Resistance suppression by high-intensity, short-duration aminoglycoside exposure against hypermutable and non-hypermutable Pseudomonas aeruginosa

Abstract

Objectives: Hypermutable bacteria are causing a drastic problem via their enhanced ability to become resistant. Our objectives were to compare bacterial killing and resistance emergence between differently shaped tobramycin concentration-time profiles at a given fAUC/MIC, and determine the tobramycin exposure durations that prevent resistance.

Methods: Static concentration time-kill studies over 24 h used Pseudomonas aeruginosa WT strains (ATCC 27853 and PAO1) and hypermutable PAOΔmutS. fAUC/MIC values of 36, 72 and 168 were assessed at initial inocula of 10⁶ and 10⁴ cfu/mL (all strains) and 10^1.2 cfu/mL (PAOΔmutS only) in duplicate. Tobramycin was added at 0 h and removed at 1, 4, 10 or 24 h. Proportions of resistant bacteria and MICs were determined at 24 h. Mechanism-based modelling was conducted.

Results: For all strains, high tobramycin concentrations over 1 and 4 h resulted in more rapid and extensive initial killing compared with 10 and 24 h exposures at a given fAUC/MIC. No resistance emerged for 1 and 4 h durations of exposure, although extensive regrowth of susceptible bacteria occurred. The 24 h duration of exposure revealed less regrowth, but tobramycin-resistant populations had completely replaced susceptible bacteria by 24 h for the 10⁶ cfu/mL inoculum. The hypermutable PAOΔmutS showed the highest numbers of resistant bacteria. Total and resistant bacterial counts were described well by novel mechanism-based modelling.

Conclusions: Extensive resistance emerged for 10 and 24 h durations of exposure, but not for shorter durations. The tobramycin concentration-time profile shape is vital for resistance prevention and should aid the introduction of optimized combination regimens.

Figure 1.

Model diagram of the life cycle growth model including three populations, susceptible, intermediate and resistant, with two states each to describe bacterial replication. The maximum killing rate constants (K_max) and the antibiotic concentrations (KC₅₀) causing 50% of K_max are explained in Table 2. The concentration of tobramycin in broth (C_broth), intracellular tobramycin (C_intra), ADAPT₁, ADAPT₂ and all corresponding rate constants are described in the Materials and methods section.

Figure 2.

Observed viable counts (mean ± SD) and population predicted profiles (continuous lines in corresponding colours) for P. aeruginosa ATCC 27853 (a), PAO1 (b) and PAOΔmutS (c) exposed to tobramycin at an fAUC/MIC of 36 (left column), 72 (middle column) and 168 (right column) delivered over 1, 4, 10 or 24 h durations of exposure, at initial inocula (cfu_o) of 10⁶ and 10⁴ cfu/mL, excluding fAUC/MIC of 36 for ATCC 27853 and fAUC/MIC of 168 for PAOΔmutS for cfu_o 10⁴ cfu/mL. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Figure 3.

Log₁₀ PRB (mean ± SD) at 24 h on agar plates containing 1.25 (1.5 for ATCC 27853; a), 2.5 (b) and 5 (c) mg/L tobramycin. This figure shows fAUC/MIC values of 36 (left column), 72 (middle column) and 168 (right column) for the P. aeruginosa ATCC 27853, PAO1 and PAOΔmutS delivered over 1, 4, 10 or 24 h durations of exposure, at initial inocula (cfu_o) of 10⁶ and 10⁴ cfu/mL, excluding fAUC/MIC of 36 for ATCC 27853 and fAUC/MIC of 168 for PAOΔmutS for cfu_o 10⁴ cfu/mL. When a treatment yielded complete eradication or when there were no colonies on antibiotic-containing agar plates no PRB could be determined, and therefore is shown below the limit of quantification as −9.5 log₁₀.