Prevalence and characterization of extended-spectrum beta-lactamases producing Enterobacteriaceae in healthy children and associated risk factors

Background Community acquired infections due to extended-spectrum beta-lactamase-producing Enterobacteriaceae (ESBL-PE) had been increased. The fecal flora of children in the community represents a huge potential reservoir for ESBLs which are located on highly transmissible plasmids. This study examined the prevalence of ESBL-PE fecal carriage, antimicrobial susceptibility pattern, possible risk factors, and characterized the genes encoding these ESBL enzymes in Lebanese children community. Methods A total of 125 rectal swabs were taken from healthy children aged from 1 to 5 years. Detection of ESBLs was carried out using combination-disc method test and multiplex PCR. A questionnaire concerning child’s lifestyle and risk factors for ESBL carriage was illustrated. Results Thirty-one of 125 participants (24.8 %) carried ESBL-PE. Regular consumption of meat, and chicken were significantly associated with high carriage rate of ESBL-PE, while dairy products (milk, yogurt, cheese) association was non-significant. Intimate hygiene habits were found also affecting the carriage rate. Multiple bla genes were the most common, 48.4 % (15/31) of ESBL-PE carried both blaCTX-M and blaTEM, and 22.6 % (7/31) carried blaCTX-M, blaSHV, and blaTEM, 29 % (9) carried blaCTX-M only. Concerning CTX-M-types, CTX-M-9 was the most predominant (24/31) and mostly in combination with CTX-M-15 type. Conclusion High rate of colonization in healthy children with ESBL-PE was observed, regular consumption of dietary products from animal source (meat or chicken) were associated with this colonization in the community in non-hospitalized children. To our best knowledge it is the first study about regular consumption of dairy product as a risk factor for ESBL-PE community carriage, the first data about the carriage rate of ESBL-PE in community children in Lebanon and Middle East, and for the wide dissemination of CTX-M-9 type in this population.

largest and most prevalent group of enzymes [3]. Many ESBL producers are multi-resistant to non-β-lactam antibiotics, including fluoroquinolones and aminoglycosides [4], trimethoprim, tetracyclines, sulfonamides, and chloramphenicol as well as aminoglycosides, and this is often encoded by the same plasmids that determine the ESBL [5]. Consequently, effective antibiotic therapy for treating these infections is limited to a small number of drugs [6] such as carbapenems and thus increasing the chance of resistance to carbapenems among the Enterobacteriaceae. Of major concern is the coexistence of multiple ESBL, carbapenemase genes, and other antibiotic resistance determinants on mobile elements which may lead to the emergence of organisms with resistance to all antibiotic [2].
Rapid dissemination of Escherichia coli and Klebsiella pneumoniae isolates producing ESBLs widely in the community settings, resulted in both community-onset and hospital-associated infections on a global scale [7]. More than 200 types of ESBLs have been described in various species of the Enterobacteriaceae family and other non-enteric organisms, including Pseudomonas aeruginosa and Acinetobacter spp. These organisms produce variants of the TEM, SHV and CTX-M β-lactamases [8]. TEM-and SHV-type β-lactamases, mainly produced by K. pneumoniae, have spread throughout hospital settings, and CTX-M enzymes, mainly produced by E. coli, have become predominant in the community [9]. CTX-M enzymes were discovered in 1989 [10] but they did not become predominant over the other ESBL enzymes until the first decade of the twenty-first century during which an extraordinary spread of these enzymes was observed [11] in both hospital and community settings [12]. The bla CTX-M-15 gene are found mainly in Enterobacteriaceae and were recently named "plasmids of resistance responsible for outbreak" because of their capacity to acquire genes of resistance and to transfer among bacteria [13].
The community may thus represent a reservoir for ESBLs not detected in clinical isolates [14].
Most data on prevalence, risk factors and molecular characterization of ESBL -producing organisms are from studies in adult patients in both hospital and community settings, or from studies confined to hospitalized infants [15], only few data are available concerning the healthy pediatric populations. It is critical to better define the prevalence, risk factor, and molecular characterization of ESBLs-producing organism carried by children to adopt best-practice infection control measures and help in the appropriate choice of empirical antimicrobial coverage for infections in these populations.
The aim of this study is to investigate the prevalence and predisposing factors of intestinal carriage of ESBLs-producing Enterobacteriaceae (ESBL-PE) among Lebanese community children and to determine the molecular characterization of the resistance genes (bla TEM , bla CTX , bla SHV , bla CTX-M-2 , bla CTX-M-9 , and bla CTX-M- 15 ).

