Population pharmacokinetic meta-analysis with efavirenz.

A population-based pharmacokinetic (PK) model has been developed for efavirenz based on 16 phase I studies. The combined data set consisted of 334 healthy volunteers, 2,907 efavirenz dose administrations and 9,342 measured plasma concentrations across a range of doses from 100-600 mg. The pharmacokinetic structural model was a 2-compartment model with first-order absorption with differentiation between single- and multiple-dose exposure to account for known hepatic cytochrome P450 induction of efavirenz metabolism. Model-building was performed on the index data set (66% of the total database), as a data-splitting technique was used to validate the final model using NONMEM. The final model confirmed the appropriateness of separate clearance terms for single and multiple dose administration (2.65 versus 10.2 l/h, respectively). Clearance increased with dose and frequency of administration. A lower clearance was predicted in Asians and Blacks relative to Caucasians. A slightly lower clearance was observed in females relative to males (9.08 compared to 10.2 l/h in males) and interactions on clearance due to co-administration of fluconazole, ritonavir, rifampin, indinavir and azithromycin were identified. The magnitudes of these effects were small and did not suggest dose adjustment in the various subpopulations. With little exception, these results agree with the findings from the non-compartmental analyses. The residual variability was 21% CV and the intersubject variation in CL/F and V/F was 48 and 85%, respectively. The phase I meta-analysis was able to substantiate the pharmacokinetic characteristics of efavirenz derived from the composite of individual well-defined studies. The model was deemed adequate for subsequent evaluation in HIV-infected patients. Covariates and outlier classes identified in this phase I meta-analysis were similarly identified in subsequent analyses of patient data.

Population pharmacokinetic-pharmacodynamic modeling of filgrastim (r-metHuG-CSF) in healthy volunteers.

The pharmacokinetic-pharmacodynamic (PK-PD) relationship of the granulopoietic effects of Filgrastim in healthy volunteers was characterized via a population approach. Healthy male volunteers were enrolled into a four-way crossover clinical trial. Subjects received four single doses of Filgrastim (375 and 750 micrograms i.v. and s.c.) with an intervening washout period of 7 days. Serum concentrations of Filgrastim were determined using an enzyme-linked immunosorbent assay. Absolute neutrophil count (ANC) was determined. Data analysis was performed using mixed-effects modeling as implemented in the NONMEM software package. The final PKPD model incorporates a two-compartment PK model with bisegmental absorption from the s.c. site, first-order and saturable elimination pathways, and an indirect PD model. A sigmoidal Emax model for the stimulation of ANC input rate (kin) was superior to the conventional Emax model (mean +/- SE: Emax = 12.7 +/- 1.7; EC50 = 4.72 +/- 0.72 ng/ml; Hill = 1.34 +/- 0.19). In addition, a time-variant scaling factor for ANC observations was introduced to account for the early transient depression of ANC after Filgrastim administration. The absolute bioavailability of subcutaneously administered Filgrastim was estimated to be 0.619 +/- 0.058 and 0.717 +/- 0.028 for 375 micrograms and 750 micrograms s.c. doses, respectively. The time profiles of concentration and ANC, as well as the concentration approximately ANC relationship of Filgrastim in healthy volunteers were well described by the developed population PK-PD model.

Time-variant increase in methylprednisolone clearance in patients with acute respiratory distress syndrome: a population pharmacokinetic study.

Methylprednisolone (MP) disposition was evaluated in 20 individuals who participated in an ongoing randomized, double-blind, placebo-controlled study designed to evaluate the efficacy of MP in the treatment of acute respiratory distress syndrome (ARDS). MP (1 mg/kg) was given as a loading infusion over 30 minutes followed by a 1 mg/kg/day continuous i.v. infusion. Patients were switched to oral MP upon restoration of oral intake. MP plasma concentrations (n = 110) were determined using a specific HPLC method. Population pharmacokinetic analysis was performed using nonlinear mixed-effects models, implemented in NONMEM, version V. MP plasma concentration data were described by a one-compartment open model with a time-dependent, non-linear increase in the clearance (CL) of MP during the course of therapy. Initial clearance of MP (CLo) in ARDS patients at the start of therapy increased to a maximal value (CLmax) after approximately 7 days. The estimate of CLmax was similar to the CL of MP in healthy individuals reported previously. Population mean estimates (+/- SE) of parameters in the model were as follows: CLo = 13.2 +/- 2.4 L/h, CLmax = 25.0 +/- 3.6 L/h, time of half-maximal increase in CL (T50) = 41.1 +/- 8.2 h, gamma (Hill coefficient) = 3.8 +/- 0.6, and volume of distribution (Vd) = 137 +/- 30.2 L. Disease progression indices and patient demographics were evaluated as covariates, and no significant correlation was found. Means (+/- SD) of plasma protein binding differed between healthy individuals (72% +/- 4%) and ARDS patients (46% +/- 11%) (p < 0.001). The pharmacokinetics of MP in ARDS patients has not been described previously.

