE, a correlation-based analysis was conducted using antipyrine as an example drug.

RESULTS

In the overweight group (BMI >25 kg/m2), LLV could describe 36% of the variation in antipyrine CL (R² = 0.36), which was >2-fold higher than that was explained by LV (R² = 0.17). In the normal-weight group (BMI ≤25 kg/m2), the coefficients of determination were 58% (R² = 0.58) and 43% (R²= 0.43) for LLV and LV, respectively.

CONCLUSION

The analysis indicates that LLV is potentially a more appropriate descriptor of functional liver size than LV, particularly in overweight individuals. Therefore, LLV has a potential application in IVIVE of CL in obesity.

Integrated Data Analysis of Six Clinical Studies Points Toward Model-Informed Precision Dosing of Tamoxifen

Introduction

At tamoxifen standard dosing, ∼20% of breast cancer patients do not reach proposed target endoxifen concentrations >5.97 ng/mL. Thus, better understanding the large interindividual variability in tamoxifen pharmacokinetics (PK) is crucial. By applying non-linear mixed-effects (NLME) modeling to a pooled ‘real-world’ clinical PK database, we aimed to (i) dissect several levels of variability and identify factors predictive for endoxifen exposure and (ii) assess different tamoxifen dosing strategies for their potential to increase the number of patients reaching target endoxifen concentrations.

Methods

Tamoxifen and endoxifen concentrations with genetic and demographic data of 468 breast cancer patients from six reported studies were used to develop a NLME parent-metabolite PK model. Different levels of variability on model parameters or measurements were investigated and the impact of covariates thereupon explored. The model was subsequently applied in a simulation-based comparison of three dosing strategies with increasing degree of dose individualization for a large virtual breast cancer population. Interindividual variability of endoxifen concentrations and the fraction of patients at risk for not reaching target concentrations were assessed for each dosing strategy.

Results and Conclusions

The integrated NLME model enabled to differentiate and quantify four levels of variability (interstudy, interindividual, interoccasion, and intraindividual). Strong influential factors, i.e., CYP2D6 activity score, drug–drug interactions with CYP3A and CYP2D6 inducers/inhibitors and age, were reliably identified, reducing interoccasion variability to <20% CV. Yet, unexplained interindividual variability in endoxifen formation remained large (47.2% CV). Hence, therapeutic drug monitoring seems promising for achieving endoxifen target concentrations. Three tamoxifen dosing strategies [standard dosing (20 mg QD), CYP2D6-guided dosing (20, 40, and 60 mg QD) and individual model-informed precision dosing (MIPD)] using three therapeutic drug monitoring samples (5–120 mg QD) were compared, leveraging the model. The proportion of patients at risk for not reaching target concentrations was 22.2% in standard dosing, 16.0% in CYP2D6-guided dosing and 7.19% in MIPD. While in CYP2D6-guided- and standard dosing interindividual variability in endoxifen concentrations was high (64.0% CV and 68.1% CV, respectively), it was considerably reduced in MIPD (24.0% CV). Hence, MIPD demonstrated to be the most promising strategy for achieving target endoxifen concentrations.

Development and validation of a UPLC-MS/MS analytical method for dofetilide in mouse plasma and urine, and its application to pharmacokinetic study

A novel method using UPLC with tandem mass-spectrometric detection (UPLC-MS/MS) with positive electrospray ionization was developed for the detection of the antiarrhythmic drug, dofetilide, in mouse plasma and urine. Protein precipitation was performed on 10 μL of plasma and 2 μL of urine samples using dofetilide-D4 as an internal standard, and separation of the analyte was accomplished on a C18 analytical column with the flow of 0.40 mL/min. Subsequently, the method was successfully applied to determine the pharmacokinetic parameters of dofetilide following oral and intravenous administration. The calibration curve was linear over the selected concentration range (R² ≥ 0.99), with a lower limit of quantitation of 5 ng/mL. The intra-day and inter-day precisions, and accuracies obtained from a 5-day validation ranged from 3.00 to 7.10%, 3.80-7.20%, and 93.0-106% for plasma, and 3.50-9.00%, 3.70-10.0%, 87.0-106% for urine, while the recovery of dofetilide was 93.7% and 97.4% in plasma and urine, respectively. The observed pharmacokinetic profiles revealed that absorption is the rate-limiting step in dofetilide distribution and elimination. Pharmacokinetic studies illustrate that the absolute bioavailability of dofetilide in the FVB strain mice is 34.5%. The current developed method allows for accurate and precise quantification of dofetilide in micro-volumes of plasma and urine, and was found to be suitable for supporting in vivo pharmacokinetic studies.

