Fecal carriage of extended-spectrum β-lactamase-and AmpC-producing Escherichia coli among healthcare workers

Introduction: Commensal E. coli can be considered a reservoir of genes coding for antibiotic resistance that may be transmitted in hospitals by healthcare workers (HCWs). This study aimed to determine the fecal carriage rate of extended-spectrum β-lactamase (ESBL)-producing E. coli among HCWs. Methodology: Stool samples were collected from 200 HCWs. Phenotypic screening for ESBL and AmpC β-lactamases was performed using disk diffusion and minimum inhibitory concentration methods followed by the combined disks test and double synergy differential test for confirmation. Multiplex polymerase chain reaction (PCR) was used to detect blaSHV, blaTEM, blaCTX-M, and CIT groups for AmpC genes. Results: Of 200 E. coli isolates, 100% were susceptible to imipenem, and 59 (29.5%) were resistant to one or more third-generation cephalosporins. By molecular analysis, 21% (42/200) were colonized by ESBL-producing E. coli, and 3% (6/200) were colonized by AmpCproducing E. coli. The blaSHV gene was the predominant ESBL gene, detected in 81.8% of the resistant E. coli isolates. Conclusions: These findings highlight the increase in fecal carriage of E. coli carrying ESBL and AmpC genes among HCWs, which may be one of the causes of the spread of ESBL-producing bacteria in hospitals and requires sound infection control measures. This is the first study of the fecal carriage rate of E. coli carrying AmpC genes in HCWs.


Introduction
E. coli-producing extended-spectrum β-lactamases (ESBLs) are considered one of the most multidrug resistant (MDR) pathogens in hospitals.Infections caused by ESBL-producing organisms have resulted in reduced rates of clinical and microbiological responses, longer hospital stays, and great hospital expenses.Gastrointestinal colonization is one of the most frequent reservoirs of infection [1].
The importance of detecting carriers of antimicrobial-resistant bacteria in patients and healthy people has recently increased.The increase in the proportion of carriers heightens the risk that other individuals will become carriers as a consequence of human-to-human transmission of resistant bacteria.Also, environmental enriching by the resistance gene pool may facilitate the acquisition of resistance mechanisms by susceptible bacteria [2].
Several studies have evaluated fecal carriage of ESBL-producing isolates from the stools of healthy individuals in community settings [3][4][5], but aside from one recent study [6], they did not specifically investigate ESBLs in healthcare workers (HCWs).E. coli that produces both ESBLs and AmpC has been increasingly reported worldwide.The AmpCproducing organisms can act as a hidden reservoir for ESBLs; the high-level expression of the AmpC βlactamases may mask the recognition of the ESBLs [7].
The present study aimed to determine the prevalence of fecal carriage of ESBL-producing E. coli among HCWs by conventional microbiological methods and on a molecular basis, and also to determine the prevalence of fecal carriage of AmpC βlactamases and combined ESBLs including AmpC βlactamases.

Methodology
The present study was conducted at Fayoum University Hospital.Fecal samples were collected from 200 HCWs accepted to participate in the study who had not been recently hospitalized and had no recent exposure to antibiotics (≥ 3 months); 46 (23%) isolates were from internal medicine and intensive care units (ICUs), 40 (20%) were from operating theaters, 39 (19.5%) were from surgical wards, 27 (13.5%)were from outpatients clinics, 22 (11%) were from pediatric wards, 17 (8.5%)were from obstetrics, and 9 (4.5%) isolates were from laboratory staff.The study was approved by Fayoum University's ethics committee.

Bacterial isolates
Fresh stools samples were collected (between January 2013 and July 2013) and transported to the Department of Medical Microbiology and Immunology.The samples were cultured on MacConkey agar plates (Oxoid, Basingstoke, UK) and then incubated for 18-20 hours at 37°C.E. coli was isolated and identified according to conventional microbiological methods [8].
Minimum inhibitory concentrations (MIC) of ceftazidime and cefotaxime at concentrations ranging from 0.025 to 500 μg/mL were determined by microdilution method.A MIC of the cephalosporins of ≥ 2 μg/mL is suggestive of ESBL production [11].
Isolates with one or more of the following criteria were considered to be potential ESBLs and were listed for confirmation of ESBL production by the combined disks method.The criteria included an inhibition zone to ceftazidime ≤ 22 mm, cefotaxime ≤ 27 mm, aztreonam ≤ 27 mm, extension of the zone of inhibition of any cephalosporins towards amoxicillinclavulanic acid by the disk diffusion method, or MIC of ceftazidime or cefotaxime ≥ 2 µg/mL [10].Isolates resistant to cefoxitin (inhibition zone < 18 mm), in addition to being resistant to one or more of the third-generation cephalosporins and being intermediate or resistant to amoxicillin-clavulanic acid were considered as putative AmpC β-lactamase producers [9,12].

