Key points are not available for this paper at this time.
Over the last decade, antibiotic resistance and particularly multiple resistance (to four or more drugs) has increased dramatically in salmonellae isolated from infected patients in Western Europe and North America. An important factor in this increase has been the epidemic spread, since 1990, of multiresistant Salmonella typhimurium definitive phage type (DT) 104 with resistance to ampicillin, chloramphenicol, streptomycin, sulphonamides and tetracyclines (ACSSuT resistance type).1 Strains of this phage type/ resistance type combination have become widely distributed in cattle in the UK since the early 1990s,2 and have subsequently been transmitted to humans through the food chain.3 In England and Wales, isolations of multiresistant DT 104 from humans increased from about 200 in 1990 to >4000 in 1996.4 However, over the last two years there has been a significant decline in isolations, with numbers dropping by 48% in 1998, to 2090.5 Antibiotic-susceptible isolates of S. typhimurium DT 104 have been obtained from cases of of human infection in England and Wales since the early 1960s. However, multiresistant DT 104 of resistance type ACSSuT was not identified until the early 1980s, with the first isolations from gulls and exotic birds. With the exception of a small outbreak in Scotland in the mid-1980s, there were no isolations of multiresistant DT 104 from humans until 1989, by which time the strain had begun to be isolated from cattle. Over the next 5 years, multiresistant DT 104 became epidemic in cattle throughout the UK.6 However, in contrast to previous epidemic multiresistant phage types of S. typhimurium, such as DTs 29, 204, 193 and 204c,7–9 which were for the most part confined to cattle, multiresistant DT 104 has also become common in poultry (particularly turkeys), pigs and sheep.2 Human infection with multiresistant DT 104 has been associated with the consumption of chicken, beef, pork sausages and meat paste3 and, to a lesser extent, with occupational contact with infected cattle.10 Although some variants have been identified, the majority of multiresistant DT 104 have a distinctive XbaI-generated macrorestriction fingerprint when studied by pulsed-field gel electrophoresis.11,12 Over the last 4 years, this clone has caused outbreaks of infection in food animals and humans in numerous European countries including the Irish Republic,13 Denmark,14,15 Germany,12,16 Austria,12 France,17 the Czech Republic,18 Italy (A. Caratolli, personal communication) and Sweden (R. Wollin, personal communication). Infections in humans have also been recognized in Trinidad, South Africa, The Netherlands, Northern Ireland, the United Arab Emirates and the Philippines,11,12 and in Israel.19 In 1996 infections with multiresistant DT 104 were recognized in cattle and humans in North America, both in Canada20 and the USA.21–23 In the USA a considerable number of outbreaks have been associated with the consumption of cheese made from raw milk.24–26 Of particular concern has been the resistance of the organism to a wide range of therapeutic antimicrobial agents. Furthermore, in some countries there have been reports of an apparent predilection of the organism to cause serious disease.3,14,15,21,25,26 However, a recent UK study has demonstrated that multiresistant DT 104 appears no more invasive than other common serotypes and phage types in terms of isolations from blood culture in comparison with faecal isolations.4 Genetic analysis has demonstrated that in multiresistant DT 104 of resistance type ACSSuT, the complete spectrum of antibiotic resistance is chromosomally encoded.1 Investigation of a selection of multiresistant strains by PCR demonstrated integron hot spots in two copies as two discrete bands of approximately 1.0 and 1.2 kb. Direct nucleotide sequencing of the individual amplicons indicated that the 1.0 kb gene product was the enzyme ANT (3′′)-1a, responsible for resistance to streptomycin (and also to spectinomycin); the 1.2 kb amplicon contained the gene blaP1, encoding the β-lactamase CARB-2 (PSE-1).11,27 It is particularly noteworthy that all isolates of multiresistant DT 104 of the ACSSuT phenotype have contained the same gene cassettes irrespective of source (food animal or human) or country of origin.11 An indistinguishable β-lactamase has also been identified in France in strains designated as ‘12 atypic’, which probably corresponds to multiresistant DT 104.17 Subsequent studies have demonstrated that the genes for chloramphenicol and tetracycline resistance are grouped between the ampicillin and streptomycin/spectinomycin integrons described above.28 Antibiotic resistance in multiresistant DT 104 of resistance type ACSSuT therefore comprises a sequence of approximately 1.4 kb, containing the ampicillin and streptomycin/ spectinomycin integrons, and intervening plasmid-derived genes coding for resistance to chloramphenicol and tetracyclines. Since 1992 a disturbing feature of infections with multiresistant DT 104 has been