Population pharmacokinetic and covariate analyses of intravenous trastuzumab (Herceptin®), a HER2-targeted monoclonal antibody, in patients with a variety of solid tumors

PURPOSE

The aim of the study was to characterize the population pharmacokinetics (PK) of the intravenous formulation of trastuzumab, assess the impact of patient and pathological covariates on trastuzumab PK, and perform simulations to support dosing recommendations in special situations.

METHODS

Serum trastuzumab concentrations were obtained from 1582 patients with metastatic breast cancer (MBC), early breast cancer (EBC), advanced gastric cancer (AGC), or other tumor types/healthy volunteers in 18 phase I, II, and III trials and analyzed by nonlinear mixed-effects modeling.

RESULTS

A two-compartment model with parallel linear and nonlinear elimination best described the data. During treatment, linear clearance (CL) dominated, resulting in a total CL of 0.173-0.337 L/day, which is similar to other IgG1 monoclonal antibodies. Covariates influencing CL were baseline body weight, aspartate aminotransferase, albumin, gastric cancer, and the presence of liver metastases. MBC and EBC had similar PK parameters, while CL was higher in AGC. Simulations indicated that at least 95% of patients with BC reach concentrations < 1 µg/mL (~ 97% washout) by 7 months. A dose delay in BC or AGC patients of > 1 week would take approximately 6 weeks to get back within steady-state exposure range.

CONCLUSIONS

Trastuzumab PK for the intravenous formulation was well-described across cancer types, disease status, and regimens. No dose adjustment is required for any of the identified patient covariates. A 7-month serum washout period for trastuzumab is recommended. A reloading dose is required if a maintenance dose is missed by > 1 week.

Modelling and simulation of edoxaban exposure and response relationships in patients with atrial fibrillation

Edoxaban is a novel, orally available, highly specific direct inhibitor of factor Xa and is currently being developed for the treatment and prevention of venous thromboembolism and prevention of stroke and systemic embolism in patients with non-valvular atrial fibrillation (NVAF). The objectives of the present analyses were to characterise edoxaban population pharmacokinetics (PPK) and identify potential intrinsic and extrinsic factors affecting variability in edoxaban exposure, determine if there are relationships between edoxaban pharmacokinetics or biomarkers and the risk of bleeding in patients with NVAF using an exposure-response model, and to use the PPK and exposure-response model to support dose selection for a phase III trial of edoxaban in patients with NVAF. PPK analysis of data from 1,281 edoxaban-dosed subjects with intrinsic factors such as renal impairment or NVAF and extrinsic factors such as concomitant medications revealed significant effects of renal impairment and concomitant strong P-glycoprotein (P-gp) inhibitors on the pharmacokinetics of edoxaban. Exposure-response analysis found that in patients with NVAF, the incidence of bleeding events increased significantly with increasing edoxaban exposure, with steady-state minimum concentration (Cmin,ss) showing the strongest association. Clinical trial simulations of bleeding incidence were used to select 30 mg and 60 mg once-daily edoxaban with 50% dose reductions for patients with moderate renal impairment or receiving concomitant strong P-gp inhibitors as the treatment regimens in the ENGAGE AF-TIMI 48 (NCT00781391) trial.

The results of this study were previously presented at the 2009 International Society on Thrombosis and Haemostasis, July 2009, Boston, Massachusetts, USA.

Population pharmacokinetics of lenalidomide in multiple myeloma patients

PURPOSE

Lenalidomide disease-specific toxicity profiles and potentially life-threatening adverse events support the consideration of diversity in starting doses. The aim of this study was to conduct a population pharmacokinetic analysis of lenalidomide in multiple myeloma patients to identify and evaluate non-studied covariates that could be used for dose individualization.

