Figure 1.
Work flow of PKPD modeling.
Figure 2.
Model structure of the PKPD model.
(A) Semi-PBPK model for the systemic kinetics of GL and GA in rat. (B) Semi-PBPK model for the systemic kinetics of GL and GA in human. (C)11β-HSD2 associated renin-angiotensin-aldosterone-electrolyte biological system PD model.
Table 1.
Abbreviations of parameters of the physiologically based pharmacokinetic model for GL, GA and GAM.
Table 2.
Physiological parameters in a 250 g rat and 70 kg human.
Table 3.
Biochemical parameters for the physiologically based model for GL, GA and GAM in rat.
Figure 3.
GL plasma concentration and accumulated biliary excretion after i.v. GL in rats with bile fistulas.
Plasma concentration-time profiles of GL at (A) 5–50 mg/kg [15], (C) 10 mg/kg [13], (D) 25 and 100 mg/kg [18], [19], [31]; accumulated biliary excretion of GL at (B) 5–50 mg/kg [15] and (C) 10 mg/kg [13]. Experimental data (fitset and testset) are shown as symbols; the lines represent the prediction of the GL model.
Figure 4.
GA plasma and kidney concentration after i.v. GA and p.o. GA and GL in rats.
(A) Plasma concentration-time profiles of GA after i.v. 2–20 mg/kg GA in rats with bile fistulas [29]; (B) i.v. 60 mg/kg GA in rats with bile fistulas [18], i.v. 5.7 mg/kg GA in rats without bile fistulas and p.o. 5.7 mg/kg GA in rats [30]; (C) p.o. 10 [30] and 100 (experimental studies) mg/kg GL; (D) plasma and kidney exposure of GA after p.o. 200 mg/kg GL (experimental studies). Experimental data (fitset and testset) are shown as symbols; the lines represent the forecast of the GA model.
Figure 5.
Plamsa concentration and accumulated biliary excretion of GAM.
Left, plasma concentration and accumulated biliary excretion of 3MGA after i.v. 5 mg/kg 3MGA in rats [5]; Right, accumulated biliary excretion of GAM after i.p. 25 mg/kg GA in rats [25]. Experimental data are shown as symbols; the lines represent the predictions of the GA model.
Figure 6.
Plasma concentrations of GL and GA in human.
(A) GL plasma concentration after i.v. 40–120 mg GL in human [16]; (B) GA plasma concentration after p.o. administration of 130 mg/kg GA at (B, left) single dose and (B, right) multiple dose [10]; GA plasma concentration after p.o. (C, left) 225 mg GL and (C, right) 150 g licorice containing 225 mg GL [9]. Experimental data (fitset and testset) are shown as symbols; the lines represent the predictions of the human PBPK model.
Table 4.
Biochemical parameters of the physiologically based model for GL, GA and GAM in human.
Figure 7.
Time courses of urinary excretion of cortisol, cortisone and their ratio in different scenarios: (A) p.o. GA 130 mg/day for 5 days and withdrawn for another 5 days [10]; (B–D) p.o. GA 500 mg/day, 2 times/day for 7 days [34].
Figure 8.
The effects of different levels of (A) angiotensin II [40] and (B) potassium [41] on aldosterone concentration.
Experimental data (fitset) are shown as symbols; the lines represent the predictions of the human PBPK model. i.f. in B stands for intravenous infusion.
Table 5.
Optimized parameters in the PD model.
Table 6.
Estimated urinary cortisol∶cortisone ratio for p.o. administration of GA at 510 mg/day administered 3 times/day for 2 days compared with the observed value.
Table 7.
Estimated biomarkers of pseudoaldosteronism for p.o. administration of GA at 500 mg/day 2 times/day for one week compared with the observed value.
Table 8.
Estimated biomarkers of pseudoaldosteronism for p.o. administration of GL at 0.7 g/day (9 subjects) or 1.4 g/day (5 subjects) for one to four weeks compared with observed values.
Figure 9.
The simulated effect of three sensitivity factors on GA pharmacokinetics and potassium level.
Sensitivity factors: (A, B) sinusoidal transport function, (C, D) colonic transit time and (E) activity of 11β-HSD 2. The causal effect: (A, C) GA plasma concentration and (B, D, E) serum potassium level. GL was administered p.o. at 200 mg/day for one week. N stands for the normal value of the corresponding parameter.
Figure 10.
The simulated probability distribution of the individual dose limit in 1000 virtual elderly people.
(A) The simulated dose data and fitted probability density by lognormal distribution; (B) The simulated dose data and fitted cumulative probability by log normal distribution; (C) The second derivative of the cumulative probability function and the critical value.
Table 9.
Distribution of CLup, Kco, and kox,0 related physiological factors in simulation of the individual dose limit for 1000 subjects by the Monte Carlo method.