Relationship between gentamicin susceptiblity criteria and therapeutic serum levels for Pseudomonas aeruginosa in mouse infection model.

In this study estimations of in vivo and in vitro gentamicin susceptibility for a series of strains of Pseudomonas aeruginosa were compared. The series included an extremely susceptible strain, typically susceptible strains by current susceptibility criteria, and strains with enzymatic and permeability-mediated resistance. In vivo testing was done by using a mice protection test involving six 1-h doses of gentamicin and an inoculum of 50 50% lethal doses of P. aeruginosa. Both normal mice and cyclophosphamide-treated mice were used. It was found that peak serum levels and serum levels of gentamicin obtained just prior to the sixth dose (fifth dose trough levels) required for protection were much higher than minimal inhibitory concentrations (MICs) or minimal bactericidal concentrations (MBCs) obtained in high-cation medium. However, first dose trough levels were similar to MICs or MBCs. Only an extremely susceptible strain, 280, could be treated at antibiotic dosages and serum levels which are considered likely to be safe in humans. A distinct inoculum effect was found in the mice tests, with a 10-fold increase in inoculum producing a 4-fold increase in the amount of gentamicin required, but no inoculum effect was found for MICs. These results suggest that current susceptibility criteria in use for gentamicin and P. aeruginosa overestimate gentamicin susceptibility, particularly when low-cation growth medium is used for susceptibility testing and when treating disseminated infection.