Population pharmacokinetics and pharmacodynamics of sotalol in pediatric patients with supraventricular or ventricular tachyarrhythmia.

AIMS: To derive useful pharmacokinetic (PK) and pharmacodynamic (PD) information for guiding the clinical use of sotalol in pediatric patients with supraventricular (SVT) or ventricular tachyarrhythmia (VT). METHODS: Two studies were conducted in-patients with SVT or VT in the age range between birth and 12 years old. Both studies used an extemporaneously compounded formulation prepared from sotalol HCl tablets. In the PK study, following a single dose of 30 mg/m2 sotalol, extensive blood samples (n = 10) were taken. The PK-PD study used a dose escalation design with doses of 10, 30, and 70 mg/m2, each administered three times at 8-hr intervals without a washout. Six ECG recordings for determination of QT and RR were obtained prior to the initial dose of sotalol. Four blood samples were collected six ECG's were determined during the third interval at each dose level. Plasma concentrations of sotalol (C) were assayed by LC/MS/MS. The data analysis used NONMEM to obtain the population PK and PD parameter estimates. The individual PK and PD parameters were estimated with empirical Bayes methodology. RESULTS: A total of 611 C from 58 patients, 477 QTc and 499 RR measurements from 23 and 22 patients, respectively, were available for analysis. The PK of sotalol was best described by a linear two-compartment model. Oral clearance (CL/F) and volume of central compartment (Vc/F) were linearly correlated with body surface area (BSA), body weight or age. CL/F was also linearly correlated with creatinine clearance. The best predictor for both CL/F and Vc/F was BSA. The remaining intersubject coefficients of variation (CV's) in CL/F, and Vc/F were 21.6% and 20.3%, respectively. The relationship of QTc to C was adequately described by a linear model. The intersubject CV's in slope (SL) and intercept (E0) were 56.2 and 4.7%, respectively. The relationship of RR to C was also adequately described by a linear model in which the baseline RR and SL were related to age or BSA. The intersubject CV's for SL and E0 were 86.7 and 14.4%, respectively. CONCLUSIONS: BSA is the best predictor for the PK of sotalol. Both QTc and RR effects are linearly related to C. No covariates are found for the QTc-C relation, while the RR-C relation shows age or BSA dependency.

Balanced designs in longitudinal population pharmacokinetic studies.

A simulation study was performed using a balanced design to determine the sample size required for accurate and precise estimation of a parameter at a given level of intersubject variability in a longitudinal population pharmacokinetic study. A two-compartment model parameterized in terms of clearance (Cl), volumes of the central (V1) and peripheral (V2) compartments, and intercompartmental clearance (Q) with multiple intravenous bolus inputs was assumed. Six samples were obtained from each subject using the informative profile (block) randomized design. Variability (in terms of coefficient of variation, CV) in model parameters was varied between 30% and 100%, and residual variability was fixed at 15%. Sample sizes ranging from 30 to 1,000 subjects were studied, and a hundred replicate data sets were generated and analyzed with NONMEM for each sample size at each CV. A sample size of 30 was required for accurate and precise estimation of structural model parameters when CV < or = 75%, except for Cl where it is adequate for CV < or = 100%. A sample size of 80 was required for intersubject variability estimation with CV < or = 60%. Robust estimates of variability in Cl were obtained with sample sizes of 30 (CV < or = 45%), 60 (CV 60-75%), and 100 (CV > or = 75%). Positively biased estimates of residual variability were obtained irrespective of sample size at > or = 60% CV. This indicates that estimates of residual variability obtained in study situations where CV > or = 60% should be interpreted with caution. In such situations model misspecification may not be the issue, because in this simulation study concentration-time profiles were generated and analyzed with the same model. Although these results should be interpreted within the context of the study, they provide a framework for addressing the issue of sample size in longitudinal population pharmacokinetic study with a balanced sampling design. The result of a population pharmacokinetic study can be anticipated by comparing the results of several simulations in which the various input factors have been varied.

