Carriage of multi-drug and extensive-drug resistant Gram negative bacteria (MDR-GNB) is associated with an increased risk of infections by these bacteria for the carriers and a high risk of dissemination both in the healthcare setting and the community; the main MDR-GNB reservoir is the fecal microbiota. To prevent both infections and dissemination, effective measures to decolonize subjects carrying MDR-GNB are urgently needed. Animal models, case reports and cohort studies suggest fecal microbiota transplantation (FMT) may be efficient for MDR-GNB decolonization.

The issue of antibiotic resistance is considered as one of the major Public Health threats of the 21th century by the World Health Organisation (WHO). Given the rapid increase in the dissemination of multi-drug resistant organisms globally, the real implications of spreading drug resistance will be felt the world over, with developing countries and large emerging nations bearing the brunt of this problem. Routine surgeries and minor infections will become life-threatening once again and the hard won victories against infectious diseases of the last fifty years will be jeopardised. Hospital stays and expenses will increase significantly. Drug resistant infections are already on the rise with numbers suggesting that up to 50,000 lives are lost each year to antibiotic-resistant infections in Europe and the US alone. Globally, at least 700,000 die each year of drug resistance in bacterial infections. Estimates have found that antibiotic resistance may kill more than cancer in 2050, with figures reaching 10 million deaths a year globally, if nothing is done.

In this context, one of the most worrying problems is the dramatic increase in infections caused by multi-drug resistant Gram Negative bacteria (MDR-GNB), including extended spectrum β-lactamase and/or carbapenem-resistant Enterobacteriaceae (ESBL-E and CRE) with few effective therapeutic options remaining in the armamentarium of clinicians. There is a striking lack of new antimicrobial agents especially against MDR-GNB while dissemination of these MDR-GNB is accelerating. The rising epidemic of CRE is especially worrying. Southern European countries such as Italy and Greece are already at epidemic levels of CRE infections, which are associated with much higher death rates and are more expensive to treat because of the lack of effective non-toxic antibiotics.

Even if new agents were discovered, the lead in time is considerable as there are no truly new agents expected on the market in the short or medium term. Alternative effective measures to contain resistance and limit the spread of MDR-GNB are therefore urgently needed. The gut microbiota is the main reservoir of MDR-GNB. Patients carrying MDR-GNB are at higher risk of clinical infections with their own bacteria and of dissemination in the community and in the hospital settings. In the last 10 years, the number of MDR-GNB carriers has increased strikingly (for instance, one report found a 10-fold increase in France from 2006-2011 from 0.6% to 6% of subjects in the community; another found a 12% ESBL-E colonization rate in Parisian hospitals upon admission in 2015) and continues to rise. Finding efficient decolonization strategies is urgent to attempt to limit the risk of infection at the individual level and the spread of MDR-GNB at the population level.

Fecal microbiota transplantation (FMT), which has been shown to be highly effective for the treatment of recurrent Clostridium difficile infections (CDI), has also been suggested as a decolonization strategy for patients carrying MDR-GNB: animal models and several clinical case reports suggest that this strategy may be interesting to decolonize MDR-GNB carriers. To date, the question of whether FMT may be efficient to decolonize MDR-GNB carriers in human is not resolved. Because decolonization may also be spontaneous, performing a randomized controlled trial is essential to answer this question.

The hypothesis for this study is that FMT may be efficient to decolonize patients carrying MDR-GNB in their gut microbiota. To determine whether this hypothesis is true, we shall conduct a phase III randomized controlled trial comparing capsule-delivered FMT with a placebo to decolonize subjects carrying either CRE or ESBL-E.

Elimination of carriage should benefit directly the individual patient by preventing infections with these resistant organisms and limiting isolation precautions, which impacts their quality of care, the possibilities of transfer to other healthcare settings (such as rehabilitation centres) and their psychological health. Additionally, the elimination of MDR-GNB carriage should benefit the community by decreasing the risk of inter-individual transmission in particular in the healthcare setting. For this intervention, we will use frozen capsules rather than nasogastric infusion or colonoscopy, as they limit adverse events, facilitate donor screening, allow for outpatient treatment, and are adapted to treat large populations. Thanks to the above-mentioned reasons, capsules are becoming the standard-of-care for FMT. In the context of the major threat that antibiotic resistance in Gram negatives carries, both on an individual and a collective point-of-view; this is a highly important question.

Inclusion Criteria:

Inclusion Criteria for patients:

• ≥ 18 years and < 105 years
• Patient with at least one positive rectal swab for enterobacteria:

extended-spectrum β-lactamase-producing Enterobacteriaceae (ESBL-E) or carbapenem-resistant Enterobacteriaceae (CRE), or who have had an ESBL-E or CRE infection within the year For ESBL-E carriers: an ESBL-E infection within the year is mandatory

- Patient able to take 50 capsules orally in a day and without swallowing disorders

Inclusion Criteria for healthy volunteers donors:

• Healthy subjects ≥ 18 years and < 50 years
• Body mass index < 30 kg/m^2
• Regular bowel movement defined as at least 1 stool every 2 daysand maximum than 3 stools per day

Exclusion Criteria:

Exclusion Criteria for patients:

• Current antibiotic treatment with te exception of long term antibiotic prophylaxis (duration of at least 3 months/year)
• Patients hospitalized in the intensive care unit
• Pregnancy or breastfeeding during the study
• Women of childbearing potential who are unwilling or unable to use an acceptable method of birth control [such as oral contraceptives, other hormonal contraceptives (vaginal products, skin patches, or implanted or injectable products), or mechanical products such as an intrauterine device or barrier methods (condoms)] to avoid pregnancy for the entire study
• Patient under legal protection
• Participation in another interventional study
• Non-affiliation to a social security scheme
• Patient under AME
• Refusal to participate to the study

Exclusion Criteria for healthy volunteers donors:

• Any history of or current proctologic disease or any acute condition, which in the investigator's judgment could harm the volunteer and/or compromise or limit the evaluation of the protocol or data analysis (for details, please see protocol)
• Subject under legal protection
• Participation in any other interventional study
• No-affiliation to a social security scheme
• Subject under AME
• Refusal to participate to the study

Randomization criteria:

• Patient colonized with a carbapenem-resistant Enterobacteriaceae (CRE) and/or colonized with an extended spectrum β-lactamase producing Enterobacteriaceae (ESBL-E) at inclusion on stool culture
• Patient with an ESBL-E infection in the previous 12 months (only for participants no colonized with CRE).
• Compatible transplant (FMT) based on patient's serological profile (CMV/EBV) available

Double-blinded placebo-controlled randomized multicentric controlled trial with a 1:1 randomization.