Impact of inoculum effect on in vitro activity of intermittent and continuous piperacillin-tazobactam infusion against ESBL-producing Escherichia coli in a hollow-fiber infection model

@article{giry2026impact,
  title = {{Impact of inoculum effect on in vitro activity of intermittent and continuous piperacillin-tazobactam infusion against ESBL-producing Escherichia coli in a hollow-fiber infection model}},
  author = {Marion Giry and Charles Baulier and Fabien Lamoureux and Manuel Etienne and Nicolas Nessi and François Gravey and Kévin Alexandre},
  journal = {Microbiology Spectrum},
  year = {2026},
  month = oct,
  doi = {10.1128/spectrum.01457-26},
  url = {https://doi.org/10.1128/spectrum.01457-26},
  note = {CC BY 4.0},
}

TY  - JOUR
TI  - Impact of inoculum effect on in vitro activity of intermittent and continuous piperacillin-tazobactam infusion against ESBL-producing Escherichia coli in a hollow-fiber infection model
AU  - Marion Giry
AU  - Charles Baulier
AU  - Fabien Lamoureux
AU  - Manuel Etienne
AU  - Nicolas Nessi
AU  - François Gravey
AU  - Kévin Alexandre
PY  - 2026
DA  - 2026/10/13
JO  - Microbiology Spectrum
DO  - 10.1128/spectrum.01457-26
UR  - https://doi.org/10.1128/spectrum.01457-26
AB  - Use of piperacillin-tazobactam (TZP) in the treatment of severe extended-spectrum β-lactamase (ESBL)-producing Escherichia coli (ESBL-EC) infections is controversial. Two features may explain treatment failures: the inoculum effect and suboptimal pharmacokinetic/pharmacodynamic (PK/PD) target attainment, particularly in critically ill patients. To study the combined impact of the inoculum effect and a PK/PD optimization on the in vitro efficacy of TZP against ESBL-EC infections. Five-day hollow-fiber infection model simulating TZP pharmacokinetics in critically ill patients was performed: two TZP infusion regimens (4/0.5 g q6h over 30 min, and 16/2 g q24h continuous infusion), three ESBL-EC clinical strains: 1146 (blaCTX-M-15, MIC = 2 mg/L), 3383 (blaCTX-M-15 + blaTEM-1, MIC = 4 mg/L), and 2262 (blaCTX-M-15 + blaOXA-1, MIC = 8 mg/L), and two initial inocula (5 and 7 log10 CFU/mL). Intermittent infusion was associated with systematic bacterial regrowth within 24 h, consistent with failure to reach the 100%fT >4× MIC threshold. Continuous infusion resulted in undetectable bacterial count for the strains 3383 (MIC = 4 mg/L, low inoculum) and 1146 (MIC = 2 mg/L, low and high inocula). For strain 3383, the inoculum effect was demonstrated by total bacterial killing in continuous infusion with low inoculum, whereas early regrowth (>10 log10 CFU/mL) with resistance emergence (MIC > 256 mg/L) was observed with high inoculum. Strain 2262 (MIC = 8 mg/L) exhibited systematic treatment failure regardless of infusion mode or inoculum size. Continuous infusion of TZP may be considered as a carbapenem-sparing agent in ESBL-EC infection with a TZP MIC ≤ 2 mg/L regardless of inoculum, or a TZP MIC of 4 mg/L at low inoculum. TZP should not be used for strains with an MIC of 8 mg/L or suspected OXA-1 co-production, irrespective of the infusion mode.  IMPORTANCE: Infections caused by extended-spectrum β-lactamase (ESBL)-producing Escherichia coli represent a growing therapeutic challenge, particularly in critically ill patients. Carbapenems remain the reference treatment, but their overuse drives resistance emergence. Piperacillin-tazobactam is a potential carbapenem-sparing agent, but its clinical efficacy remains debated, partly due to the inoculum effect (a reduction in antibiotic activity at high bacterial densities) and suboptimal dosing regimens. Using a hollow-fiber infection model simulating the pharmacokinetics of critically ill patients, this study demonstrates that continuous infusion of piperacillin-tazobactam can overcome the inoculum effect under specific conditions, defined by the MIC of the infecting strain and the estimated bacterial density. These findings support a stratified approach to piperacillin-tazobactam use in severe ESBL-producing E. coli infections, based on MIC, inoculum size, and β-lactamase profile, and may help clinicians identify patients who could safely receive piperacillin-tazobactam instead of a carbapenem.
SN  - 2165-0497
ER  -