Carbapenemase-Producing Aeromonas veronii Disseminated in the Environment of an Equine Specialty Hospital

Introduction

Carbapenem antimicrobials are an important last resort to treat multidrug-resistant invasive gram-negative infections (Webb et al. 2016). The World Health Organization has recently implemented an action plan to lessen the dissemination of antimicrobial resistance, with a focus on hospital and community settings as well as animal populations (Finley et al. 2013). The use of broad-spectrum antibiotics in healthcare settings facilitates selection pressure for the emergence and dissemination of antimicrobial resistance genes, including carbapenem-resistant Enterobacteriaceae (CRE) (Webb et al. 2016). Mechanisms of carbapenem resistance include production of carbapenemase that can hydrolyze extended-spectrum β-lactam antibiotics that are frequently encoded on transmissible plasmids, or reduced permeability of the outer cell membrane potentially in conjunction with the expression of outer membrane efflux pumps (Nordmann et al. 2012, Webb et al. 2016).

CREs have only rarely been reported in equine populations. CRE-expressing bla _OXA-23 and bla _OXA-48 were isolated from fecal samples of hospitalized horses in Belgium and Germany, respectively (Smet et al. 2012, Schmiedel et al. 2014). bla _OXA-23 is chromosomally mediated and dependent upon an ISAbaI promoter for upregulation and mobilization (Smet et al. 2012), whereas bla _OXA-48 is a conjugative plasmid-mediated carbapenemase gene (Poirel et al. 2012). However, CREs have not been reported in horses in the United States. Veterinary hospitals may provide selective pressure favoring commensal flora with mobile, plasmid-mediated resistance genes (Finley et al. 2013). Furthermore, regularly exposing bacteria to antibiotic selection pressure could accelerate the development of novel resistance mechanisms (Finley et al. 2013). Our objective was to determine the prevalence of CRE in the environment of an equine hospital. We report here carbapenem-resistant organisms recovered from environmental surfaces in an equine specialty veterinary hospital.

Materials and Methods

Samples were collected from surfaces commonly contacted by staff or patients in a large referral equine hospital in Kentucky using electrostatic cloths. The same surfaces and locations were sampled over seven visits (May 2015–April 2016). For smaller locations (drains, door handles, equipment handles, etc.), the entire surface was sampled. For larger locations (stalls, preparation counters, surgical suites, etc.), a composite sample was obtained representing ∼25% of the location focusing on frequently contacted areas. Environmental samples were incubated in nutrient broth with 2 μg/mL cefotaxime at 37°C and then inoculated onto MacConkey agar with 0.5 μg/mL meropenem and 70 μg/mL ZnSO₄. Lactose-positive isolates were tested for carbapenemase enzymes using the Carba NP test (Nordmann et al. 2012). Minimum inhibitory concentrations (MICs) to standardized panels of antimicrobial drugs were generated using a semiautomated broth microdilution system (CMV3AGNF and ESB1F MIC panels; Thermo Fisher Scientific, Oakwood Village, OH) following Clinical and Laboratory Standards Institute guidelines (CLSI 2014). Plasmid content and size were determined by electrophoresis using a standard plasmid profiling procedure (Kado and Liu 1981). Identification of isolate species, antimicrobial resistance genes, and multilocus sequence type was obtained by whole-genome sequencing (Illumina MiSeq, San Diego, CA). Genetic analysis of CRE plasmid replicons was performed using PlasmidFinder 1.3 (Center for Genomic Epidemiology).

Results

Four carbapenemase-producing isolates termed B1-V3, B1-V4, B9-V5, and FB-V7 were recovered from 315 surfaces (1.3%). Isolates B1-V3 and B1-V4 were obtained from the floor drains of an all-purpose/adult patient ICU barn. Isolate B9-V5 was obtained from a stall in a barn used for patients with colic and other gastrointestinal disease and isolate FB-V7 was found on the floor drains of a barn demarcated for ill mare/foal pairs. All isolates were identified as Aeromonas veronii, a gram-negative, facultative anaerobic environmental bacterium and opportunistic pathogen commonly associated with aquatic reservoirs (Igbinosa et al. 2012). Although all four isolates were confirmed to produce carbapenemase, the corresponding imipenem and meropenem MICs were within the susceptible range (Table 1).

