Creatinine- and cystatin C-based population pharmacokinetic models for predicting mycophenolic acid clearance in patients with autoimmune diseases.

PURPOSE: To quantitatively compare the effectiveness of serum creatinine- and cystatin C-based equations in population pharmacokinetic (PPK) models in predicting renal function for mycophenolic acid (MPA) clearance in patients with autoimmune diseases. METHODS: In this single-centre retrospective study, 176 plasma samples from 37 adult patients (≥ 18 years) with autoimmune diseases who received mycophenolate mofetil were analysed. Renal function was estimated using creatinine clearance (Ccr) based on the Cockcroft-Gault equation and Chronic Kidney Disease Epidemiology Collaboration equation (creatinine-, cystatin C-, and creatinine+cystatin C-based). The PPK model was developed using Phoenix NLME 8.1 software. Model diagnostics were performed using goodness-of-fit plots, and the model adequacy was assessed using visual predictive check and bootstrap methods. RESULTS: Plasma MPA concentrations were modelled using a two-compartment model with first-order elimination and a transit compartment for the absorption process. In the PPK analysis, MPA clearance was most effectively explained by the Ccr. Moreover, Ccr and albumin levels were identified as covariates significantly influencing MPA clearance. CONCLUSION: For adults with autoimmune diseases, creatinine-based equations (particularly Ccr) are superior in predicting MPA clearance. Cystatin C levels reflect glomerular filtration alone, whereas creatinine is eliminated via both glomerular filtration and tubular secretion. Since creatinine and a major MPA metabolite are eliminated by tubular secretion, Ccr is the optimal covariate in this analysis. However, cystatin C-based equations may better explain MPA clearance in paediatric patients or older adults with low muscle mass. Therefore, selecting the appropriate renal function equation according to patient characteristics is crucial for optimal mycophenolate mofetil dosing.