The main constraints to the administration of aminoglycosides (AG) are risks of nephrotoxicity and ototoxicity, which can lead to renal and vestibular failure. AG accumulation in the kidney may be related to the dosing schedule. As a result, administration of larger doses on a less frequent basis may reduce the drug accumulation in the renal cortex. Many methods have been proposed to reduce AG nephrotoxicity. (1) Molecular modeling and analog synthesis could lead to intrinsically less toxic AG but this approach is time consuming and expensive. Protective approaches such as the coadministration of polyaspartic acid or defferoxamine appear to be very promising in clinical practice. (2) Population pharmacokinetic computer programs, used to control AG serum concentrations, are correct predictors of efficacy but the estimated concentrations in the second compartment are not reliable predictors of nephrotoxicity because they do not take into account non-linear processes such as the AG uptake in the renal cortex or the tubuloglomerular feedback. (3) Finally, modelling the AG nephrotoxicity with probabilistic approaches and / or with deterministic approaches seems to be very promising. These two approaches appear to be not competitive but very complementary in clinical practice. The probabilistic model can be used to predict nephrotoxicity at the beginning the treatment. The deterministic model can be used to simulate and control nephrotoxicity when it is already unfolding and the treatment must be given for a long period of time.
Keywords: aminoglycoside, nephrotoxicity, review, model, clinical, practice