Fecal carriage and molecular epidemiology of carbapenem-resistant Enterobacteriaceae from outpatient children in Shanghai

Background Fecal colonization with carbapenem-resistant Enterobacteriaceae (CRE) is a risk factor for bacterial translocation resulting in subsequent endogenous infections. The purpose of this study is to investigate the prevalence of CRE strains colonization in stool samples of outpatient in a tertiary pediatric hospital of Shanghai, China. Methods In a retrospective study, fecal samples were consecutively obtained from patients in 2016 and screening test for CRE was conducted by using home-made MacConkey agar. Antimicrobial susceptibility was determined by the broth microdilution method and β-lactamases were characterized by polymerase chain reaction (PCR) assays and DNA sequencing. Multilocus sequence typing (MLST) was performed for the genetic relationships of the isolates. Results A total of 880 fecal samples were included for this screening test and 32 CRE strains were identified in 32 non-duplicate fecal samples from 32 children (1.3 ± 1.5 years), with a carriage rate of 3.6%. These strains mainly distributed in Klebsiella pnuemoniae (37.5%) and Escherichia coli (37.5%). All CRE strains showed high resistance to most of the routinely used antibiotics (> 90%) except for polymyxin B and tigecycline. The blaNDM gene was the major carbapenemase gene harbored by gastrointestinal CRE strains, followed by blaKPC-2, blaIMP-26, and blaIMP-4. Other β-Lactamase genes including blaCTX-M, blaSHV, blaTEM-1, and blaDHA-1 were also detected. MLST analysis revealed that various sequence types (STs) were detected in these strains, with ST11 and ST37 being more prevalent in K.pneumoniae and ST101 in E.coli. Conclusions This study revealed the prevalence of CRE fecal carriage in children from outpatient and urgent implementation of infection control measure should be conducted to limit the spread of CRE strains.


Background
Carbapenems such as imipenem and meropenem are often recommended as antimicrobial agents of the last resort, especially in cases where extended-spectrum β-Lactamase (ESBL) producing organisms involved [1]. Since the KPC1 carbapenem-resistant Klebsiella pneumoniae was first isolated in the United States of America [2], the prevalence of carbapenem-resistant Enterobacteriaceae (CRE) has emerged as a public health problem and given rise to epidemic outbreaks in hospital surroundings all around the world [3][4][5]. According to the CHINET surveillance of bacterial resistance data in China, CRE strains showed high levels of resistance to most antibiotics and the rates of carbapenem resistance in E. coli and K. pneumoniae had increased from 0 and 2.9% in 2005 to 1 and 13.4% in 2014, respectively [6]. In general, the mechanism of resistance to carbapenems is mainly associated with the production of carbapenemases.
Enterobacteriaceae are a family of bacteria that mainly colonize in normal human intestines. Gastrointestinal tract colonization with CRE strains can be a risk factor for bacterial translocation resulting in subsequent endogenous infections in immunocompromised children [7]. As Das et al. reported that neonates with Gram negative bacilli in the gut had a higher incidence of clinical sepsis than those without these bacteria [8]. The acquisition of mobile genetic elements may be involved in the horizontal transmission. Several studies have been conducted to investigate the fecal carriage characteristics of CRE strains among inpatients [9][10][11], however, there are few data regarding the prevalence of CRE colonization in community settings, especially in China. Therefore, this research retrospectively aimed to conduct a surveillance of molecular epidemiology of CRE strains colonized in stool samples of outpatient children in a tertiary pediatric hospital of Shanghai, China.

Specimen collection and bacterial screening
This study was retrospectively conducted in 2016 from January to December at Shanghai Children's Hospital, which is one of largest pediatric hospitals with 700 beds. Epidemiological information including patient demographics, prior hospitalization, previous receipt of antibiotic therapy, and invasive operation during hospitalization was obtained from the medical records of each patient.
Fecal samples were consecutively obtained from patients of ≤18 year old attending in outpatient clinics, who received the fecal culture testing. Approximately 0.5 mg of stool or a rectal swab from each patient was stored in Cary-Blair transport medium (Hopebio, Qingdao, China) in 4°C condition up to 1 days prior to testing. Screening test of CRE was performed as follows: fecal samples were inoculated in home-made MacConkey agar supplemented with meropenem at 1 μg/ml and then incubated at 35°C in a 5% CO2 incubator. After incubation for 24-48 h, the growth of at least one colony per agar plate was considered as positive screening result and the strain was further investigated. All CRE strains were stored at − 80°C in 40% glycerol broth medium for further analysis.

