An exploration on retro-construction of plasma drug concentration-time curves from corresponding urine excretion data and single-point plasma concentrations using a simplified and idealized method

Background: Despite the availability of various tools of modeling and simulation, clinical pediatric pharmacokinetic (PK) studies remain far less efficient than those on adults due to ethical constraints. One of the optimal solutions is to substitute urine to blood sampling based on explicit mathematic relationships between them. However, this idea is limited by three main knowledge gaps associated with urine data, i.e., complicated excretion equations with excessive parameters, insufficient frequency that is hard to fit, and the mere expression of amounts with no in vivo distribution volume information involved. Methods: To overcome these obstacles, we sacrificed the precision from mechanistic PK models with complex excretion equations to expediency of compartmental model in which a constant k(e) is used to cover all the internal parameters. And the total cumulative amounts of urinary drug excretion (X-u(infinity)) were estimated and introduced to the excretion equation so that urine data were likely to be fitted using a semi-log-terminal linear regression method. In addition, urinary excretion clearance (CLr) could be calculated by single point plasma data to anchor the plasma concentration-time (C-t) curve based on the assumption that CLr was kept constant throughout the PK process. Results: After sensitivity analysis of two subjective judgements (the selection of the compartmental model and the selection of plasma time point to calculate CLr), the performance of the optimized models was assessed using desloratadine or busulfan as model drugs in a variety of PK scenarios, from i.v. bolus/infusion to p.o. administration, from a single dose to multiple doses, and from rats to children. The fitting plasma drug concentrations of the optimal model were close to the observed value. Meanwhile, the drawbacks inherent to the simplified and idealized modeling strategy were fully identified. Conclusions: The method proposed by this tentative proof-of-principle study was able to deliver acceptable plasma exposure curves and shed light on the future refinements.