Background: A population pharmacokinetic (PK) protein-binding model was developed to (1) predict free mycophenolic acid (fMPA) based on total MPA (tMPA) concentrations in renal transplant patients, to establish the therapeutic range of fMPA through pharmacokinetic-pharmacodynamic studies; and (2) provide a guideline for dosing mycophenolate mofetil (MMF).
Methods: Full PK profiles of 56 patients (from five different occasions) during the first year after transplantation who were treated with oral MMF and cyclosporine, or macrolides (either tacrolimus or sirolimus), were analysed. fMPA protein-binding was modelled using nonlinear mixed effects modelling (NONMEM). The influence of physiological factors and coadministered immunosupressant was studied.
Results: A two-compartment model with first-order absorption and elimination, linear protein binding and enterohepatic circulation (EHC) best described the PK of MPA. Different recycling rate constants were considered depending on the coadministered immunosuppressant. The protein-binding rate constant (KB [relative standard error, RSE%]) increased nonlinearly with renal function according to K B = 43.1 (3.13)·(CLCR/59.51)0.394(10.66) h-1. Furthermore, fMPA plasma clearance, given by clearance of the free mycophenolic acid (CLfMPA), CLfMPA = 410 (RSE%3.00)·(1+CsA·0.594 (22.39)) L/h, was 59.4% greater in cyclosporine-treated patients than in macrolide-treated patients, leading to lower MPA exposures. External evaluation proved acceptable area under the plasma concentration-time curve and trough concentration predictions.
Conclusions: A reliable protein-binding population PK model was developed for prediction of fMPA or tMPA from each other and for dose guiding in stable renal transplant recipients.