Ignorability and parameter estimation in longitudinal pharmacokinetic studies.

In the analysis of longitudinal pharmacokinetic data, both balanced (equal number of samples per subject) and unbalanced data are used. It is implicitly assumed that the process that caused the missing data can be ignored. A simulation study was performed to determine the effect of ignoring the missing data (i.e., "ignorability") on the accuracy and precision of parameter estimation in longitudinal pharmacokinetic studies. A two-compartment model with multiple intravenous bolus inputs was assumed. Subjects with balanced data sets had six samples, and those with unbalanced data had 1 to 5 samples missing (i.e., supplied in a decreasing order from 5 to 1 samples). The proportion of subjects with 1 to 5 samples missing varied from 25% to 75% in a fixed sample size of 100. The effect of ignorability was studied at intersubject variability ranging from 15% to 60% for a drug assumed to be dosed at its elimination half-life. One hundred replicate data sets of 100 subjects each were simulated for each missing data scenario. The accuracy of parameter estimation was not significantly affected by the amount of ignorable missing data at any given level of variability. However, the precision of parameter estimation was affected by the degree of "missingness."

Pharmacokinetic model of ascorbic acid in healthy male volunteers during depletion and repletion.

PURPOSE: To develop a new pharmacokinetic model for ascorbic acid (vitamin C) since no previously published model describes ascorbic acid absorption and disposition over a broad physiologic range of doses and plasma concentrations. METHODS: A new model was developed through exploratory simulations. The model was fitted to pharmacokinetic data obtained from seven healthy volunteers who underwent ascorbic acid depletion then gradual repletion. Concentrations of ascorbic acid were measured in plasma and urine. Final pharmacokinetic model parameter estimates were obtained using nonlinear regression analysis. RESULTS: The new model included saturable absorption, distribution and renal tubular reabsorption parameters. The model described ascorbic acid concentrations in plasma, cells, and urine during depletion and gradual repletion phases with a residual error less than 15%. CONCLUSIONS: The model was useful for obtaining a new understanding of the likely causes for the complex concentration-time profile observed during gradual repletion. At doses of 200 to 2500 mg per day, the plateau in pre-dose concentrations is largely due to apparent saturation of tissue uptake and less a function of oral bioavailability and renal excretion than previously thought.

Effect of albumin on the estimation, in vitro, of phenytoin Vmax and Km values: implications for clinical correlation.

The effect of bovine serum albumin (BSA) on human liver metabolism, in vitro, of 14C-phenytoin (PHT) was studied. Michaelis Menten parameters were determined for the conversion of PHT to p-hydroxy phenytoin in seven different microsomal preparations with the addition of 0, 2, and 4% BSA. The unbound Km (Kmu) values were 30.8 +/- 18.6, 1.57 +/- 0.21 and 1.50 +/- 0.17 microM (mean +/- S.D.), respectively; however, there was excellent agreement among the Vmax values (29.1, 31.8 and 31.5 pmol/min/mg). With intact tissue slices, BSA (4%) added to incubations of PHT had a minimal effect on the Vmax values in two of the four livers studied and resulted in a mean Kmu value of 2.20 +/- 0.59 microM, although the Kmu in the absence of BSA was 6.64 +/- 3.17. In scaling-up to the whole body, Vmax values were 3.9 and 1.0 mg/kg/day for microsomes and slices, respectively, compared to 5.9 mg/kg/day, in vivo. The Kmu values determined in the presence of albumin in both microsomes and slices were similar to those based on in vivo human steady state data (Kmu = 2-3 microM), and the intersubject variation, in vitro, was decreased in the presence of BSA. These findings for phenytoin metabolism suggest that the addition of albumin to incubation media for slices or microsome experiments may yield Km estimates that are more representative of in vivo values.

Ocular penetration and pharmacokinetics of topical fluconazole.