Ethical clearance
The study protocols were approved by Institutional Review Board Committee of Beirut Arab University. Written informed consents were obtained from all patients parents (at least one parent of each child) before enrollment.

Sample population and questionnaire
The study was performed between January 2013 and May 2013 in three different pediatric clinics in Lebanon. Only 125 healthy children population aged from 1 to 5 years old coming for vaccination or general checkup were chosen for this study. However, the children who were under antibiotic treatment within 5 days before the enrollment date were excluded from this study. A questionnaire was completed for each participant regarding name, age, gender, medical history (previous antibiotic or antacid intake, previous hospital admission), dietary habits (milk, yogurt, cheese, meat, or chicken consumption), and intimate hygiene habits.

Bacterial isolation and ESBL detection and confirmation
Rectal swabs were taken by a sterile swabs moistened with sterile saline and were immediately plated onto MacConkey agar (Oxoid, UK) plates supplemented with 2 mg/l ceftazidime (used within 5 days of preparation), incubated at 37 °C for 24 h. When a sample showed positive growth on the selective medium, at least three colonies plus each distinct morphotype were selected for subsequent characterization. Bacterial identification was performed using Gram staining, biochemical testing (indole, methyl red, vogus-prauskaur, citrate, and urease), and API 20E system (bioMerieux,Marcy l'Etoile, France).
The isolates were first screened for ESBL production using ceftazidime, cefepime, cefotaxime, cefpodoxime, ceftriaxone, and aztreonam disks (Oxoid Ltd, Basingstone,UK) and then phenotypic confirmatory test was carried out by double disc synergy and the combination-disc method on Mueller-Hinton agar using ceftazidime, ceftazidime-clavulanic acid, cefotaxime, and cefotaxime-clavulanic acid. The organisms were considered to be ESBL producing when a ≥5-mm increase in a zone diameter for either antimicrobial agent tested in combination with clavulanic acid compared with the zone diameter of the agent when tested alone [16]. Then molecular analysis was done on all positive isolates screened.

Characterization of genes encoding ESBLs
Two conventional multiplex PCR was performed, the first one for the detection of TEM, SHV, CTX-M genes and the second for detection of CTX-M-2, CTX-M-9, CTX-M-15 encoding genes. Crude genomic DNA was extracted from the isolates by heat lysis. Briefly, one pure colony was suspended in 40 µl of sterile distilled water, and the cells were lysed by heating up at 95 °C for 5 min.
For the first PCR multiplex, suitable primers (Sigma-Aldrich) [17] each targeting selected region the bla TEM , bla SHV , and bla CTX-M were used (Table 1).
For the second multiplex PCR, suitable primers each targeting selected region the bla CTX-M-2 , bla CTX-M-9 , and bla CTX-M-15 were used [17] (Table 2). Reactions were performed in a DNA thermal cycler under the following conditions: denaturation at 94 °C for 5 min followed by 30 cycles at 94 °C for 15 s, 56 °C for 15 s and 72 °C for 45 s with a final extension of 72 °C for 5 min.
After PCR amplification, 2.5 μl of each reaction was separated by electrophoresis in 1.5 % agarose gel for 30 min at 100 V in 0.5× TBE buffer. DNA was stained with ethidium bromide (1 μg/ml) and the bands were detected using UV transilluminator (Cleaver Scientific Ltd).

Statistical analysis
The data were analyzed using Yates corrected χ 2 test. Chi square test was done to establish statistically difference in proportions for categorical data. Statistical significance was set as P values of <0.05 with confidence interval of 95 %. For multivariate logistic regression analysis, the two independent variables were dairy products and meat. For dairy products data were counted for consumer of any of the dairy products (milk, yogurt, or cheese). For meat variable data were counted for consumer for either animal meat or chicken collectively as meat. ESBL-PE carriage was the dependent variable. Statistical analysis was performed using Minitab program (Minitab 14 statistical software, PA, USA) and The Statistical Package for Social Sciences (SPSS, Version 20) program (IBM, Armonk, NY).

