A Simulation Study of the Comparative Performance of Partial Area under the Curve (pAUC) and Partial Area under the Effect Curve (pAUEC) Metrics in Crossover Versus Replicated Crossover Bioequivalence Studies for Concerta and Ritalin LA.

Concerta and Ritalin LA are methylphenidate (MPH) drugs with different release mechanisms. Generic bioequivalence (BE) to these conventionally uses pAUC (partial area under the curve) as metrics in addition to Cmax (maximum concentration), AUC0-t (area from time 0 to time t), and AUC0-infinity. The recommended BE design was a standard two-formulation, two-sequence, and two-period crossover; however, the currently recommended design is a replicated crossover to better define subject-by-formulation interaction variance. The current purpose was to compare via simulation, using literature MPH models, the performance of the pAUC metrics in establishing BE via the standard crossover design versus a replicated design, and the relationship of the pAUC metrics to PD (pharmacodynamics) metrics, e.g., SKAMP (Swanson, Kotkin, Agler, M-Flynn, and Pelham rating scale) composite scores pAUEC (partial area under the effect curve). One-thousand, 40-subject studies were simulated with model literature parameters. An indirect response model described the SKAMP composite scores corrected for placebo. Performance of the pAUC metrics was demonstrated by the calculation of 90% confidence intervals (CIs) for each k0fast (fast absorption rate constant) and kaslow (slow absorption rate constant) test/reference (T/R) ratio. The 90% CIs resulting from changes in the k0fast and kaslow ratios, e.g., T/R, showed greater sensitivity to changes in the ratios at quotients below 0.8 than above for both Concerta and Ritalin LA. Ritalin LA pAUC values were insensitive to increases in either ratio once the ratio exceeded 1.0 and the study design. Correlations between least squares means (LSM) for pAUC and the SKAMP pAUEC for the composite scores were near 90%.

The US Food and Drug Administration's perspective on the new antidepressant vortioxetine.

OBJECTIVE: This article summarizes the US Food and Drug Administration's (FDA's) review of the New Drug Application for vortioxetine, especially the clinical efficacy and safety data. It emphasizes the issues that were important to the FDA's approval decision, particularly the difference in the effective dose in domestic and foreign studies, and notes several new labeling features, specifically, description of time course of treatment response and detailed sexual dysfunction evaluation. DATA SOURCES: The data sources were the original raw data sets for all clinical trials included in the development program for vortioxetine, as well as the sponsor's original analyses of these data. Data were available from 51 human trials involving vortioxetine, and included a total of 7,666 healthy volunteers and patients with a diagnosis of major depressive disorder (MDD) or generalized anxiety disorder who were exposed to at least 1 dose of vortioxetine for a total of 2,743 patient-years. RESULTS: Vortioxetine was effective in treating MDD in the United States at a dose of 20 mg/d. The recommended starting dose is 10 mg once daily without regard to food, with increase to 20 mg/d if the 10 mg/d dose is tolerated. For patients who do not tolerate 20 mg/d, 10 mg/d can be used and 5-mg/d dose can be considered. Vortioxetine can be discontinued abruptly, but it is recommended that doses of 15 mg/d or 20 mg/d be reduced to 10 mg/d for 1 week prior to full discontinuation to avoid potential withdrawal symptoms. Although the non-US maintenance study showed that maintenance doses of 5 to 10 mg/d were effective, a clinical judgment needs to be made to decide the maintenance dose in the United States. The applicant has agreed to conduct a US maintenance dose-response study covering the US-approved dose range. Vortioxetine's adverse event profile is similar to that of other selective serotonin reuptake inhibitors (SSRIs). Nausea is the most common adverse event and is dose dependent. No dose adjustment is needed based on age, gender, or the presence of renal or mild to moderate hepatic impairment. The maximum recommended dose is 10 mg/d in known cytochrome P450 2D6 poor metabolizers. CONCLUSIONS: Vortioxetine is a new treatment for MDD, and its adverse event profile is similar to that of other SSRIs.

In silico, experimental, mechanistic model for extended-release felodipine disposition exhibiting complex absorption and a highly variable food interaction.

The objective of this study was to develop and explore new, in silico experimental methods for deciphering complex, highly variable absorption and food interaction pharmacokinetics observed for a modified-release drug product. Toward that aim, we constructed an executable software analog of study participants to whom product was administered orally. The analog is an object- and agent-oriented, discrete event system, which consists of grid spaces and event mechanisms that map abstractly to different physiological features and processes. Analog mechanisms were made sufficiently complicated to achieve prespecified similarity criteria. An equation-based gastrointestinal transit model with nonlinear mixed effects analysis provided a standard for comparison. Subject-specific parameterizations enabled each executed analog's plasma profile to mimic features of the corresponding six individual pairs of subject plasma profiles. All achieved prespecified, quantitative similarity criteria, and outperformed the gastrointestinal transit model estimations. We observed important subject-specific interactions within the simulation and mechanistic differences between the two models. We hypothesize that mechanisms, events, and their causes occurring during simulations had counterparts within the food interaction study: they are working, evolvable, concrete theories of dynamic interactions occurring within individual subjects. The approach presented provides new, experimental strategies for unraveling the mechanistic basis of complex pharmacological interactions and observed variability.

Quantitative drug benefit-risk assessment: utility of modeling and simulation to optimize drug safety in older adults.

BACKGROUND: More than 50 % of individuals affected by adverse drug events (ADEs) are older adults. Establishing a drug dosing regimen that balances benefit and risk, and minimizes ADEs in older populations can be challenging. OBJECTIVE: The aim of this study is to evaluate the use of modeling, simulation, and risk-benefit acceptability methods to establish a drug dosing regimen that balances benefit and risk. METHODS: The study population comprised nondiabetic patients with schizophrenia from the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) ≥50 years old, who had been on oral olanzapine for ≥2 weeks. We used mixed-effects modeling based on a preexisting pharmacokinetic model to derive clearance estimates, which were then used to determine the olanzapine area under the concentration-time curve (AUC). Subsequently, with multivariate regression and Monte Carlo simulation, we estimated the olanzapine dose corresponding to the benefit-risk AUC breakpoint. RESULTS: The study population (n = 34) was predominantly male (82.3%) and white (67.6%), with a mean age of 54.4 years and treatment duration of 361.8 days. The mean AUC was 747.6 ng h/mL (95% CI = 524.5, 970.7) for the benefit group (n = 16) and 754.1 (95% CI = 505.9, 1002.4) for the risk group (n = 15). The benefit-risk AUC breakpoint was 524.5 ng h/mL and the corresponding oral olanzapine dose that optimizes benefit-risk balance was 17.8 mg/d. CONCLUSIONS: Our study introduces a real-world approach for finding the safe drug dosing regimen without extensive exposure of a vulnerable and older population to drugs. Further studies into the use of modeling, simulation, and risk-benefit acceptability methods to enhance geriatric drug safety are needed.

Impact of release mechanism on the pharmacokinetic performance of PAUC metrics for three methylphenidate products with complex absorption.

PURPOSE: Investigate the performance of partial area under the drug concentration-time curve (PAUC) metrics (0-3 h) and (3-24 h), for Concerta, Ritalin LA and Focalin XR (different Methylphenidate modified-release formulations). The metrics have been chosen as additional BE metrics for Ritalin LA by the FDA to establish BE for these products due to the early and late peak concentrations critical for treatment of morning and afternoon symptoms of attention deficit hyperactivity disorder (ADHD). METHODS: Two-stage analysis was performed on plasma data for the methylphenidate modified-release products. Simulations using the fitted parameters determined how changes in fast absorption rate constant k0fast, and slow absorption rate constant KAslow affected curve shape and BE determination using Cmax, AUCINF and PAUC. RESULTS: Sensitivity of the mean PAUC(test)/PAUC(reference) ratios to changes in k0fast and Kaslow were product dependent. Focalin XR mean PAUC(test)/PAUC(reference) ratios for PAUC0-3 h and PAUC3-24 h were most responsive to changes in k0Fast and Kaslow than Concerta and Ritalin LA. The PAUC(test)/PAUC(reference) ratios for (0-3 h) were not responsive to changes to Kaslow. Concerta PAUC (3-24 h) ratios were responsive to changes in Kaslow at ratios less than 1. CONCLUSIONS: Response to PAUC(0-3 h) in the formulations was greater for k0fast than was PAUC(3-24) to changes in KAslow.

Use of partial AUC (PAUC) to evaluate bioequivalence--a case study with complex absorption: methylphenidate.

PURPOSE: Methylphenidate modified-release products produce early and late peak concentrations critical for treatment of morning and afternoon symptoms of attention deficit hyperactivity disorder (ADHD). Standard bioequivalence (BE) criteria cannot be applied to these products. The performance of partial area under the drug concentration-time curve (PAUC), Cmax and AUCINF to assess BE were independently evaluated for two products. METHODS: A two-stage analysis was performed on plasma data for two methylphenidate modified-release products (Product 1 and 2). Simulations using the fitted parameters determined how changes in fast absorption rate constant (K0Fast) and fraction available (F1) affected curve shape and BE determination using Cmax, AUCINF and PAUC. RESULTS: The sensitivity of the mean PAUC(test)/PAUC(reference) ratios to changes in K0Fast(test) are product dependent. Product 1 mean PAUC(test)/PAUC(reference) ratios for PAUC0-4h are more responsive to both decreases and increases in K0Fast(test) than Product 2. Product 2 showed a greater response in the mean PAUC(test)/PAUC(reference) ratio for PAUC0-4h when the K0Fast(test) is decreased and less response as the value is increased. CONCLUSIONS: PAUC estimated curve shape is sensitive to changes in absorption and are product specific, and may require a new PAUC metric for each drug. A non-product specific metric to assess curve shape is warranted.

Individualized, discrete event, simulations provide insight into inter- and intra-subject variability of extended-release, drug products.

OBJECTIVE: Develop and validate particular, concrete, and abstract yet plausible in silico mechanistic explanations for large intra- and interindividual variability observed for eleven bioequivalence study participants. Do so in the face of considerable uncertainty about mechanisms. METHODS: We constructed an object-oriented, discrete event model called subject (we use small caps to distinguish computational objects from their biological counterparts). It maps abstractly to a dissolution test system and study subject to whom product was administered orally. A subject comprises four interconnected grid spaces and event mechanisms that map to different physiological features and processes. Drugs move within and between spaces. We followed an established, Iterative Refinement Protocol. Individualized mechanisms were made sufficiently complicated to achieve prespecified Similarity Criteria, but no more so. Within subjects, the dissolution space is linked to both a product-subject Interaction Space and the GI tract. The GI tract and Interaction Space connect to plasma, from which drug is eliminated. RESULTS: We discovered parameterizations that enabled the eleven subject simulation results to achieve the most stringent Similarity Criteria. Simulated profiles closely resembled those with normal, odd, and double peaks. We observed important subject-by-formulation interactions within subjects. CONCLUSION: We hypothesize that there were interactions within bioequivalence study participants corresponding to the subject-by-formulation interactions within subjects. Further progress requires methods to transition currently abstract subject mechanisms iteratively and parsimoniously to be more physiologically realistic. As that objective is achieved, the approach presented is expected to become beneficial to drug development (e.g., controlled release) and to a reduction in the number of subjects needed per study plus faster regulatory review.

Bioequivalence of long half-life drugs--informative sampling determination--using truncated area in parallel-designed studies for slow sustained-release formulations.

A simulation study was done to determine if 72 h is the most informative sampling duration for bioequivalence (BE) determination in parallel-designed BE studies with drugs that have half-lives of at least 30 h. The impact of absorption and elimination half-lives on informative sampling was evaluated. Two-treatment parallel-designed BE studies using a one-compartment oral absorption model with half-lives of 30 and 350 h was simulated. Area under the curve (AUC) values were truncated at 12-360 h. Experimental BE data [median time to reach the maximum concentration (T(max) ) = 20 h and low clearance = 0.192 L/h) indicated a decrease and then an increase in the intrasubject variability [root mean square error (RMSE)] for truncated AUC as a function of time. Simulations supported these findings with the highest probability of passing the BE confidence interval criteria being between 60 and 96 h, depending on half-life and percent coefficient of variation. The 30-h simulation exhibited a minimum in RMSE at 24-h truncation that continued to increase up to 360 h, whereas the 350-h simulation exhibited a minimum at 60 h, which increased after 96 h. Power curves at the 350-h half-life showed higher probabilities of rejection of BE for true test/reference ratios greater than 0.9. For parallel-designed BE studies, sampling beyond 120 h will not affect the BE decision and therefore is unnecessary.

Bioavailability considerations in evaluating drug-drug interactions using the population pharmacokinetic approach.

Applying a comedication (COMD) covariate to apparent clearance (CL(app) = CL/F) is a common practice when using population pharmacokinetics (PopPK) to study metabolism-based drug-drug interactions (DDI). This study evaluates the importance of independently applying COMD to F and CL to account for DDI at the level of first-pass metabolism. A known DDI between single oral doses of the CYP3A substrate midazolam (5 mg) and the inhibitor ketoconazole (400 mg) was simulated using a physiologically based pharmacokinetic simulator SimCyp in virtual subjects. The simulated midazolam data were analyzed by PopPK method under the following scenarios by applying COMD effect to (1) CL(app) only, (2) CL and F, and (3) CL(app) and apparent volume of distribution (V(app) = V/F), assuming V is unchanged. The mean simulated degree of interaction, measured by midazolam AUC ratio with and without ketoconazole (AUCR), was 10.28. Scenario 1 underestimated AUCR. When COMD was independently applied to F and V(app) in scenarios 2 and 3, lower objective function values of the PopPK analysis and more accurate AUCR estimates were achieved. AUCR estimates were also dependent on sampling. The authors conclude that when significant inhibition of the first-pass metabolism of the substrate is anticipated, COMD effects should be applied to both CL and F in PopPK analysis.