Phenotypic confirmatory tests for ESBL production
The combined discs method was used; Cefotaxime (30 µg) and ceftazidime (30 µg) disks with or without clavulanate (10 µg) were used for confirmatory phenotypic detection of the ESBL-producing strains [11].
Isolates with negative results by the combined disks method were tested using the double synergy differential test (DSDT) for detection of AmpC βlactamases or combined ESBL and AmpC βlactamases [12].

Detection of ESBLs and plasmid AmpC genes
Automated DNA extraction was performed with MagNA Pure-DNA Isolation and Purification Kit I using a MagNA Pure LC 2.0 Instrument (Roche Lifescience, Penzberg, Germany).Multiplex polymerase chain reaction (PCR) amplification was performed as described previously by group-specific primers to bla TEM , bla SHV , bla CTX-M group 1 and CIT group for the AmpC gene [13,14] (Table 1).

Statistical analysis
All statistical calculations were done using computer programs SPSS (Statistical Package for the Social Science) version 15 for Microsoft Windows.Data were statistically described in terms of frequencies and percentages.P values less than 0.05 were considered statistically significant.

Detection of ESBL and plasmid AmpC genes
Among 59 isolates resistant to third-generation cephalosporins, 44 were positive with multiplex PCR to one or more of the tested genes.All 44 ESBL-and AmpC-producing E. coli isolates were susceptible to imipenem, and 100%, 52.3%, 43.1%, 41%, 29.5%, 27.3%, 13.6%, and 9.1% were resistant to ceftazidime, cefotaxime, amoxicillin-clavulanic acid, cefepime, cefoxitin, aztreonam, ciprofloxacin, and amikacin, respectively.The prevalence of genes encoding ESBLs and AmpC among HCWs is illustrated in Table 2, while the distribution of tested genes among the 44 resistant isolates is illustrated in Table 3.

Discussion
The evolution of antibiotic resistance is a continuous threat to human health worldwide [4].E. coli may act as a reservoir of genes coding for antibiotic resistance and may also be responsible for a group of endogenous infections.Large unrecognized reservoirs of HCWs with ESBL-and AmpC-producing E. coli may lead to transmission of antimicrobialresistant strains in hospitals from HCWs to patients, especially if the HCWs do not adhere to infection control measures at their hospitals [7].
In the current study, we determined the prevalence of fecal carriage of ESBLs, AmpC, and combined ESBL and AmpC β-lactamases among HCWs at Fayoum University Hospital and found that the prevalence of fecal carriage of ESBL-producing E. coli among HCWs was 21%.This result partially  [5].Much lower rates were obtained in different regions of the world: 12.3% in Saudi Arabia [15], 11% in Egypt [4], 7.3% in Tunisia [16], and 1.8% in Portugal [3].All these studies investigated the prevalence of ESBL-producing Enterobacteriaceae in community settings rather than in HCWs, which may explain the lower rates.A recent study conducted in multiple rehabilitation centers examined colonization with ESBL-producing Enterobacteriaceae in healthcare personnel by rectal swabs and found that the prevalence was 3.5% among the isolated E. coli [6].These results are still lower than ours.This may reflect non-adherence of HCWs to hand hygiene in our region.It could also be due to the fact that we collected stool samples from HCWs serving different hospital wards, not only rehabilitation wards.It should be noted that the percentage of resistant organisms isolated from inpatient wards is usually higher than that from outpatient areas [17].
In the current study, the prevalence of fecal carriage of AmpC genes among HCWs was 3%.To our knowledge, this is the first study investigating the prevalence of AmpC-producing E. coli isolated from stool samples among HCWs.These results are similar to those obtained from patients in other works; Naseer et al. reported a prevalence of 3.7% of AmpC among outpatients in Norway [18], and Li et al. reported a prevalence of 1.9% [19].These findings are nearly similar to those reported in the Czech Republic and Denmark (rates were1.1% and 2.4%, respectively) [20,21].
In our study, the distribution of genes encoding ESBLs among 44 isolates was: 36 (81.8%) isolates with bla SHV , 7 (15.9%)with bla TEM , and 2 (4.5%) with bla CTX-M .It appears that bla SHV is the predominant ESBL gene among the E. coli-resistant strains in HCWs.This is in agreement with another Egyptian study, which found that the main type of ESBL gene isolated from patients in the ICUs at Assiut University Hospitals was bla SHV [22].
Our results disagree with many studies all over the world that reported bla CTX-M was the predominant gene among ESBL-producing E. coli isolated from fecal samples of healthy subjects [4,23,24], which may be explained by the variability in resistance patterns and encoding genes according to different geographic regions.

Conclusions
In conclusion, an alarming high rate of fecal carriage of ESBL-producing E. coli among HCWs has been detected, with bla SHV being the most predominant gene.These results emphasize the need for continuous surveillance in hospitals for both patients and HCWs to detect resistant strains by implementing strict guidelines for antibiotic therapy and infection control measures to reduce the increasing burden of antibiotic resistance and to prevent its spread.

Table 1 .
Group-specific primers used in the study

Table 2 .
Prevalence of resistant genes encoding ESBLs and AmpC among HCWs

Table 3 .
Distribution of tested genes among 44 resistant isolates