The high bioavailability and low toxicity of fluconazole, a stable, water-soluble, low-molecular-weight bis-triazole antifungal, makes it a good candidate for consideration as a topical ocular agent. The penetration of fluconazole (0.2%) into the corneas and aqueous humors of New Zealand white rabbits was assayed by gas liquid chromatography (GLC). Peak corneal levels occurred essentially immediately at 5 min in the corneas [debrided, 8.2 +/- 1.2 micrograms/g; nondebrided, 1.6 +/- 0.6 microgram/g; (mean +/- SEM)] and at 15 min after application in the aqueous [debrided, 9.4 +/- 2.3 micrograms/ml; nondebrided, 1.6 +/- 0.6 microgram/ml; (mean +/- SEM)]. Estimating from semilogarithmic plots of the data, the halflife (t1/2) in the debrided eyes was 15 min; in the nondebrided eyes, t1/2 was 30 min. A loading dose of a 20-microliter drop per min for 5 min yielded levels of 59.9 +/- 11.3 micrograms/g (mean +/- SEM) in the debrided corneas and 32.4 +/- 1.9 micrograms/ ml (mean +/- SEM) in the corresponding aqueous humor. A regimen consisting of this loading dose followed by one 20 microliters drop/h for 6 h showed 45.9 +/- 3.5 micrograms/g (mean +/- SEM) in the debrided corneas and 8.8 +/- 1.7 micrograms/ml (mean +/- SEM) in the corresponding aqueous. The same regimen yielded values of 3.1 +/- 0.2 micrograms/g in the nondebrided corneas and 1.3 +/- 0.2 micrograms/ml (mean +/- SEM) in the aqueous. Minimal inhibitory concentrations (MIC) at 24 h for yeasts ranged from < 1.25 to 20 micrograms/ml, for hyaline molds from 2.5 to > 20 micrograms/ml, and dematiaceous molds from < 1.25 to > 20 micrograms/ml. Topical fluconazole exhibits pharmacokinetics and selective MICs that merit further evaluation for its ophthalmic use as a topical antifungal agent.

Ketotifen pharmacokinetics in children with atopic perennial asthma.

OBJECTIVE: Published pharmacokinetic data on ketotifen are sparse, although it is a commonly used prophylactic agent in various allergic disorders in adults and children. The aim of this study was to assess the steady-state pharmacokinetics of ketotifen in children with atopic perennial asthma who were participating in a clinical trial. METHOD: The NONMEM population approach with sparse sampling was utilized. The data set consisted of 239 samples from 48 children who were randomized to receive either 1 mg or 2 mg oral ketotifen daily. Patients underwent a clinical examination and had a blood sample taken at 2-week intervals for 12 weeks. The ketotifen concentrations were measured by RIA. RESULTS: A one-compartment model with first-order absorption was fit to the data. Volume was estimated at 394 l and clearance (CL) at 97.4 l.h-1 (3.6 l.h-1.kg-1). Weight or body surface area were the most influential covariates for explaining interindividual variability in CL. The 2-mg dose appeared to have a relative bio-availability of 85% of the 1-mg dose. CONCLUSION: Children have a faster clearance of ketotifen than adults and would therefore require a higher dose per kilogram body weight to give comparable steady-state levels.

On the recording of sample times and parameter estimation from repeated measures pharmacokinetic data.

A pharmacokinetic screen has been advocated for the characterization of the population pharmacokinetics of drugs during Phase 3 clinical trials. A common perception encountered in the collection of such data is that the accuracy of sampling times relative to dose is inadequate. A prospective simulation study was carried out to evaluate the effect of error in the recording of sampling times on the accuracy and precision of population parameter estimates from repeated measures pharmacokinetic data. A two-compartment model with intravenous bolus input(s) (single and multiple doses) was assumed. Random and systematic error in sampling times ranging from 5-50% using profile (block) randomized design were introduced. Sampling times were simulated in EXCEL while concentration data simulation and analysis were done in NONMEM. The effect of error in sampling times was studied at levels of variability ranging from 15-45% for a drug assumed to be dosed at its elimination half-life. One hundred replicate data sets of 100 subjects each were simulated for each case. Although estimates of clearance (CL) and variability in clearance were robust for most of the sampling time errors, there was an increase in bias and imprecision in overall parameter estimation as intersubject variability was increased. If there is interest in parameters other than CL, then the design of prospective population studies should include procedures for minimizing the error in the recording of sample times relative to dosing history.