The four A. veronii isolates had identical multilocus sequence types using an online database (Center for Genomic Epidemiology) with a previously unreported allelic profile: grol_128 (100%), glta_155 (99.39%), gyrb_14 (98.53%), metg_172 (98.21%), ppsa_107 (97.21%), and reca_194 (98.57%). NCBI protein sequence search revealed that all four isolates harbored the metallo-β-lactamase gene bla _cphA capable of hydrolyzing carbapenem antibiotics. Multiple other genes conferring resistance to β-lactam antibiotics were also identified, including bla _TEM-1, bla _OXA-2, bla _OXA-12, and bla _FOX-5 (Table 1). Aminoglycoside, chloramphenicol, and sulfanomide resistance genes were also detected (Table 1). Plasmid electrophoresis (Kado and Liu 1981) revealed five plasmids for isolates B1-V3 and B1-V4 as well as six plasmids for B9-V5 and FB-V7 ranging from 3.8 to 210 kb. Genetic analysis of CRE plasmid replicons identified no known plasmid replicon types in the A. veronii isolates.

Discussion

These results suggest that a single clonal carbapenemase-producing A. veronii was transferred between hospital locations, likely through movement of patients or staff. Antimicrobial use within this hospital may have provided selection pressure favoring the persistence of these carbapenemase-producing Aeromonas spp. Aeromonas spp. are not a common equine pathogen but have caused infections in both healthy and immunocompromised hosts because of ingestion of contaminated food or water (Igbinosa et al. 2012) and have been recently identified as a pathogen for humans and animals within the past decade (Rossolini et al. 1995). It has been suggested that Aeromonas spp. are an overlooked cause of diarrhea in horses because of its higher isolation from diarrheic horses (55%) than Salmonella (20%) (Hathcock et al. 1999). Aeromonas spp. may also be correlated with diarrhea in young foals or enteric lesions (Hathcock et al. 1999). As such, minimizing patient exposure to Aeromonas spp. is essential to maintaining a sanitary clinical environment.

The carbapenemase gene associated with these A. veronii isolates, bla _cphA, is normally located on the bacterial chromosome. This reduces the risk of rapid horizontal dissemination of resistance compared with plasmid-mediated genes, but these organisms can pose a threat to individual patients. Our recovery of these four carbapenemase-producing A. veronii suggests that carbapenem resistance can be present in veterinary hospital environments posing a risk to patients. Nosocomial infections are possible because of the high-density patient population typical of healthcare settings and the immunodeficient status of many hospitalized horses. Environmental contamination also poses a health hazard for veterinary staff and any visiting personnel to the contaminated area, as previous cases have shown that ingestion of Aeromonas spp. can cause bacterial infection in healthy individuals (Igbinosa et al. 2012). The presence of carbapenemase-producing bacteria in an equine hospital environment highlights the need for regular surveillance and risk assessment for both patients and practitioners to prevent CRE dissemination and nosocomial infection in veterinary hospitals.

Carbapenemase-producing Aeromonas spp. isolates, harboring bla _cphA-like genes, have also been recovered from enteric flora of dairy cattle (Webb et al. 2016). bla _cphA is an intrinsic chromosomal adaptation of most wild-type Aeromonas, but many of the wild strains are susceptible to carbapenem antimicrobials because of inexpression of enzymatic activity (Rossolini et al. 1995). However, Aeromonas species have been observed to overexpress β-lactamase and carbapenemase production when under selective pressure (Ko et al. 1998), such as clinical antibiotic use. Our recovery of bla _cphA2-harboring A. veronii is the first report of carbapenemase-producing bacteria in the environment of an equine veterinary hospital. However, the low recovery rate (1.3%) suggests that environmental contamination within this equine hospital is uncommon. It is likely that selection pressure resulting from the clinical use of carbapenems favors the survival of Aeromonas that overexpress carbapenemase genes. This suggests that continued appropriate hygiene, including effective cleaning and disinfection techniques, is needed to lower the risk to patients and personnel at this hospital.