OXA-48-producing ST372Escherichia coliin a French dog

Sir, In veterinary medicine, extended-spectrum cephalosporins (ESCs) are widely used, contrary to carbapenems, which have no legal indication. Consequently, enzymes conferring resistance to ESCs, such as ESBLs and plasmidic AmpCs (pAmpCs), have been repeatedly reported in animals, whereas carbapenemase-producing Enterobacteriaceae (CPE) are still uncommon in non-human sources and have been recognized only sporadically in domestic animals and wildlife in certain countries and continents.1–5 Moreover, these cases were largely observational and not reported from prevalence studies. Here, our aim was to ascertain the prevalence of faecal carriage of acquired resistance to ESCs and carbapenems in a large set of healthy cats and dogs in France. Rectal swabs of unrelated healthy cats (n = 227) and dogs (n = 166) attending the Veterinary School of Maisons-Alfort (suburb of Paris) for routine vaccination and/or general check-up were sampled between January and June 2015. Each owner filled in a short questionnaire about their own health and their animal’s health, and contacts and travels in the last 12 months. Most animals were young (median age of 10 months for cats and 4 years for dogs) and very few presented a history of hospitalization and/or antibiotic treatment within the last 12 months. Selective plates for the detection of ESBL/pAmpC producers (ChromID ESBL) and carbapenem-resistant isolates (CarbaSMART) were used (bioMérieux, Marcy-l’Étoile, France), and the positive isolates were identified and characterized molecularly. Six non-clonal isolates, as assessed by PFGE of Xba1-digested DNA and MLST, recovered from four dogs and two cats, were identified as Escherichia coli by MALDI-TOF (Table 1). Susceptibility testing by disc diffusion (www.sfm-microbiologie.org) allowed the detection of three ESBL and two AmpC phenotypes, whereas one isolate presented a reduced susceptibility to ertapenem while remaining susceptible to ESCs. This latter isolate was susceptible to colistin and tigecycline (Etest, bioMérieux), but presented MICs of imipenem, ertapenem and meropenem of 1.5, 0.75 and 0.75 mg/L, respectively. Molecular and epidemiological characteristics of the E. coli isolates investigated in this study NT, non-typeable; ND, not determined; NAL, nalidixic acid; STR, streptomycin; TMP, trimethoprim; GEN, gentamicin. Molecular and epidemiological characteristics of the E. coli isolates investigated in this study NT, non-typeable; ND, not determined; NAL, nalidixic acid; STR, streptomycin; TMP, trimethoprim; GEN, gentamicin. The three ESBL-producing E. coli originated from one dog and two cats and belonged to ST973, ST68 and ST141 (Table 1). As detected by PCR/sequencing, each isolate harboured a different blaCTX-M gene, i.e. blaCTX-M-1, blaCTX-M-15 and blaCTX-M-14, located on IncFIB, IncI1 and IncF plasmid types, respectively, according to the PCR-based replicon typing (PBRT; Diatheva, Fano, Italy) and PFGE-S1 gels followed by Southern blot using adequate probes. One cat had suffered from a respiratory tract infection in the last 12 months and had been treated with antibiotics (unrecorded molecule), whereas the two other animals presented no history of illness, travel or contact with people working in healthcare centres. The two AmpC-producing E. coli originated from dogs, presented a blaCMY-2 gene located on non-typeable plasmids and belonged to ST55 and ST963. Neither of these dogs presented any identified risk factor for being colonized with ESC-resistant bacteria. Altogether, this study shows a very weak prevalence of ESC-resistance carriage in cats (2/227, 0.9%) and dogs (3/166, 1.8%) in a large set of randomly chosen animals. The ertapenem-non-susceptible E. coli isolate belonged to ST372 and displayed a unique blaOXA-48 gene located on an ∼60 kbp IncL plasmid. The blaOXA-48 gene was located in a Tn1999.2 transposon as revealed by the disruption of the upstream IS1999 element by IS1R using PCR mapping. This isolate was recovered from a 3.5-year-old German shepherd dog that had been hospitalized for 3–8 days in the last 12 months (abdominal infection) and treated with amoxicillin/clavulanic acid. The owner had also been hospitalized in France for a short period of time, but had never travelled outside Europe. This is the first report of CPE in an animal in France. This finding adds to the very limited number of studies reporting CPE in animals worldwide. In particular, OXA-48-producing Enterobacteriaceae in animals have been reported only in four countries so far, i.e. in poultry in Lebanon and in dogs in Algeria, Germany and the USA.1–5 Of note, most reports were from pets, which may indicate a human-to-animal transfer through close contacts. In our study, the OXA-48-positive animal was living in France, where OXA-48 is not endemic in humans.6 This suggests that OXA-48 producers in animals might be underreported in OXA-48-endemic countries. The origin of the OXA-48 E. coli in this dog could not be traced. No blaCTX-M-15 gene was identified along with blaOXA-48, a combination that is, on the contrary, rather common in human isolates. Also, carbapenem-susceptible ST372 E. coli isolates have already been associated with human and canine infections.7 In conclusion, even though resistance to carbapenems is uncommon in animals, carbapenemase genes are associated with a high potential for dissemination and their prevalence in non-human sources may be underestimated, potentially even in countries or continents where CPE are not highly prevalent in humans. CPE in animals should be more thoroughly monitored worldwide in order to clarify the role of non-human settings as possible reservoirs of carbapenemase genes. This work was supported by the French Agency for Food, Environmental and Occupational Health & Safety (ANSES) and by the National EcoAntibio Action Plan funded by the Ministry in charge of Agriculture. None to declare.