Following a request from the European Commission, the Panel on Biological Hazards (BIOHAZ) was asked to deliver a Scientific Opinion on the public health risks of bacterial strains producing extended-spectrum beta (β)-lactamases (ESBL) and/or AmpC β-lactamases (AmpC) in food and food-producing animals. In particular, the Panel was asked: (i) to propose a definition of the ESBL- and/or AmpC-producing bacterial strains and genes relevant for public health and linked to food-producing animals or food borne transmission; (ii) to review the information on the epidemiology of acquired resistance to broad spectrum cephalosporins including the genes coding for such resistance in food-producing animals and food, ensuring that differentiation was made between transmission of resistant bacterial strains and/or genes to humans by consumption or handling of contaminated food; and transmission of resistant bacterial strains and/or genes to humans through the food animal production environment; (iii) to perform a critical analysis of the methods (phenotypic and genotypic) and the interpretive criteria currently used for detection (isolation and identification) and characterisation of ESBL- and/or AmpC-producing bacterial strains, ESBL- and/or AmpC-encoding genes and associated mobile elements; (iv) to make recommendations for a harmonised monitoring of resistance (phenotypic eand genotypic) caused by ESBL- and/or AmpC in food and food-producing animals in the EU; (v) to the extent possible, to identify risk factors contributing to the occurrence, emergence and spread of ESBL- and/or AmpC-producing bacterial strains in food producing animals and food; and finally, (vi) to identify and rank possible control options, taking into account the expected efficiency in reducing public health risk caused by ESBL and/or AmpC-producing bacterial strains transmitted via the food chain or via food animal production environment, and consider the advantages and disadvantages of different options.
The BIOHAZ panel concluded that ESBLs may be defined as plasmid-encoded enzymes found in the Enterobacteriaceae, frequently in Escherichia coli and Klebsiella pneumoniae, that confer resistance to a variety of ß-lactam antibiotics, including penicillins, 2nd-, 3rd- and 4th-generation cephalosporins and monobactams (eg aztreonam), but usually not the carbapenems or the cephamycins (e.g. cefoxitin). In contrast, AmpC β-lactamases are intrinsic cephalosporinases found on the chromosomal DNA of many Gram-negative bacteria, which confer resistance to penicillins, 2nd- and 3rd-generation cephalosporins including β-lactam/inhibitor combinations, cefamycins (cefoxitin), but usually not to 4th-generation cephalosporins (cefepime, cefquinome) and carbapenems; a growing number of these AmpC enzymes are now plasmid-borne.
The potential contribution of food-producing animals or foods to public health risks by ESBL and/or AmpC-producing bacteria is related to specific plasmid-mediated ESBL and/or AmpC genes encoded by a number of organisms. Although there are a large number of genes which encode ESBL and AmpC enzymes not all are equally prevalent among human and animal bacteria. The predominant ESBL families encountered are CTX-M, TEM, and SHV. The predominant AmpC-family is CMY. The bacterial species most commonly identified with these genes are Escherichia coli and non-typhoidal Salmonella. Among E. coli, the clonal lineages: B2-E. coli O25:H4-ST131, D-E. coli O25a-ST648 and D- E. coli-ST69, -ST393, are being increasingly detected among both humans and animals. Among Salmonella the most common serovars are Typhimurium, Newport, and Heidelberg; ESBL/AmpC transmission is mainly driven by integrons, insertion sequences, transposons and plasmids, some of which are homologous in isolates from both food-production animals and humans.