the appearance of additional resistance to trimethoprim and ciprofloxacin.29 In 1998, 13% of multiresistant DT 104 in England and Wales were additionally resistant to trimethoprim (ACSSuTTm resistance type). Sixteen per cent showed decreased susceptibility to ciprofloxacin, with an MIC of ciprofloxacin ranging from 0.5 to 1.0 mg/L (ACSSuTCp resistance type), and 2% were additionally resistant to both trimethoprim and ciprofloxacin (ACSSuTTmCp resistance type).5 In isolates of resistance type ACSSuTTm, resistance to trimethoprim was encoded by a plasmid of approximately 4.6 MDa (6.9 kb), which also coded for resistance to sulphonamides.30 It has been suggested that the appearance of resistance to trimethoprim may have resulted from the use of trimethoprim-containing compounds in cattle in attempts to combat infection with DT 104 of resistance type ACSSuT.29,31 In isolates with decreased susceptibility to ciprofloxacin, this property is chromosomally encoded. Amplification and sequencing of the 120 bp quinolone resistance determining region (QRDR) in a panel of 15 isolates of multiresistant DT 104 of resistance type ACSSuTCp has identified two discrete base substitutions at codon Asp87 and further point mutations at codons Ser83 and Ala119 (just outside the QRDR), all giving rise to decreased susceptibility to ciprofloxacin.11 The most common mutation in Asp87 involved change from GAC (Asp) to AAC (Asn). This mutation was identified in 11 of the 15 strains studied, including four from food animals (two poultry, one pig, one cattle). An identical mutation giving rise to decreased susceptibility to ciprofloxacin has recently been identified in a strain of multiresistant DT 104 responsible for an outbreak in Denmark in the summer of 1998. This outbreak was associated with pork of Danish origin.15 The second mutation in codon 87 was from GAC (Asp) to GGC (Gly); this mutation was identified only in strains isolated from humans. The mutation at codon 83 was from TCC (Ser) to TTC (Phe) and that at codon 119 was from GCA (Ala) to GTA (Val). The former mutation was observed in a strain of human origin and the latter in a strain of porcine origin. The significance of these results is that strains of multiresistant DT 104 with decreased susceptibility to ciprofloxacin are not clonal; it is possible that such strains may have arisen independently, either at different times or in different hosts. The emergence and spread in the UK of multiresistant DT 104 with decreased susceptibility to ciprofloxacin followed the licensing for veterinary use of the related fluoroquinolone antibiotic, enrofloxacin, in November 1993.4,31 This antimicrobial agent has subsequently been used extensively in both cattle and poultry for treatment and prophylaxis.32 A consequence of this has been the rapid development of resistance to quinolone antibiotics in strains of multiresistant DT 104 from food-producing animals.33 Such resistance has been particularly apparent in strains of multiresistant DT 104 from turkeys but also in strains from chickens and cattle.33 For humans, the clinical significance of decreased susceptibility to ciprofloxacin is controversial.34 However, in the Danish outbreak in 1998, four of 11 hospitalized patients did not respond to treatment with ciprofloxacin and two died.15 This clearly demonstrates the potential of multiresistant DT 104 for causing serious disease and the clinical consequences of decreased susceptibility to ciprofloxacin in this epidemic multiresistant strain. Drug resistance in zoonotically transmitted salmonellae is an undesirable but almost inevitable consequence of the use of antimicrobial agents in food animals. Such use is quite legitimate. However, it is regrettable that recommendations propounded in the UK in 1992 by the Expert Group on Animal Feedingstuffs (the Lamming Committee), that any new antibiotics with cross-resistance to those used in human medicine should not be used for prophylaxis in animal husbandry,35 were not accepted. In this respect the consequences of the recent decision of the USA Food and Drug Administration (FDA) to approve enrofloxacin for use in pigs and cattle will require close monitoring. To combat the development of resistance in zoonotic pathogens to such important drugs as the fluoroquinolones, it is hoped that such antimicrobial agents will only be used in food animals when absolutely necessary and that codes of practice for their usage will be strictly followed. In the UK recommendations targeted at the development of a coherent strategy aimed at reducing the veterinary use of antibiotics have recently been published by the Advisory Committee on the Microbiological Safety of Food (ACMSF), in their recent report on microbial antibiotic resistance in relation to food safety.36 It is hoped that the ACMSF recommendations will now be adopted and that a real and sustained reduction in the incidence of resistance in such pathogens as S. typhimurium DT 104 will soon follow.
E. J. Threlfall (Sat,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: