Monte-Carlo Simulation of Pharmacokinetics of Nexium DR Capsule (20mg)
Purpose: The study aims to 1) run Monte-Carlo simulation to generate individual in vivo release profiles of DR Capsules and simulate individual PK profiles among 20 subjects; and 2) to run sensitivity analysis to assess the effects of release parameters on the success rate of bioequivalence of generic drugs. Methods: A two-compartment PK model was developed to describe esomeprazole PK; one depot compartment was developed to describe release and absorption (first-order absorption with rate constant Kabs and bioavailability F) of esomeprazole from DR capsules. Drug release was described by Weibull function with a lag time Tlag, a release time Trel, and a shape factor. Model parameters were sequentially estimated from PK profiles after IV infusion administration (to estimate clearance rate CL, elimination rate constant Kel, rate constants K01 & K10) and after oral solution administration (to estimate Kabs & F). They were then used to obtain, via deconvolution/numerical convolution, the in vivo release profile of DR capsules from the mean PK profile after DR capsule administration. Based on the in vitro release profile (obtained via USP-2 method) and the estimated in vivo release profile, Monte-Carlo simulation was run to generate Weibull function parameters to construct individual in vivo release profiles and to generate individual PK profiles. The simulated mean PK profile was compared to the observed mean PK profile to validate the model and the method. Sensitivity analysis was run to test the effect of each release parameter on the success rate of bioequivalence. Results: A two compartment model was sufficient to describe esomeprazole PK of dose 20mg; this does not exclude other possible models. In vitro release profile and the estimated in vivo release profile differ significantly; this does not exclude IVIVC since time transformation can be conducted to obtain IVIVC. Variation in individual in vivo release profiles can almost completely explain the observed mean PK profile; variation in individual PK parameters will add more complexity to the model. The mean in vivo release profile has great potential to be used to obtain IVIVC. Sensitivity analysis indicated that Trel and Tlag has have more effects on bioequivalence than shape factor. Conclusion: Monte-Carlo simulation and clinical trial simulation, based on good estimates of model parameters and release kinetics, have great potential to simulate and reproduce PK data for drug products. They can be used as powerful tools to test bioequivalence, design and optimize drug formulations, and expedite development of drug products. Their application in drug development can help decision making, design better studies, reduce costs, save time, and ultimately improve success rates.