Infections due to extended-spectrum beta-lactamase (ESBL)-producing Enterobacteriaceae are a major public health concern, in particular in the intensive care unit (ICU), due to the increase in their incidence. Carbapenems are the treatment of choice of these infections, but their increased use may select for carbapenem resistance in Gram-negative bacilli, which currently represents the greatest threat in terms of antibiotic resistance. Several retrospective studies have shown that the use of non-carbapenem antibiotics (mainly the association of piperacillin/tazobactam, but also cefepime and temocillin) may be safe alternatives to carbapenems to treat these pathogens when the strain is susceptible to the corresponding antibiotic. However, one recent randomized controlled study, the Merino trial, failed to demonstrate the non-inferiority of piperacillin/tazobactam, as compared to meropenem, in patients with Gram-negative bacilli bacteremia resistant to third generation cephalosporins (mainly ESBL producers). However, the patients included in that study were not ICU patients, dosing and modalities of piperacillin/tazobactam administration were not optimal (30-min infusion whereas 4-hours infusion may be associated with better outcome), and cause of death of patients in the piperacillin/tazobactam arm were not due to antimicrobial treatment failure (mostly death due to care withdrawal in cancer patients). Recently, a retrospective bicenter study performed in ICU patients showed that outcome of patients with severe infection (i.e. sepsis and septic shock according to the Sepsis-3 definition) due to ESBL-producing Enterobacteriaceae susceptible to non-carbapenem agents treated with a non-carbapenem agent was similar to that of patients treated with carbapenems.

Given the scarcity of data in ICU patients, the disputable results of the Merino trial, we will therefore conduct a multicenter, randomized, open-label trial of non-carbapenem beta-lactam (piperacillin/tazobactam or temocillin) treatment vs. meropenem treatment for ESBL-producing Enterobaceriaceae severe infection in ICU patients. Our hypothesis is that a non-carbapenem beta-lactam treatment is non-inferior to carbapenem treatment in patients with ESBL-producing Enterobacteriaceae severe infection in the ICU.

Inclusion Criteria:

• Patients ≥ 18-year-old
• Hospitalized in the ICU
• Severe infection, eg sepsis or septic shock (according to the Sepsis-3 definition) Sepsis is defined as life-threatening organ dysfunction caused by a dysregulated host response to infection, characterized by an increase of Sequential Organ Failure Assessment (SOFA) score of 2 points or more. This increase in 2 points will be calculated the day infection is diagnosed (day of positive culture serving as reference for the infection) as compared to the day before infection onset.

Septic shock is defined as sepsis and persisting hypotension requiring vasopressors to maintain mean arterial pressure ≥65 mmHg and having a serum lactate level >2 mmol/l despite adequate volume resuscitation.

This criterion (sepsis or septic shock) has to be fulfilled within a time frame of +/- 24 hours from the day of infection diagnosis (i.e. the day of positive bacteriological sample).

• Pathogen responsible for infection is an ESBL-producing Enterobacteriaceae susceptible to meropenem and either to piperacillin/tazobactam (minimum inhibitory concentration <8 mg/L) or to temocillin (minimum inhibitory concentration ≤8 mg/L)
• Signed Informed consent from patient/a legal representative/a family member/a close relative. According to the specifications of emergency inclusion, randomization without the close relative or surrogate consent could be performed if the patient is unable to give his/her consent and when the legal representative/family member or close relative are absent except patients included in another study for which emergency inclusion has already been used (see exclusion criteria n° 8). For these patients, emergency inclusion cannot be used). Close relative/surrogate/family consent will be asked as soon as possible. The patient will be asked to give his/her consent for the continuation of the trial when his/her condition will allow
• Affiliation to social security (AME excluded)

Exclusion Criteria:

• Pregnancy or breastfeeding
• Known allergy to beta-lactam
• Patient with severe neutropenia, as defined by absolute neutrophil count <0.5x109/L
• Infection requiring prolonged antimicrobial treatment (endocarditis; mediastinitis; osteomyelitis/septic arthritis; undrainable/undrained abscess; unremovable/unremoved prosthetic-associated infection)
• Moribund, defined by a SAPS II score at inclusion >75
• Decision of withholding/withdrawing care
• Patient with concomitant infection requiring antibiotics with activity against Gram-negative bacilli, including patient with polymicrobial infection with pathogen resistant to study drugs
• Participation in another interventional study evaluating drugs or being in the exclusion period at the end of a previous study evaluating drugs .
• Hypersensitivity to any components of the formulations