Factors associated with ESBL-PE carriage
Subjects who used antibiotics or antacid in the last 8 weeks, had higher rate of ESBL-PE carriage than their counterparts (Table 3), though no significant differences between these variables, while subjects admitted to the    (Table 3). A univariate and multivariate logistic regression analysis were performed to test for regular consumption (3 times per week or more) of milk, yogurt, cheese, chicken, or meat and ESBL-PE carriage. As seen in Table 3 the univariate analysis for each predictor revealed significant impact (P < 0.05) for regular consumption of milk, yogurt, cheese, meat, or chicken. However the multivariate logistic regression analysis (Table 4) failed to reveal any of these dietary products consumed as a predictor of ESBL-PE carriage. To delineate the food type as dairy products or meat (meat or chicken) as a predictor, the analysis revealed the meat and chicken as significant (P < 0.02) predictor for ESBL-PE carriage while dairy products impact was non-significant (P > 0.2) ( Table 5).
For intimate hygiene habits, participant who wash with water or with soap and water then dry with tissue have lower ESBL-PE carriage rate 20.3 % (15/74) compared

Discussion
ESBL-producing organisms are increasingly common worldwide, they are recognized as important nosocomial pathogens in children, and are often associated with outbreaks [19]. Screening for intestinal carriage is crucial to predict the risk of ESBL infection, as the colon serves as a reservoir for extra-intestinal pathogenic E. coli [20].
This study was the first survey conducted in Lebanon on the intestinal carriage of ESBL-PE in the children community, and very rare studies were reported from  Males appear to have higher colonization rate than females. This observation is difficult to explain and needs to be further explored.
High carriage rate was in those individuals that used antibiotics or antacid in the last 8 weeks, or hospitalized in the last 12 months. Risk factors mentioned in the literature were prolonged hospital stays, living in nursing homes or long-term care facilities, underlying medical conditions, recent surgery, haemodialysis, and also prior use of antibiotics, particularly quinolones and third-generation cephalosporins, but also co-trimoxazole, aminoglycoside, and metronidazole [9].
Regarding antibiotic intake, although the selected children were healthy and not taking antibiotics (5 days before enrollment), it is noticed that 44.8 % of them have been taken antibiotics within last 8 weeks prior to being screened, this rate considered very high and reflects the high exposure of children to antibiotics and might explain the comparably high colonization rate. As known one of the main risk factors for the development of bacterial resistances is the increase of the consumption of several antibiotics [27]. In Lebanon there is no control measure and guidelines for antibiotic use in this age group. In addition, antibiotics are taken without medical prescription particularly in lower socioeconomic areas [28].
To the best of our knowledge, it is the first time to study regular consumption of dairy products (milk, yogurt, and cheese) and ESBL-PE carriage. In the present study, univariate analysis revealed significant association between regular consumption of dairy products, chicken, or meat with high ESBL-PE carriage rate. However analysis using multivariate logistic regression to test for the confounding factors, failed to show significant impact of any of these dietary products tested. This could be explained in part by the consumption of more than one type by most of the participants. Also the number of cases that consumed one type only is few to show the power of the test. This was obvious as gathering milk, yogurt and cheese as dairy products and meat and chicken together as meat showed significant association between meat and ESBL-PE colonization. A conclusion that meat is a predictor for ESBL-PE carriage compared to dairy products was also apparent by the insignificant impact of dairy products on ESBL-PE carriage, although both had higher incidence of ESBL positive isolates compared to control rare or never users. The odd ratio was 3 and 2 for meat and dairy products respectively. These dietary products are from animal source, and as known antibiotics are used for various purposes in agriculture and livestock production, for instance to boost growth or as therapeutic treatment and disease prophylaxis. E. coli and some other Enterobacteriaceae colonize the intestinal tract of both animals and humans, and recently the number of studies describing the prevalence of ESBL-PE in meat and raw milk has increased [29,30]. Many studies had shown significant genetic similarities among ESBL-positive E. coli isolates from chicken meat and human according to mobile resistance elements, virulence genes and genomic back ground [31][32][33]. However other fewer contrasting observations showed that considerable differences in ESBL types between poultry and