Bioequivalence: individual and population compartmental modeling compared to the noncompartmental approach.

PURPOSE: The purpose of this study were to evaluate the use of individual compartmental and population compartmental methods for bioequivalence determination, and to determine their utility as adjuncts to the current methods used for bioequivalence assessment. METHODS: Data from three bioequivalence studies of chlorthalidone were analyzed with PCNONLIN using individual compartmental modeling and NONMEM for population analyses. These results were compared with results obtained from the traditional noncompartmental or SHAM (slopes, heights, areas, and moments) approach for bioequivalence assessment and the 90% confidence interval procedure. RESULTS: Individual compartmental modeling and population compartmental modeling techniques performed well on this routine set of bioequivalence data which displayed simple pharmacokinetic properties. A direct assessment of the analysis methods was made by comparing the final estimates and 90% confidence intervals for the test to reference ratios (T/R) of AUC and CMAX. The final estimates and 90% confidence intervals for AUC T/R and CMAX T/R were similar and suggest consistency of results, independent of the method used. CONCLUSIONS: These results demonstrate the utility of modeling techniques as adjuncts to the traditional noncompartmental approach for bioequivalence determination.

Population pharmacokinetic modeling: the importance of informative graphics.

PURPOSE: The usefulness of several modelling methods were examined in the development of a population pharmacokinetics model for cefepime. METHODS: The analysis was done in six steps: (1) exploratory data analysis to examine distributions and correlations among covariates, (2) determination of a basic pharmacokinetic model using the NON-MEM program and obtaining Bayesian individual parameter estimates, (3) examination of the distribution of parameter estimates, (4) multiple linear regression (MLR) with case deletion diagnostics, generalized additive modelling (GAM), and tree-based modelling (TBM) for the selection of covariates and revealing structure in the data, (5) final NONMEM modelling to determine the population PK model, and (6) the evaluation of final parameter estimates. RESULTS: An examination of the distribution of individual clearance (CL) estimates suggested bimodality. Thus, the mixture model feature in NONMEM was used for the separation of subpopulations. MLR and GAM selected creatinine clearance (CRCL) and age, while TBM selected both of these covariates and weight as predictors of CL. The final NONMEM model for CL included only a linear relationship with CRCL. However, two subpopulations were identified that differed in slope and intercept. CONCLUSIONS: The findings suggest that using informative graphical and statistical techniques enhance the understanding of the data structure and lead to an efficient analysis of the data.

Evaluation of bioequivalence of highly variable drugs using Monte Carlo simulations. I. Estimation of rate of absorption for single and multiple dose trials using Cmax.

PURPOSE: A Monte Carlo simulation study was done to investigate the effects of high intrasubject variation in clearance (CL), and volume of distribution (V) on the calculation of the 90% confidence interval (CI) for Cmax for single dose and multiple dose studies. METHODS: Simulations were done for both immediate release and sustained release scenarios. The simulated data were compared with clinical data from bioequivalence studies performed on indomethacin and verapamil. RESULTS: Previous reviews and simulations have shown that the probability of failure for the Cmax for single dose studies was always greater than that for multiple dose studies. However, the results for the simulated scenarios currently investigated indicate that if intra-subject (period-to-period) variation in CL and V is high (% CV's above 25%, and 12%, respectively), multiple dose studies can exhibit a higher probability of failure for Cmax than do single dose studies. Furthermore, Cmax values from studies performed with a sustained release scenario are more sensitive to changes in Ka, CL, and V than are results of studies on immediate release products. As an example, the probability of failure for immediate release products in simulated single dose studies is about 11% and 21% when the mean difference in Ka is 10% and 20%, respectively; while, the probability of failure for multiple dose studies is about 36% regardless of the difference in Ka. The corresponding values for the probability of failure for sustained release products were 25%, 53% for single dose studies and 39% for multiple dose studies. The simulations also indicate that changes in the fraction absorbed have a greater effect on the estimation of Cmax in multiple dose regimens than in single dose studies. CONCLUSIONS: The results from these investigations indicate that multiple dose studies do not necessarily always reduce variability in Cmax.

A spreadsheet program for simulation of bioequivalence and bioavailability studies.

