Pharmacodynamics of piperacillin/tazobactam against Pseudomonas aeruginosa: antibacterial effect and risk of emergence of resistance

Abstract

Full text

Piperacillin/tazobactam has a broad spectrum of antibacterial activity, which makes it suitable therapy for severe sepsis in the hospital setting, especially in critical care.^1^ Despite this, there remains a paucity of published preclinical pharmacodynamic (PD) characteristics of piperacillin/tazobactam to aid our optimization of dosing.^2^ Studies indicate that the pharmacokinetic/PD (PK/PD) index of effectiveness for β-lactams is the percentage of time that the free drug concentration remains greater than the MIC, fT>MIC.^3^ A PD index target of fT>MIC ≥ 50% is widely used to define antibacterial breakpoints for piperacillin/tazobactam, but the appropriate PD index targets for critical care patients is less clear.^4^ In critical care, alternative β-lactam targets have been employed, such as exceeding the β-lactam trough concentration by 4 ×  the pathogen MIC.^5^ PD factors have also led to the adoption of prolonged or continuous infusion therapies of piperacillin/tazobactam to treat aerobic Gram-negative rods and, most especially, Pseudomonas aeruginosa infection.^6–8^ However, there are burgeoning concerns regarding these ‘supratherapeutic’ doses, with the prevalence of β-lactam toxicity in the critical care environment thought to be underestimated.^9,10^ Here we aim to increase our preclinical data and evaluate the activity of piperacillin/tazobactam against P. aeruginosa, focusing on the fT>MIC target for antibacterial effect (ABE) and emergence of resistance (EOR).

Four clinical strains of P. aeruginosa were used: 45966 (piperacillin/tazobactam MIC 4 mg/L); 46042 (MIC 6 mg/L); 27853 (MIC 4 mg/L); and 46172 (MIC 6 mg/L). MICs were obtained following EUCAST guidelines.^11^ Clinical strains were obtained from North Bristol NHS Trust and tested alongside the QC ATCC strain 27853 (MIC 4 mg/L). fT>MIC dose ranging from 0% to 100% fT>MIC for piperacillin/tazobactam was tested against these strains in a dilutional single compartment in vitro PK model (IVPKM). A pharmacy preparation of piperacillin/tazobactam (4 g/0.5 g; Milpharm, UK) was used for both dose ranging and piperacillin/tazobactam dose escalation simulations, with t½ = 1 h, 8 hourly, and a minimum of eight experiments per strain performed to simulate 0%–100% fT>MIC. The fT>MIC ratios were based on the piperacillin concentrations with a ratio of 1:0.125 piperacillin:tazobactam. Mueller–Hinton broth (100%; Thermo Fisher, Basingstoke, UK) was used in all experiments. Nutrient agar plates were used to recover cfu enumerations per timepoint, 0–8 h, and every subsequent 24 h increment per IVPKM/per simulation, and blood agar plates were poured containing multiples of piperacillin/tazobactam MIC, measured in mg/L every 24 h increment from T0 to determine EOR. The inoculum was 10^6^ cfu/mL and experiments were performed over 72 h.

The relationship between fT>MIC and ABE were described by log reduction in viable count at 24, 48 and 72 h. This was done using a Boltzmann sigmoid Emax equation with the software package GraphPad Prism: y = bottom + (top−bottom)/{1 + exp(V_50_−x)/slope]} (San Diego, CA, USA). EOR was assessed by changes in population profile from baseline at time 0 and 24 h by culture onto media containing ×4 and ×8 piperacillin/tazobactam MIC. Viable counts (log cfu/mL) were determined over the 72 h of piperacillin/tazobactam exposure.

At 24 h, the piperacillin/tazobactam fT>MIC for ABE static effect was 39.8% ± 7.6%, for a −1 log drop was 51.7% ± 11.7%, and for a −2 log drop was 61.6% ± 17.7%; −3 log drop was not achieved in two of the four strains. At 48 h, fT>MIC for static effect was 76.5% ± 14.2%, for a −1 log drop was 81.7% ± 10.2%, −2 log drop was not achieved (>100%), and for a −3 log drop was >100%. At 72 h, fT>MIC for static effect was 86.6% ± 3.1%, for a −1 log drop was 89.5% ± 4.4%, −2 log drop was not achieved (>100%), and for −3 log drop was >100% (Table 1).

The risk of EOR, as indicated by the recovery of resistant isolates on ×4 MIC or ×8 MIC plates exhibited an inverted U pattern, demonstrating that lower doses exhibit less selective pressure. Initially the lower dose produced a small amount of resistance, and as this dose increased so did the EOR; this then continued up to a critical point whereby the bacteria were overcome by the higher concentrations of piperacillin/tazobactam. At 24 h, EOR steadily rises with the maximum risk of EOR being at an fT>MIC of 40%–60%, then decreases at higher targets, with fT>MIC of >60%–80%. The complete population profiles for ×4 MIC and ×8 MIC recovery plates are shown in Table S1 (available as Supplementary data at JAC-AMR Online).

In conclusion, an fT>MIC of 50% for piperacillin/tazobactam is associated with a −1 log reduction in bacterial counts of P. aeruginosa after 24 h exposure. fT>MIC targets of >80% are required for static or −1 log kill over 72 h. fT>MIC in the range of 40%–60% maximally amplifies resistance at 24 h, decreasing at higher targets of fT>MIC > 60%.

Our in vitro data for piperacillin/tazobactam align well with the current tenet of a PD index target of fT>MIC ≥ 50% for reduction in bacterial counts. Less studied is EOR; here, selection of resistance is most pronounced in the range of 40%–60% fT>MIC , and so prevention of EOR in vitro requires targets of fT>MIC > 60%. This data suggest that while the conventional dose of piperacillin/tazobactam can be used successfully to treat P. aeruginosa infection, dose adjustment to achieve trough concentrations of >4 ×  pathogen MIC may be unnecessary. In addition, it may be associated with unintended toxicity.^5^ To prevent EOR in vitro requires targets of fT>MIC > 60%.

This was an in vitro experiment, and the authors acknowledge the challenges in translating this to an in vivo/clinical recommendation.

Supplementary Material

dlae108_Supplementary_Data