Avoparcin and virginiamycin as animal growth promoters: a plea for science in decision-making

Transcription

1 IN SEARCH OF OPINION Avoparcin and virginiamycin as animal growth promoters: a plea for science in decision-making J. Acar 1, M. Casewell 2, J. Freeman 3, C. Friis 4 and H. Goossens 5 1 University of Pierre et Marie Curie, Paris, France; 2 King's College School of Medicine, London, UK; 3 Harvard School of Public Health, Boston, MA, USA; 4 RoyalVeterinary and Agricultural University, Copenhagen, Denmark; 5 University of Antwerp, Antwerp, Belgium INTRODUCTION In the past few years there have been renewed calls to phase out or even ban antibiotics given to animals as growth promoters. For example, these were made at a WHO conference in Berlin in 1997, at a European meeting in Copenhagen in September 1998, and in recent reports from the House of Lords and from Directorate General XXIV of the European Commission. However, an interdisciplinary conference was held at Georgetown University in spring 1999 and included physicians, congressional sta, and regulatory agencies, including the Food and Drug Administration and Centers for Disease Control and Prevention (CDC). It was agreed that a rigorous risk assessment should be undertaken before conclusions were reached on the future of feed additives in livestock production, and, indeed, it has recently been announced by the FDA that a quantitative risk assessment will be carried out of the impact of growth promoters on enterococcal resistance. Used as growth promoters from the 1950s, various antibiotics were found to increase growth in young animals by up to 8% [1]. The possible actions in the gut include altering the ora, increasing nutrient absorption, and diminishing toxin production. Initially, a wide range of antibiotics was used, but because resistant organisms emerged, various countries introduced restrictions. Thus, in 1969, the Swann Report recommended that therapeutic antibiotics should not be used as growth promoters. Later, the European Union (EU) banned certain antibiotics as growth promoters, but the USA did not introduce similar restrictions. In April 1997, avoparcin was banned in the EU, as were virginiamycin, bacitracin, spiramycin, and tylosin in July The EU directive (70/524) stipulates that not only must additives favorably a ect livestock production, but that they should not endanger human health. Corresponding author and reprint requests: H. Goossens, Department of Microbiology, Universitair Ziekenhuis Antwerpen, Wilrijkstraat 10, B-2650 Edegem, Belgium Tel: Fax: Herman.Goossens@uza.uia.ac.be Clearly, any decision to ban or phase out should be based on scienti c studies. The EU avoparcin ban was implemented despite scienti c advice from the EU's own Scienti c Committee on Animal Nutrition (SCAN) that there were insu cient data to support it. Moreover, SCAN reviewed the evidence for the unilateral Danish ban on virginiamycin in 1998, also concluding that this was insu cient. A major problem is the non-rigorous literature: so many of the data are recent and not peer-reviewed that drawing rm conclusions is di cult. In particular, there are few wellplanned and conducted studies, as pointed out by Phillips [2]. Nevertheless, we review here publications and recent communications for two key antibioticsöavoparcin and virginiamycinöbut not for the others, given the paucity of the data. AVOPARCIN Avoparcin is a glycopeptide structurally related to vancomycin and teicoplanin, and was widely used in Europe as a growth promoter from the early 1970s until banned. Vancomycinresistant enterococci (VRE) in Europe were rst reported in the mid-1980s [3,4], and are now an important cause of hospital-acquired infection, almost exclusively in the USA [5]. Their acquired resistance mechanisms to glycopeptides are complex, and three phenotypes (vana, vanb and vand)canbe distinguished [6]. The vana resistance phenotype consists of high-level vancomycin and teicoplanin resistance. The genes are often found on a transposon that in vitro is easily transferred to other enterococci [6]. In the laboratory, the vana gene cluster has also been successfully transferred from Enterococcus faecalis to Staphylococcus aureus [7].The vanb resistance phenotype consists of variable levels of vancomycin resistance and susceptibility to teicoplanin. The genes encoding the vanb resistance phenotype are more commonly chromosomal but can also be transferred. VanD is very rare. The van gene clusters confer resistance to glycopeptides through similar mechanisms. In the USA, avoparcin was never used in animals, whereas in Europe it was widely used until banned [8]. In the USA, only vancomycin is used for human disease, whereas in Europe, both vancomycin and teicoplanin are used. So, despite = 2000 Copyright by the European Society of Clinical Microbiology and Infectious Diseases