This paper presents a spreadsheet for Excel for Windows, which simulates bioequivalence (BE) trials. The program incorporates intersubject and intrasubject variability in drug absorption and disposition as well as assay precision and the uniformity of the administered dose. The output provides confidence intervals and a pass/fail code for each study. This program is useful for simulating BE trials using widely available and simple-to-use spreadsheet programs. An example of the application of the program in assessing the influence of intrasubject variability on the outcome of BE testing of two identical formulations is also presented.

Comparison of the Akaike Information Criterion, the Schwarz criterion and the F test as guides to model selection.

In pharmacokinetic data analysis, it is frequently necessary to select the number of exponential terms in a polyexponential expression used to describe the concentration-time relationship. The performance characteristics of several selection criteria, the Akaike Information Criterion (AIC), and the Schwarz Criterion (SC), and the F test (alpha = 0.05), were examined using Monte Carlo simulations. In particular, the ability of these criteria to select the correct model, to select a model allowing estimation of pharmacokinetic parameters with small bias and good precision, and to select a model allowing precise predictions of concentration was evaluated. To some extent interrelationships among these procedures is explainable. Results indicate that the F test tends to choose the simpler model more often than does either the AIC or SC, even when the more complex model is correct. Also, the F test is more sensitive to deficient sampling designs. Clearance estimates are generally very robust to the choice of the wrong model. Other pharmacokinetic parameters are more sensitive to model choice, particularly the apparent elimination rate constant. Prediction of concentrations is generally more precise when the correct model is chosen. The tendency for the F test (alpha = 0.05) to choose the simpler model must be considered relative to the objectives of the study.

Comparison of single and multiple dose pharmacokinetics using clinical bioequivalence data and Monte Carlo simulations.

The purpose of this study was to evaluate the relative performance and usefulness of single dose (SD) and multiple dose (MD) regimens for bioequivalence (BE) determination. Drugs such as indomethacin, procainamide, erythromycin, quinidine, nifedipine were tested for BE under SD and MD dose regimens. Drugs characterized by low accumulation indices (AI) showed virtually no change in the 90% confidence interval (CI) of AUC and CMAX upon multiple dosing. On the other hand, drugs with higher AI appeared to have smaller CI at steady-state. For example, the CI range of AUC and CMAX of quinidine (AI of 1.54) decreased from 26 to 12 and from 22 to 12, respectively, upon multiple dosing. A Monte Carlo simulation study of SD and MD bioequivalence trials was performed. The probability of failing the bioequivalence test was evaluated for several situations defined by different levels of variability and correlation in ka constants, presence or absence of inter- and/or intra-individual variability in clearance (CL) and volume of distribution (V), and different degrees of accumulation. All the possible combinations of these factors were tested with SD and MD study designs. All simulations used 1000 data sets with 30 subjects in each data set for a total of 144 unique designs (total of 144,000 simulations of bioequivalence trials). Upon multiple dosing, narrowing of CI ranges was observed for drugs simulated to have high AI high variability and a large difference in absorption constants (ka) between test and reference formulations.(ABSTRACT TRUNCATED AT 250 WORDS)

Bioavailability of controlled release carbamazepine estimated by mixed effect modelling.

The absorption properties of a conventional tablet of carbamazepine (T) and a controlled release form of carbamazepine (TCR) have been compared using a nonlinear mixed effect model (NONMEM). Plasma carbamazepine concentration data were obtained from an open, steady-state, crossover bioavailability study in which 494 measurements were obtained from 13 patients, with an equal number of samples per patient for each dosage form. The pharmacokinetic model used as a one-compartment open model with first-order absorption and elimination. The objective function was used as a measure of the goodness of fit of the model to the data. Body weight was an important determinant of carbamazepine clearance (CL) but not volume of distribution (V). Accounting for the interindividual variability in volume of distribution did not significantly influence the objective function. Including different rates of absorption (ka) for the two dosage forms resulted in a significant improvement in the objective function, as well as reducing the interindividual variability in the rate of absorption. Adding a parameter for relative bioavailability (f) of TCR improved the objective function statistically, but an unrealistic value for V was obtained, and the absorption and elimination rates appeared to be transposed in the classical "flip-flop" manner. Fixing V to the value obtained before introducing f did not change the objective function and permitted estimation of f without the confounding influence of excessive parameters.(ABSTRACT TRUNCATED AT 250 WORDS)