Cefotaxime is used as the drug of choice for optimum detection of blaESBL and/or blaAmpC genes in Salmonella and E. coli. From the results presented in the Community Summary Report it can be concluded that the prevalence of resistance to cefotaxime in food animals varies by country and animal species. High levels are observed in E. coli and Salmonella from poultry in Spain, Italy, the Netherlands and Poland. From raw meat from poultry only limited cefotaxime resistance prevalence data are available. Belgium and the Netherlands reported high to moderate cefotaxime resistance prevalences in Salmonella and E. coli from poultry meat. In pigs and cattle the prevalences were low. Since 2000, the presence of ESBL- and/or AmpC-producing Salmonella and E. coli in animals and food has been increasingly reported in both Europe and globally. Although these enzymes have been described in bacteria from all major food-producing animals, poultry and poultry products are most frequently reported to carry ESBL and/or AmpC-producing bacteria. The most frequently reported ESBL subtypes in the EU in both Salmonella and E. coli in food-producing animals and foods are CTX-M-1, CTX-M-14, TEM-52 and SHV-12; the predominant plasmidic AmpC variant described globally to occur in Salmonella and E. coli from food-producing animals or foods since the mid-1990s is CMY-2. A wide range of additional CTX-M subtypes (CTX-M-1, -2, -3, -8, -9, -14, -15, -17/18, -20, -32, -53) have been detected in food-producing animals and food in European countries. A range of additional TEM (TEM-20, -52, -106, -126) and SHV (SHV-2, -5, -12) variants have similarly been detected in different European countries. Epidemic plasmids belonging to the incompatibility groups F, A/C, N, HI2, I1 and K groups carrying particular ESBL-encoding genes (blaTEM-52, blaCTX-M-1, -9, -14, -32,) or AmpC-encoding genes (blaCMY-2) have been detected among farm and companion animals, food products and humans. There are few studies that describe clear evidence of direct transmission of ESBL- or AmpC-producing E. coli isolates from food-producing animals or food to humans. Data do exist about common clones of ESBL- and/or AmpC-producing E. coli isolates in humans and food-producing animals and foods, which provide indirect evidence about this transmission. Comparison of E. coli derived from humans and poultry has shown that antibiotic-resistant E. coli isolates from both reservoirs are more frequently genetically related than antibiotic-susceptible isolates. Recent findings indicate transmission of ESBL genes, plasmids and clones from poultry to humans is most likely to occur through the food chain.There is limited evidence for spread of ESBL/AmpC-carrying organisms via direct contact with animals or indirectly via the environment. Nevertheless people working with poultry have a higher risk for intestinal carriage of ESBL/AmpC-producing bacteria.
The preferred method for selective isolation of ESBL- and/or AmpC-producers is using cephalosporin-supplemented agar preceded by selective enrichment in a broth. The preferred method for selective isolation of ESBL- and/or AmpC-producers is chromogenic (e.g. MacConkey agar) with 1 mg/L cefotaxime or ceftriaxone. Using low concentrations will result in optimum sensitivity to detect all relevant beta-lactamase families. Pre-enrichment may be performed in a general broth like Mueller-Hinton, Brain Heart Infusion or Luria-Bertani broth with 1 mg/L cefotaxime or ceftriaxone.
Identification is performed by determination of susceptibility to cefotaxime, ceftazidime and cefoxitin. ESBL producers are resistant to cefotaxime, variably resistant to ceftazidime and susceptible to cefoxitin. Confirmation of ESBLs is performed by testing for synergy with clavulanic acid by combination disks, ESBL-etests or broth micro-dilution including cefotaxime, and ceftazidime as single drugs, and in combination with clavulanic acid. Confirmation of AmpC producers is performed by determination of susceptibility to cefepime. AmpC producers are susceptible to cefepime and resistant to cefotaxime, ceftriaxone and cefoxitin. To identify ESBL and/or AmpC-suspected Enterobacteriaceae by broth micro-dilution susceptibility tests, optimum breakpoints or interpretive criteria need to be used. Although CLSI has recently redefined MIC breakpoints for 3rd- and 4th-generation cephalosporins, the R-breakpoints for ceftazidime, cefoxitin and cefepime are still one to two dilution steps higher than those defined by the European Committee on Antimicrobial Susceptibility Testing (EUCAST). In order to harmonize the interpretation of susceptibility data and for optimum phenotypic detection of ESBL and/or AmpC producers, it is important to use EUCAST clinical breakpoints for interpretation of susceptibility or resistance and EUCAST epidemiological cut-off values (ECOFFs), to determine if an isolate belongs to the wild-type population or not.