Identification and antimicrobial susceptibility testing
All CRE bacteria were identified by matrix-assisted laser desorption ionization time of flight (MALDI-TOF) mass spectrometry using MALDI Biotyper (Bruker Daltonik GmbH, Bremen, Germany). Antimicrobial susceptibility testing of all CRE strains against amikacin, cefotaxime, ceftazidime, cefepime, piperacillin-tazobactam, ertapenem, imipenem, meropenem, ciprofloxacin, fosfomycin, aztreonam, colistin, and tigecycline was carried out by using broth microdilution method for determining the minimal inhibitory concentrations (MICs). The breakpoints used for interpretation were recommended by the Clinical and Laboratory Standards Institute (CLSI) 2017 [12]. Quality control was managed by using E.coli ATCC 25922. The EUCAST breakpoints were used for colistin. The interpretive criterion for tigecycline was based on the breakpoints of the Food and Drug Administration (FDA).

Statistical analysis
Statistical analysis was performed by using SPSS 19.0 for Windows (version 19.0; SPSS Inc., Chicago, IL, USA). The quantitative data were expressed as means ± standard deviations (SD). Categorical variables, expressed as numbers and percentages, were compared by the Chi-square or Fisher's exact test. A value of P ≤ 0.05 was considered statistically significant. Antimicrobial resistance data was analyzed by using WHONET 5.6.
The CRE strains were isolated from 32 individual children (19 male and 13 female) whose mean age was 1.3 ± 1.5 years (rang: 1 m-5y). Among these children, 28.1% (9/32) had a history of prior hospitalization in recent 3 months and these patients also had received therapy with antibiotics (e.g. imipenem, cefperazone-sulbactam, ampicillin/sulbactam, or cefotaxime). Of the 32 CRE carriers and 848 non-CRE carriers identified, 9 (28.1%) and 38 (4.5%) children, respectively, had been hospitalized and received antibiotics within the past 3 months (P < 0.05). Among nine previously hospitalized CRE carrier children, only two children had received invasive operation during their hospitalization.

Antimicrobial susceptibility
The results of the antimicrobial susceptibility testing of the 32 CRE strains are shown in Table 1. All CRE isolates showed high resistance to cephalosporins and carbapenems (> 95%). The rates of susceptibility to amikacin, ciprofloxacin, aztreonam and fosfomycin were 75, 25, 21.6, and 15.6%, respectively. Tigecycline retained excellent activity against all 32 CRE isolates, with a susceptibility rate of 100%, regardless of the species. Additionally, the CRE strains showed a high susceptibility rate to colistin (90.6%). However, the MIC ≥4 μg/mL of colistin was observed in E.coli and E.cloacae isolates.

Characteristics of STs
MLST analysis revealed that STs distributed dispersedly in different species colonized in fecal samples from community and a total of twenty distinct STs were identified among 32 CRE isolates, with various STs in different species. As depicted in Table 3, nine STs were observed among carbapenem-resistant E.coli and ST101 was main type, accounting for 33.3%. We also found seven STs in carbapenem-resistant K.pneumoniae isolates, with ST11 being the major type (Table 2). Then four STs were identified in the six E.cloacae isolates.