humans in Europe exist [34], and meat consumption is not related to ESBL transmission suggesting environmental sources such as water for drinking and food preparation [35]. There is no direct study between consuming dairy products and ESBL-PE carriage. A recent study in Tunisia isolated E. coli from cattle milk harbored the bla CTX-M-15 gene on an F2: A-:B-plasmid, a combination frequently found in humans [36]. Although the dissemination of resistance genes in E. coli may occur through multiple routes, our study show a potential threat posed by animal food products as sources for human ESBL-positive isolates [33]. Studies concerning ESBL-PE in dairy products and meat in the Lebanese market should be assessed to determine the impact of food products as reservoirs and disseminators of such strains through the food production chain to human. Washing with water, or soap and water, the intimate area after urination, fecal evacuation, or diaper change might have tendency towards protective effect against ESBL-PE carriage compared to using only dry tissue or wipes, although P = 0.07, this might be further studied and explained.
Apart from a high rate of susceptibility to imipenem and amikacin, most isolates were resistant to other antimicrobials tested. ESBL-producing isolates were found highly resistant to nalidixic acid, tetracycline, gentamycin and to trimethoprim-sulfamethoxazole, and this result confirm carriage of multidrug-resistant ESBL-PE in this selected population of asymptomatic healthy children in Lebanon. This high carriage rate of ESBL-PE and associated resistance to aminoglycosides and trimethoprim-sulfamethoxazole, as well as high frequency of coexistance of fluoroquinolone resistance, increases the risk of infection with multi-resistant bacteria, and thus the need for usage of last resort antibiotics, such as carbapenems and colistin, in the treatment of common infections [6]. Majority of ESBL-PE carry multiple bla genes, where CTX-M-type was the most predominant. CTX-M ESBLs has been increased worldwide among E. coli and Klebsiella spp., this is the first study in the Lebanon that specifically determined the high prevalence of CTX-M-type ESBLs isolated from a healthy asymptomatic children, where CTX-M type beta-lactamases enzyme were produced by all isolates, and the majority of the isolates were found to be co-producers of the TEM gene, hence, TEM type was considered also a frequent disseminated betalactamases type. This was in contrast to a study in Egypt HCW where SHV was the predominant [25].This massive dissemination of CTX-M-type ESBLs in our result and its rapid and global spread almost reported everywhere could be referred as 'CTX-M pandemic' [37,38], as the CTX-M-encoding plasmids are often transmissible by  conjugation in vitro; this property explains the easy dissemination of bla CTX-M -harboring plasmids [39]. This dissemination of the CTX-M-type ESBLs is not restricted to the nosocomial setting but also involves the community. This phenomenon is acting to modify the epidemiology of ESBLs, whereas those enzymes were, previously, mostly restricted to the nosocomial setting [40], this may be correlated with food processing channel especially those from animal origin. Molecular characterization of CTX-M-type ESBL revealed that the majority was CTX-M-9, followed by CTX-M-15 type. The greater number of isolates was co-producer of either 2 or 3 genes (bla CTX-M-9 , bla CTX-M-15 , bla CTX-M-2 ) together. CTX-M-9 was first reported from E. coli obtained from urine specimen from Spain and Brazil in1996 [41,42], and recently one report from India [43]. Where CTX-M-type was predominant in E. coli obtained from Libyan children with diverse groups of CTX-M-1, CTX-M-15, and CTX-M-3 [21], major type was CTX-M-15 and no CTX-M-2 and CTX-M-9-types in E. coli isolate obtained from stool of Tehran children suffering from diarrhea [44], and CTX-M-1, CTX-M-15, and CTX-M-14 were predominant types of ESBL-PE obtained from French children rectal swabs [45]. This is the first study from Lebanon that demonstrates such a high prevalence of CTX-M-9 enzymes among Enterobacteriaceae strains isolated from community children. The bla CTX-M-9 was the predominant ESBL gene in this setting. Although CTX-M-15 are the most frequent CTX-M enzymes isolated in humans, animals, as well as in the environment, worldwide [46], and associated with the global CTX-M pandemic, but in our study CTX-M-9-type ESBL appears the most predominant. This is an indication that new mechanisms of resistance might be emerging in Lebanon.

Conclusion
The carriage of ESBL-PE in young children in the Lebanese community is very high compared to the other countries. A significant correlation was found between this colonization rate and regular consumption of meat or chicken. Alarming multidrug resistance ESBL-PE is of great concern especially in this age group. Furthermore ESBL-PE multiple gene carriage of TEM, SHV, and CTX-M was the predominant, where CTX-M-9 and CTX-M-15 were the mostly spreader ESBL type.