Biol. Pharm. Bull. 29(7) 1454—1459 (2006)

ease (ESRD) is quite different from that in patients with normal renal function. Numerous approaches are undertaken to appropriately adjust the dosage regimens of renal excretory drugs in ESRD patients because the systemic exposure in this population unexpectedly increases and frequently causes adverse events. In contrast, hemodialysis sometimes results in an inadequate decrease of the plasma drug level from the therapeutic range. Consequently, pharmacokinetic information for ESRD patients is requisite in the process of new drug development in order to ascertain the effect of renal failure on tolerability and exposure and to determine the dosage regimen. However, the clinical study of ESRD patients has not often been performed due to difficulties with patient enrollment. In such cases, prediction of the pharmacokinetics might be an alternative to clinical studies and provide useful information for optimizing the dosage regimen. Modeling and simulation have been explored by many scientists, and recently virtual clinical trials based on pharmacokinetics and pharmacodynamics (PK/PD) modeling and subsequent simulation have been employed for decisionmaking in drug development prior to cost consuming phase III/IV trials. In particular, this kind of methodology is informative for the development of antibiotic agents, since there is clear evidence that the pharmacological effect is well correlated to some PK/PD parameters. Hemodialysis is a physical process that can be described by a mathematical model. There are many factors affecting drug removal during hemodialysis, including molecular weight, lipid and water solubility, surface area, porosity of the dialysis membrane, blood and dialysate flow rates, protein binding and red blood cell partitioning. Even though some methods were previously developed considering these factors, they are complicated for clinical use. Thus, we are trying to develop a simple method based on several assumptions. Utilization of the chemical structure and pharmacokinetic relationship can be a useful approach for the prediction of pharmacokinetics. For example, apparent oral clearances in humans could be adequately extrapolated from animal data, while the structural parameters could be evaluated by a multivariate analysis of various drugs selected in view of their structure, molecular weight, partition coefficient, number of hydrogen bond acceptors, pharmacological activities, and pharmacokinetic characteristics. Antibiotics are suitable for this approach because they are diverse enough to evaluate the method in terms of various physicochemical and pharmacokinetic properties. The purpose of this study was to provide a new method for the prediction of pharmacokinetics in ESRD patients and to confirm its validity and applicability using an experimental renal failure model.