All isolates confirmed phenotypically to be either ESBL or AmpC producers may be screened for β-lactamase gene families using micro-array or (multiplex) PCR. The ESBL and/or AmpC subtypes may be identified by dedicated PCRs and sequence analysis of the amplicons. Characterisation of plasmids on which blaESBL and/or blaAmpC-genes are located is essential to study the epidemiology of these genes and plasmids. Since in Enterobacteriaceae several different plasmids are often present in each isolate, a structured approach is needed to identify the characteristics of the plasmid on which the β-lactamase genes are located. If the presence of an ESBL and/or AmpC gene in a bacterial isolate is confirmed, plasmid isolation is performed to determine the number and sizes of plasmids present. Subsequently, by conjugation or electroporation, tranconjugants or transformants are isolated on selective agar plates with only the plasmid that harbours the β-lactamase gene present. The plasmid can be typed using replicon typing and sub-typed by fingerprinting or plasmid MLST. Ultimately whole plasmid sequence analyses may replace the current typing and subtyping techniques. The choice of the molecular typing method to be used for isolates is determined by epidemiological relatedness of the isolates. Next to phenotypic methods such as serotyping and phage typing, PFGE or MLVA can be used to identify clusters of isolates that are related to a certain ‘outbreak’ in a restricted time frame. MLST is generally the method of choice to identify relatedness of isolates of the same species from different backgrounds (eg. animal versus human).
The establishment of risk factors for occurrence of ESBL/AmpC-producing bacteria is particularly complicated by the data unavailability or lack of its accuracy. Few studies designed to assess risk factors for ESBL and/or AmpC occurrence in animals are available. The use of antimicrobials is a risk factor for selection and spread of resistant clones, resistance genes and plasmids. Most ESBL- and AmpC-producing strains carry additional resistances such as to sulphonamides and other commonly-used veterinary drugs. Therefore, generic antimicrobial use is a risk factor for ESBL/AmpC and it is not restricted specifically to the use of cephalosporins. Currently there are no pan-European data available on the use of antimicrobials. The European Surveillance of Veterinary Antimicrobial Consumption (ESVAC), coordinated by the European Medicines Agency (EMA), is collecting information. An additional risk factor is extensive trade of animals in EU member states (MS), with few countries leading the production and the export, and with a small number of companies producing pure line breeding animals. How widespread ESBL-carrying bacteria are in food-producing animals in the breeding/rearing/fattening sectors is generally unknown, although a few reports suggest that ESBL/AmpC are not uncommon in the top of some production pyramids (breeding). ESBL- and/or AmpC-producing E. coli are introduced in poultry production chain through day-old grandparent chickens. Moreover, some data indicate that the occurrence of these organisms in the different levels of the poultry production chain is the result of vertical transmission, local recirculation and selection.
There are no data on the comparative efficiency of individual control options presented in this document in reducing public health risks caused by ESBL and/or AmpC-producing bacteria related to food-producing animals. Prioritisation is complex, and the effectiveness of measures discussed in this Opinion is based on the best available evidence and expert opinion. As such it is considered that a highly effective control option to reduce selection of ESBL/AmpC-producing bacteria at an EU level, would be to stop all uses of cephalosporins/systemically active 3rd/4th generation cephalosporins, or to restrict their use (use only allowed under specific circumstances). Measures intended to minimize off label use should focus on increased compliance with existing legislation. As co-resistance is an important issue, it is also of high priority to decrease the total antimicrobial use in animal production in the EU. Also of importance (more so after the ESBL/AmpC-producing microorganisms have emerged) are the measures to control dissemination, for example by implementing increased farm biosecurity and controls on animal trade (of ESBL/AmpC-carriers), and by improving hygiene throughout the food chain, and implementing other general post-harvest controls for food-borne pathogens. Because most evidence is available for high prevalence of ESBL/AmpC-producing bacteria in the poultry production pyramid, and their consequent involvement in public health, it is of high priority to reduce selection pressure impossed by the use of antimicrobials, to prevent vertical transmission from the top of the poultry production pyramid, and to prevent local recirculation within subsequent flocks.
Recommendations for the harmonised monitoring of resistance caused ESBL- and/or AmpC-producing bacteria have been provided.