Long-term carriage of NDM-1-producing Escherichia coli

Sir, The carbapenemase NDM-1 was initially identified in Escherichia and Klebsiella pneumoniae in 2009, and subsequently in many enterobacterial species worldwide, predominantly from patients hospitalized in India, Pakistan and Bangladesh.1 While most reports to date have indicated nosocomial acquisition of NDM-1 producers,1 we have also reported community acquisition of an NDM-1 producer.2 This case involved a patient who, when hospitalized in France in early 2010, was found to be colonized on her skin by an NDM-1-producing Escherichia coli. Although the patient had been living in Darjeeling, India, there was no prior history of hospitalization in that country.2 The source of colonization of this patient was not identified, but recent reports have demonstrated the extensive isolation of NDM-1 from tap and environmental water in New Delhi,3 leading us to speculate that exposure to contaminated water may account for this case. We report here on the long-term follow-up of this patient over a period of 13 months, from initial hospitalization until her death. Screening of this patient using rectal swabs plated on selective culture medium (ChromID ESBL, bioMérieux, La-Balme-Les-Grottes, France), as described previously,4 yielded regular positive samples. The patient had received two courses of antibiotics comprising co-amoxiclav (3 g daily) for 10 days at the time of identification of the colonization in March 2010 and then gentamicin (250 mg daily) in an attempt to treat a urinary tract infection just prior to her demise. In our view, those courses of antibiotic treatment are unlikely to have generated sufficient selective pressure to account for the persistence of the NDM-1-positive E. coli in the intestinal flora of the patient for >1 year. Indeed, such long-term persistence of E. coli in the environment and in the intestinal flora is already known.5,6 Interestingly, the NDM-1-positive E. coli isolated from this patient was of the ST131 type, which is of the sequence type mainly associated with the dissemination of the prevalent clavulanic acid-inhibited extended-spectrum β-lactamase CTX-M-15 worldwide.7 The case we report here indicates the long-term persistence of NDM-1-positive bacteria in the intestinal flora. This sustained level of carriage may be considered as a further risk factor for the dissemination of NDM-1 producers, taking into account that up to 108E. coli per gram of faeces are commonly found in humans. This observation also further underlines the urgent need to screen for carriers worldwide4 and the fact that colonized patients should be kept in strict isolation during their entire hospital stay. This work was funded by the INSERM (U914), France. None to declare.