• Korean Journal of Clinical Pharmacy
• /
• v.31 no.2
• /
• pp.125-135
• /
• 2021

Background: Generally, pharmacokinetics (PK) models could be stratified into two models. The compartment PK model uses the concept of simple compartmentalization to describe complex bodies, and the physiologically based pharmacokinetic (PBPK) model describes the body using multi-compartment networking. Notwithstanding sharing a theoretical background in both models, there was still a lack of knowledge to enhance compatibility in both models. Objective: This study aimed to evaluate the compatibility among PBPK, lumping model and compartment PK model with voriconazole PK case study. Methods: The number of compartments and blood flow on each tissue in the PBPK model were modified using the lumping method, considering physiological similarities. The concentration-time profiles and area under the concentration-time curve (AUC) parameters were simulated at each model, assuming taken voriconazole oral 400 mg single dose. After that, those mentioned PK parameters were compared. Results: The PK profiles and parameters of voriconazole in the three models were similar that proves their compatibility. The AUC of central compartment in the PBPK and lumping model was within a 2-fold range compared to those in the 2- compartment model. The AUC of non-eliminating tissues compartment in the PBPK model was similar to those in the lumping model. Conclusion: Regarding the compatibility of the three PK models, the utilization of the lumping method was confirmed by suggesting its reliable PK parameters with PBPK and compartment PK models. Further case studies are recommended to confirm our findings.

• Korean Journal of Clinical Pharmacy
• /
• v.26 no.4
• /
• pp.312-317
• /
• 2016

Objective: Midazolam is mainly metabolized by cytochrome P450 (CYP) 3A. Inhibition or induction of CYP3A can affect the pharmacological activity of midazolam. The aims of this study were to develop a population pharmacokinetic (PK) model and evaluate the effect of CYP3A-mediated interactions among ketoconazole, rifampicin, and midazolam. Methods: Three-treatment, three-period, crossover study was conducted in 24 healthy male subjects. Each subject received 1 mg midazolam (control), 1 mg midazolam after pretreatment with 400 mg ketoconazole once daily for 4 days (CYP3A inhibition phase), and 2.5 mg midazolam after pretreatment with 600 mg rifampicin once daily for 10 days (CYP3A induction phase). The population PK analysis was performed using a nonlinear mixed effect model ($NONMEM^{(R)}$ 7.2) based on plasma midazolam concentrations. The PK model was developed, and the first-order conditional estimation with interaction was applied for the model run. A three-compartment model with first-order elimination described the PK. The influence of ketoconazole and rifampicin, CYP3A5 genotype, and demographic characteristics on PK parameters was examined. Goodness-of-fit (GOF) diagnostics and visual predictive checks, as well as bootstrap were used to evaluate the adequacy of the model fit and predictions. Results: Twenty-four subjects contributed to 900 midazolam concentrations. The final parameter estimates (% relative standard error, RSE) were as follows; clearance (CL), 31.8 L/h (6.0%); inter-compartmental clearance (Q) 2, 36.4 L/h (9.7%); Q3, 7.37 L/h (12.0%), volume of distribution (V) 1, 70.7 L (3.6%), V2, 32.9 L (8.8%); and V3, 44.4 L (6.7%). The midazolam CL decreased and increased to 32.5 and 199.9% in the inhibition and induction phases, respectively, compared to that in control phase. Conclusion: A PK model for midazolam co-treatment with ketoconazole and rifampicin was developed using data of healthy volunteers, and the subject's CYP3A status influenced the midazolam PK parameters. Therefore, a population PK model with enzyme-mediated drug interactions may be useful for quantitatively predicting PK alterations.