Role of metabolites for drugs that undergo nonlinear first-pass effect: impact on bioequivalency assessment using single-dose simulations.

We investigated the effects of dose and intrasubject variability (ISV) on bioequivalence (BE) of a parent drug with a single metabolite formed by nonlinear first-pass. A BE simulation was done using a four-compartment model at doses of 17.5, 35.0, and 70.0 mg. ISV was set at either 10% or 20% for clearance and either 20% or 50% for the absorption rate constant, K(a). The ratio of Katest/Kreference was fixed at 1.00 while fraction available ratios, F(test)/F(reference), were varied from 1.00 to 1.25. Results showed the probability of passing the 90% confidence interval (CI) BE requirement for AUC(I), area-under-the-concentration curve to time infinity, and C(max), concentration maximum, were greater for the metabolite than the parent at all F(test)/F(reference) ratios. For the parent, the probability of meeting BE criteria for AUC(I) and C(max) declined from 100% to 60% at the 70 mg dose as the ISV for K(a) increased from 20% to 50% with an increased F(test)/F(reference) ratio. For the metabolite, the probability of meeting BE criteria was above 80% for all doses and ISV values and F(test)/F(reference) ratios less than 1.15. Results show that the parent, reflected absorption, is more informative for determining BE than the metabolite. Clinical data gave a similar result.

Metabolites and bioequivalence: past and present.

Although it is widely recognised that measurement of metabolite concentrations is crucial to understanding the clinical pharmacology characteristics of a new molecular entity, a clear consensus on the role of metabolites in the assessment of bioequivalence has never been achieved within the scientific community. However, a regulatory policy for the role of metabolites in bioavailability and bioequivalence has been established by the US FDA. One school of thought believes that the parent drug alone is sensitive to picking up formulation differences, whereas another school of thought believes that establishing bioequivalence criteria on all the species that contribute to safety and efficacy is the only way to ensure the switchability of two products. In this paper, a brief review of the pharmacokinetics of metabolites under different scenarios is presented and the history of the role of metabolites in the assessment of bioequivalence is summarised. Relevant examples from the literature illustrating conflicting opinions on the need for the measurement of metabolites in bioequivalence studies are given. Cases from the literature in which the parent drug is able to meet the 90% confidence intervals while the metabolite(s) fail to do so, and vice versa, are presented to illustrate the difficulty in choosing the pertinent entity to measure. The relevant current US FDA policy and guidelines related to bioavailability and bioequivalence are discussed and contrasted with the rules and regulations applicable in Canada and Europe.

Effect of length of sampling schedule and washout interval on magnitude of drug carryover from period 1 to period 2 in two-period, two-treatment bioequivalence studies and its attendant effects on determination of bioequivalence.

The relationships between post-dosing blood sampling schedules and length of washout interval on the percent of drug carryover into period 2, in a two-treatment, two-period, crossover bioequivalence study was investigated. Observed simulations were done using a two-compartment model with a beta half-life of 100 h, sampling to 200 and 300 h, followed by 1 to 4 washout half-lives. These data were compared with simulations with sampling for 200 and 300 h and no carryover between periods (i.e. true data) and used to establish the per cent carryover. Pseudo-observed concentration period 2 data were also generated for an experimental amiodarone bioequivalence study by adding either 5%, 10% or 50% of the period 2 C(max) value to the observed concentrations of selected subjects (i.e. those with longest half-lives) to give a pseudo-plasma concentration profile. Further investigation via simulation was done by including or excluding subjects with time 0 concentrations of 1%-10% of period 2 C(max). The simulated data indicated that up to a 3% carryover of AUC into the second period of a bioequivalence study had no effect on the power of the 90% CI for AUC and C(max). For amiodarone, second period pre-dose drug concentrations equal to as much as 50% of the observed period 2 C(max) value had less effect on the 90% CI (i.e. compared with the true data) for AUC and C(max) than did subject deletion.

Determination of in vivo bioequivalence.

The May and June 2001 issues of Pharmaceutical Research contained three articles related to the determination of in vivo Bioequivalence. The articles discussed: (a) the bioequivalence of highly variable drugs, (b) novel metrics for direct comparison of bioequivalence study plasma curves, and (c) the role of a microemulsion vehicle on cutaneous bioequivalence. An analysis of the relationship and potential impact of these articles on their respective areas of bioequivalence will be addressed in this commentary.