Abstract

A physiologically based pharmacokinetic (PBPK) model of vitamin D₃ and metabolites [25(OH)D₃, 1,25(OH)₂D₃, and 24,25(OH)₂D₃] is presented. In this study, patients with 25(OH)D₃ plasma concentrations below 30 ng/ml were studied after a single dose of 5,000 I.U. (125 µg) cholecalciferol, provided with 5,000 I.U. daily cholecalciferol supplementation until vitamin D replete (25(OH)D₃ plasma concentrations above 30 ng/ml), and had serial plasma samples were collected at each phase for 14 days. Total concentrations of vitamin D₃ and metabolites were measured by ultra-high performance liquid chromatography tandem mass spectrometry. A nine-compartment PBPK model was built using MATLAB to represent the triphasic study nature (insufficient, replenishing, sufficient). Stimulatory and inhibitory effect of 1,25(OH)₂D₃ were incorporated by fold-changes in the primary metabolic enzymes CYP27B1 and CYP24A1, respectively. Incorporation of dynamic adipose partition coefficients for vitamin D₃ and 25(OH)D₃ and variable enzymatic reactions aided in model fitting. Measures of model predictions agreed well with data from metabolites, with 97%, 88%, and 98% of the data for 25(OH)D₃, 24,25(OH)₂D₃, and 1,25(OH)₂D₃, respectively, within 2-fold of unity (fold error values between 0.5 and 2.0). Bootstrapping was performed and optimized parameters were reported with 95% confidence intervals. This PBPK model could be a useful tool for understanding the connections between vitamin D and its metabolites under a variety of clinical situations.

Significance Statement This study developed a physiologically based pharmacokinetic (PBPK) model of vitamin D₃ and three metabolites for patients moving from a depleted to a repleted phase over a period of 16 weeks.