Discussion
Investigation of the fecal carriage prevalence of CRE among outpatients from the community setting can help us to better understand the origin of CRE isolates responsible for outbreak events and contribute to control CRE dissemination. Our study displayed that the rate of colonization with CRE in fecal sample collected from children in outpatient visits was 3.6%, which was lower than what had been reported in patients from community setting in other countries [10,19]. Several reasons contribute to this phenomenon of the colonization of CRE isolates in rectal gastrointestinal tract. Firstly, further analysis of patient information of CRE carriers showed that 28.1% of them had a history of prior hospitalization in recent 3 months and also had received therapy with antibiotics, but only 4.5% of non-carriers had. Previous studies reported that exposure to hospital setting or/and antimicrobial agents might increase the risk of colonization and it would be easier to acquire CRE isolates, which may also explain the occurrence in community-onset cases [20,21]. However, despite the previous studies showed an increased risk of CRE carriage with hospital exposure, there is no significance between the min, max and mean times of hospitalization for the CRE and non-CRE groups with recent hospital exposure in this study. What is more, many outpatients receive antimicrobial treatment without the advice or prescription of a physician or other trained health care provider. Last but not least, CRE strains could spread via physical contact with other people and have the propensity to acquire genetic materials mostly in the form of plasmids and transposons through horizontal gene Tigecycline 0 (0) 0 (0) 100 (32) transfer [22]. Colonization with CRE strain plays an important role in transmission of the orgnism through horizontal gene transfer in health care settings [22]. The surveillance definition for CRE are designed that Enterobacteriaceae are resistant to imipenem, meropenem, doripenem, or ertapenem or is document that the isolate possess a carbapenemase and CRE are reported to be commonly identified in K.pneumoniae, E.coli, and E.cloacae [23,24]. The major resistance mechanism of CRE is the production of carbapenemases such as NDM, KPC, and IMP, which were also observed in this study. The NDM type, including NDM-1 and NDM-5, is found to be the key carbapenemase responsible for mediating development of the carbapenem resistance phenotypes in children [3,25]. NDM-1 and its minor variants, a class B carbapenemase first clinically isolated from a patient at a hospital in New Delhi, India, has since been identified all over the world and only detected in E.coli and K.pneumoniae [26][27][28][29]. Consistent with the above findings, most of the E.coli and K.pneumoniae colonized CREs in the fecal samples in the present study harbored the bla NDM gene and a previous study in our hospital also has reported the outbreak of CRE strains caused by NDM-1 producing K.pneumoniae among neonates [3].  NDM-1 producing K.pneumoniae are highly resistant pathogens with no effective beta-lactams, including recent ones such as ceftolozane tazobactam and ceftazidime avibactam and the only one that works is aztreonam [30]. Additionally, only three K.pneumoniae strains carried bla KPC-2 gene, which was mostly prevalent in adults population [31], were found in this study. IMP type carbapenemases (IMP-4 and IMP-26) were observed in E.cloacae and K.pneumoniae, but not in E.coli. However, it is remarkable that no carbapenemase resistance genes were detected in 9.4% (3/32) of the CRE strains. This finding suggests that a new mechanism might contribute to the resistance to carbapenems and further studies into this mechanism should be conducted. Molecular analysis revealed a high level of genetic diversity, including ST101, ST11, ST37, and ST48, and several STs are clearly related to specific bacteria, and are prevalent all over the world. The most common STs of CRE in K.pneumoniae were ST11 and ST37. Interestingly, the carbapenem-resistant K.pneumoniae harboring bla KPC-2 belonged to ST11, which exhibited signs of multi-clonal dissemination and caused a series of outbreaks in China [32,33]. Additionally, another carbapenem-resistant K.pneumoniae harboring bla NDM-1 belonged to ST37 and this ST had been identified previously during an outbreak of this phenotype in our hospital [3]. Clonally diverse carbapenem-resistant E.coli (ST405, ST131, ST156 and ST101) have been reported globally, and ST101 was the predominant ST isolated in this study. The ST101 type of E.coli produced NDM-1 carbapenemase, a phenotype that has been linked to nosocomial transmission in Korea [34].
Importantly, children with CRE strains in fecal samples are considered as a high risk group by World Health Organization (WHO), which can spread CRE by intimate contact and travel [35]. The origin of CRE isolates in these children remains unknown and it is the main limitation that we don't know if the source of the described CRE carriage in these children is the result of transmission in the community or in the hospital. Screening for CRE at discharge may provide an explanation for this hypothesis. Exposure to antimicrobial agents in both the hospital setting and environments is always the key to acquisition. Also we know that children are at high risk for multi-drug organism carriage and further studies should be carried out to evaluate the phenomenon of the original source of CRE strains.

Conclusions
In summary, the current data reveal the prevalence of CRE colonization in fecal sample from pediatric outpatients and strategies to control the dissemination of antimicrobial resistant isolates from stool to other sterilized sites should be developed. Children who had hospitalization exposure should be screened for CRE at discharge, which may help to definitively establish where acquisition of CRE is truly occurring and decrease the occurrence and transmission of CRE in children group.