RESUMEN
OBJECTIVE To establish a physiologically-based pharmacokinetic （PBPK） model of amikacin in elderly patients with renal insufficiency. METHODS PK-SIM® software was adopted for model building， optimization and simulation. The physical and chemical properties and pharmacokinetic parameters related to amikacin were collected by literature review. The PBPK model on adults was established and extrapolated to the elderly population based on the built-in human model. Data from clinical PK studies were used to optimize and validate the model. The goodness of fit， relative residual， and mean folding error （MFE） were used to evaluate the performance of forecasting. The final model was employed to simulate the exposure of amikacin in the elderly population with renal insufficiency， and the efficacy and safety of commonly used clinical dosing regimens were evaluated， and the recommended regimens were proposed. RESULTS The established PBPK model of amikacin had good prediction performance in both adult and elderly populations， with the absolute mean of relative residual value of 25%； the MFE of peak concentration （cmax） and area under the plasma concentration curve （AUC0－∞） in all simulation occasions ranged ＞0.5-＜2. The simulation results showed that， compared with healthy adults， no significant clinical difference in cmax was observed in the elderly with renal insufficiency at the same dosing regimen， but the trough concentration increased significantly due to accumulation. Prolonging the administration interval of amikacin rather than reducing the dosage was more helpful to ensure the efficacy and to reduce the occurrence of nephrotoxicity. CONCLUSIONS The PBPK model for amikacin is successfully established in the elderly patient with renal insufficiency， and shows good predictive performance.