METHODS

Blood samples were collected from 15 multiple myeloma patients. Nonlinear mixed-effects modeling was used to develop a population pharmacokinetic model and perform covariate analysis. The developed model was used to simulate dose schedules in order to explore the need of different dosing regimens in patients with different covariate values.

RESULTS

The data were accurately described by a one-compartment model with first-order elimination. Absorption was best described using three transit compartments. Creatinine clearance and body surface area were identified as covariates affecting apparent clearance and apparent volume of distribution, respectively. Simulations revealed that lower starting doses than the standard 25 mg/daily could be used in patients with body surface area below 1.8 m² and even higher doses might be necessary for patients with normal renal function and large body surface area.

CONCLUSIONS

This study identified creatinine clearance and body surface area as covariates that have a clinically relevant impact on lenalidomide pharmacokinetics using population pharmacokinetics. In addition, the developed population pharmacokinetic model can be used to individualize lenalidomide dose in multiple myeloma patients, taking into account not only creatinine clearance but also body surface area.

Brivaracetam Population Pharmacokinetics and Exposure-Response Modeling in Adult Subjects With Partial-Onset Seizures

Brivaracetam is a selective high-affinity ligand for synaptic vesicle protein 2A, recently approved as adjunctive therapy in the treatment of partial-onset (focal) seizures in patients 16 years of age and older with epilepsy. A population pharmacokinetic (PK) model and a population pharmacokinetic/pharmacodynamic (PKPD) model were developed describing brivaracetam plasma concentration and the relationship with daily seizure counts in adequate well-controlled efficacy trials. The effect of body weight on clearance and volume was implemented using allometric scaling, and a range of covariates were investigated for their influence on brivaracetam clearance. The PKPD model described daily seizure counts using a negative binomial distribution, taking previous day seizures into account, and using a mixture model to separate "placebo-like" and "response" subpopulations. The PK and PKPD models provided a good description of the data, documented using visual predictive checks. Coadministration with carbamazepine, phenytoin, and phenobarbital decreased brivaracetam exposure by 26%, 21%, and 19%, respectively, without significant effects on PD response. Covariate analysis indicated that levetiracetam coadministration reduced the fraction of subjects in the mixture model response population to 4% and identified baseline seizure frequency as a strong predictor for being assigned to the mixture model response population. Simulation allowed characterization of the dose-response curve, suggesting maximum response is obtained at brivaracetam 150-200 mg/day.

Influence of donor-recipient CYP3A4/5 genotypes, age and fluconazole on tacrolimus pharmacokinetics in pediatric liver transplantation: a population approach

AIM

To characterize the effect of donor and recipient CYP3A4, CYP3A5 and ABCB1 genotypes as well as relevant patient characteristics on tacrolimus pharmacokinetics in pediatric liver transplantation.

PATIENTS & METHODS

Data from 114 pediatric liver transplant recipients were retrospectively collected during the first 3 months following transplantation. Population pharmacokinetic analysis was performed using nonlinear mixed effects modeling, including characterization of influential covariates.

RESULTS

A two-compartment model with first order elimination best fitted the data. Estimates of apparent volume of the central compartment, intestinal clearance, hepatic clearance and intercompartmental clearance were 79 l, 0.01 l/h, 10.9 l/h and 105 l/h, respectively. Time post-transplantation, recipient age, donor CYP3A5 and CYP3A4 genotypes and fluconazole administration significantly influenced tacrolimus apparent clearance while bodyweight influenced volume of distribution.

CONCLUSION

The proposed model displayed acceptable fitting performances and enabled identification of statistically significant and clinically relevant covariates on tacrolimus pharmacokinetics in the early pediatric post liver transplantation period.

Comparison of methods for handling missing covariate data

Missing covariate data is a common problem in nonlinear mixed effects modelling of clinical data. The aim of this study was to implement and compare methods for handling missing covariate data in nonlinear mixed effects modelling under different missing data mechanisms. Simulations generated data for 200 individuals with a 50% difference in clearance between males and females. Three different types of missing data mechanisms were simulated and information about sex was missing for 50% of the individuals. Six methods for handling the missing covariate were compared in a stochastic simulations and estimations study where 200 data sets were simulated. The methods were compared according to bias and precision of parameter estimates. Multiple imputation based on weight and response, full maximum likelihood modelling using information on weight and full maximum likelihood modelling where the proportion of males among the individuals lacking information about sex was estimated (EST) gave precise and unbiased estimates in the presence of missing data when data were missing completely at random or missing at random. When data were missing not at random, the only method resulting in low bias and high parameter precision was EST.

