The post-antibiotic effect of antimicrobial combinations in a neutropenic murine thigh infection model.

The post-antibiotic effect (PAE) may allow for more widely spaced dosing of antibiotics than is currently employed without loss of efficacy. Antimicrobial combinations are widely used in clinical medicine. However, dosing schedules are usually based on pharmacological profiles of the drugs used alone. Previously we have demonstrated significant prolongation of the PAE induced by antimicrobial combinations in vitro as compared to PAEs induced by the agents alone. We examined this issue further in vivo in a neutropenic mouse thigh infection model, by exposing Staphylococcus aureus, Pseudomonas aeruginosa, Escherichia coli and Klebsiella pneumoniae to several antimicrobials, either singly or in combination. The PAE in vivo was defined as the difference in time needed for the organisms in the treated animals to grow 1 log10 as compared with controls after serum drug concentrations had fallen below the MIC. Drug concentrations exceeded the MIC for 1.2-3.2 h, but bactericidal activity occurred mainly during the first hour. When the agents were used singly a negative PAE was produced by ceftazidime against P. aeruginosa, a PAE of approximately 0 h by imipenem against E. coli and K. pneumoniae, a PAE of 2-4 h by cefazolin against S. aureus, gentamicin against E. coli and K. pneumoniae, and imipenem and tobramycin against P. aeruginosa, and a PAE of 6-7 h by gentamicin against S. aureus and rifampicin against P. aeruginosa. The beta-lactam/aminoglycoside combinations when used against S. aureus and P. aeruginosa prolonged the PAE by 1.0-3.3 h, compared with the longer of the individual drug PAEs, but no prolongation was observed against E. coli and K. pneumoniae. Ceftazidime reduced the PAE when used with tobramycin against P. aeruginosa. The long PAE of rifampicin against P. aeruginosa was 'carried over' to the combination, thus prolonging the growth suppression achieved by imipenem and tobramycin alone or in combination by 5.5-8.0 h. This effect on the PAE was additive only, and synergy was not observed. In conclusion, a potentially significant prolongation of the PAE by combination of drugs was observed in vivo, but only if both (or all) agents induced a PAE when used alone. The impact of this observation needs to be examined further in studies involving multiple and different dosing regimens in an infection model.