Multiple imputation of missing covariates in NONMEM and evaluation of the method's sensitivity to η-shrinkage

Multiple imputation (MI) is an approach widely used in statistical analysis of incomplete data. However, its application to missing data problems in nonlinear mixed-effects modelling is limited. The objective was to implement a four-step MI method for handling missing covariate data in NONMEM and to evaluate the method's sensitivity to η-shrinkage. Four steps were needed; (1) estimation of empirical Bayes estimates (EBEs) using a base model without the partly missing covariate, (2) a regression model for the covariate values given the EBEs from subjects with covariate information, (3) imputation of covariates using the regression model and (4) estimation of the population model. Steps (3) and (4) were repeated several times. The procedure was automated in PsN and is now available as the mimp functionality ( http://psn.sourceforge.net/ ). The method's sensitivity to shrinkage in EBEs was evaluated in a simulation study where the covariate was missing according to a missing at random type of missing data mechanism. The η-shrinkage was increased in steps from 4.5 to 54%. Two hundred datasets were simulated and analysed for each scenario. When shrinkage was low the MI method gave unbiased and precise estimates of all population parameters. With increased shrinkage the estimates became less precise but remained unbiased.

Population pharmacokinetics of laninamivir and its prodrug laninamivir octanoate in healthy subjects and in adult and pediatric patients with influenza virus infection

Laninamivir octanoate (LO) is a new neuraminidase inhibitor for inhalation. The objectives of this study were to model the population pharmacokinetics of LO and its active metabolite laninamivir after inhaled administration of LO using a pooled population of healthy subjects, and adult and pediatric patients with influenza virus infection from 8 clinical studies, and to evaluate covariate effects on pharmacokinetics. The pharmacokinetics of LO and laninamivir in plasma and urine are well-described by structural models that consist of a 2-compartment model for LO with instantaneous bolus input and first-order elimination; and a 1-compartment model for laninamivir with formation of laninamivir via the metabolic pathway from LO in systemic circulation, entry of laninamivir from the respiratory tract compartment, and linear elimination. Creatinine clearance was identified as a covariate of apparent total clearance for LO and renal clearances for LO and laninamivir, with the largest effect on laninamivir exposure. Body weight was identified to affect distribution volumes of LO and laninamivir and the metabolic clearance of LO; however there was no notable effect on exposures across the wide body weight range evaluated. The population pharmacokinetic model also provides insight into the likely kinetics of drug disposition in the respiratory tract following inhaled administration.

Characterisation of exposure versus response of edoxaban in patients undergoing total hip replacement surgery

Edoxaban is an oral direct factor Xa inhibitor approved for the prevention of venous thromboembolism (VTE) in Japan. The objectives of this analysis were to characterise the population pharmacokinetics (PK) of edoxaban and the relationships between edoxaban exposure and clinical outcomes in a phase IIb study of surgical patients following total hip replacement (THR). A total of 1,795 subjects from a phase IIb study, 10 phase I studies, and three phase IIa studies were included in the PK analysis. The exposure-response analysis included data from surgical patients assigned to edoxaban in the phase IIb study. Edoxaban disposition in healthy and post-surgical patients was well-described with a linear, two-compartment model. Creatinine clearance was significantly correlated with edoxaban clearance and the rate of oral absorption was affected by surgery. The probability of a post-operative VTE was significantly correlated with steady-state metrics of edoxaban exposure estimated for each subject by Bayesian post-hoc methods with age and gender being the significant and expected covariates. The incidence of bleeding was low in these studies and hence no exposure-response relationship could be identified. These analyses suggest that edoxaban has a predictable anticoagulant effect in this patient population leading to dose-proportional reduction in incidence of VTE with low incidence of bleeding.

Covariate pharmacokinetic model building in oncology and its potential clinical relevance

When modeling pharmacokinetic (PK) data, identifying covariates is important in explaining interindividual variability, and thus increasing the predictive value of the model. Nonlinear mixed-effects modeling with stepwise covariate modeling is frequently used to build structural covariate models, and the most commonly used software-NONMEM-provides estimations for the fixed-effect parameters (e.g., drug clearance), interindividual and residual unidentified random effects. The aim of covariate modeling is not only to find covariates that significantly influence the population PK parameters, but also to provide dosing recommendations for a certain drug under different conditions, e.g., organ dysfunction, combination chemotherapy. A true covariate is usually seen as one that carries unique information on a structural model parameter. Covariate models have improved our understanding of the pharmacology of many anticancer drugs, including busulfan or melphalan that are part of high-dose pretransplant treatments, the antifolate methotrexate whose elimination is strongly dependent on GFR and comedication, the taxanes and tyrosine kinase inhibitors, the latter being subject of cytochrome p450 3A4 (CYP3A4) associated metabolism. The purpose of this review article is to provide a tool to help understand population covariate analysis and their potential implications for the clinic. Accordingly, several population covariate models are listed, and their clinical relevance is discussed. The target audience of this article are clinical oncologists with a special interest in clinical and mathematical pharmacology.

Pharmacokinetic evaluation of the penetration of antituberculosis agents in rabbit pulmonary lesions

Standard antituberculosis (anti-TB) therapy requires the use of multiple drugs for a minimum of 6 months, with variable outcomes that are influenced by a number of microbiological, pathological, and clinical factors. This is despite the availability of antibiotics that have good activity against Mycobacterium tuberculosis in vitro and favorable pharmacokinetic profiles in plasma. However, little is known about the distribution of widely used antituberculous agents in the pulmonary lesions where the pathogen resides. The rabbit model of TB infection was used to explore the hypothesis that standard drugs have various abilities to penetrate lung tissue and lesions and that adequate drug levels are not consistently reached at the site of infection. Using noncompartmental and population pharmacokinetic approaches, we modeled the rate and extent of distribution of isoniazid, rifampin, pyrazinamide, and moxifloxacin in rabbit lung and lesions. Moxifloxacin reproducibly showed favorable partitioning into lung and granulomas, while the exposure of isoniazid, rifampin, and pyrazinamide in lesions was markedly lower than in plasma. The extent of penetration in lung and lesions followed different trends for each drug. All four agents distributed rapidly from plasma to tissue with equilibration half-lives of less than 1 min to an hour. The models adequately described the plasma concentrations and reasonably captured actual lesion concentrations. Though further refinement is needed to accurately predict the behavior of these drugs in human subjects, our results enable the integration of lesion-specific pharmacokinetic-pharmacodynamic (PK-PD) indices in clinical trial simulations and in in vitro PK-PD studies with M. tuberculosis.

A fast method for testing covariates in population PK/PD Models

The development of covariate models within the population modeling program like NONMEM is generally a time-consuming and non-trivial task. In this study, a fast procedure to approximate the change in objective function values of covariate-parameter models is presented and evaluated. The proposed method is a first-order conditional estimation (FOCE)-based linear approximation of the influence of covariates on the model predictions. Simulated and real datasets were used to compare this method with the conventional nonlinear mixed effect model using both first-order (FO) and FOCE approximations. The methods were mainly assessed in terms of difference in objective function values (ΔOFV) between base and covariate models. The FOCE linearization was superior to the FO linearization and showed a high degree of concordance with corresponding nonlinear models in ΔOFV. The linear and nonlinear FOCE models provided similar coefficient estimates and identified the same covariate-parameter relations as statistically significant or non-significant for the real and simulated datasets. The time required to fit tesaglitazar and docetaxel datasets with 4 and 15 parameter-covariate relations using the linearization method was 5.1 and 0.5 min compared with 152 and 34 h, respectively, with the nonlinear models. The FOCE linearization method allows for a fast estimation of covariate-parameter relations models with good concordance with the nonlinear models. This allows a more efficient model building and may allow the utilization of model building techniques that would otherwise be too time-consuming.