2 ãæð Clinical Microbiology and Infection, Volume 6 Number 9, September 2000 the very low incidence of VRE infections in Europe [9], was there enough evidence to justify the ban? To answer this, it is useful to compare the di erences in the occurrence and spread of VRE in both continents, in humans and animals [10].The prevalence of VRE in isolates in US hospitals, both inside and outside intensive care units, has been rapidly increasing [5]. About 70% of isolates exhibit the vana phenotype and about 25% the vanb phenotype [11,12]. Strategies to eradicate VRE from US hospitals have generally failed. Moreover, most VRE have high-level resistance to ampicillin and the aminoglycosides, leaving few options for treatment [12]. As opposed to hospitalized patients,vre have not been found in community-based volunteers in the USA [13,14]. Likewise, VRE have not been found in American food-producing animals and pet animals [13,15,16]. These studies suggest that VRE are not present in the gut of humans and animals in the USA. However, only two studies have been performed in humans [13,14] and three studies in animals [13,15,16]; the microbiological methods were inadequate and only a few samples were studied. Clearly, better-designed studies should be performed in the USA to con rm the absence of VRE in the community and animals. In particular, a broth enrichment step in the culture of feces almost doubles the rate of detection of VRE carriage [17]. In Europe, VRE have spread only in high-risk units such as intensive care, transplant, hematology and renal units. The outbreaks were mostly caused by clonal dissemination of single strains of vana-containing E. faecium, although several strains have been involved in hospital units in the UK. In contrast to the USA, the outbreaks have been successfully contained in Europe through infection control and/or reduction of antibiotic selective pressure.thus,vre infection is not persistent in European hospitals. On the other hand,vre can be isolated from the feces of as many as 12% of healthy volunteers [18]. Several studies have shown that VRE can be found in farm and pet animals in Europe [18,19]. These VRE have always been of the vana phenotype. In contrast to the VRE isolated in US hospitals, those from humans and animals in Europe, e.g. in Belgium [20], remain generally susceptible to ampicillin and to the aminoglycosides. In addition, the susceptibility pro les of the vana phenotype of E. faecium from food animals are comparable with those from community fecal samples. On the other hand, where the vanb phenotype occurs in human feces [21], this cannot be derived from animals, as it is not found in the latter. In Australia it appears that vanb is the predominant phenotype found in humanvre cases, which contrasts with the situation in Europe, despite the fact that avoparcin has been widely used on both continents [22]. In European hospitals with a VRE problem, multiply resistant strains are reported that are similar to those in the USA, rather than those in European animals, with resistance to ampicillin, to high-level gentamicin, and to virtually all antimicrobials [23,24]. VRE may spread clonally from one host to another, or there may be horizontal transfer of vancomycin resistance genes between E. faecium of human and animal origins, and/or from a common reservoir of glycopeptide resistance. Thus the vana resistance genes may have been introduced into the community via the food chain [25]. Although the experimental ingestion by a single subject of 10 7 CFU of chicken-derived VRE resulted in the excretion of VRE for about 3 weeks [26], the interpretation of these ndings has been questioned with regard to the microbiological detection and identi cation methodology, and the lack of con rmation that the ingested and excreted strains were identical [27]. In addition, it has proved remarkably di cult to nd indistinguishable clone pairs of vana phenotypes of E. faecium in fecal samples from humans and animals, although this has been shown for a turkey farmer and his ock [28], and in two instances for epidemiologically unrelated pig and human fecal isolates by Descheemaeker et al [20]. Any generalized deduction from these observations about the unidirectional spread of VRE from animals to humans would be highly speculative. Such observations may simply indicate the wide distribution of VRE clones or the existence of a common reservoir [20]. An alternative suggestion is that the glycopeptide-resistant genetic elements spread horizontally and contribute to vancomycin resistance in humans [29,30]. The transposon originally identi ed as conferring resistance, Tn1546, displays striking diversity due to point mutations, insertion of IS elements, and deletions in the intergenic regions. These variations provide potential markers to type and trace the spread of transposonmediated glycopeptide resistance [30,31]. However, the stability of the elements must be established before they can be used as epidemiologic markers and the true signi cance of their heterogeneity can be resolved [30]. Some have found that elements of vana are shared between organisms of human and animal origins [20,29^31] and others have not [32]. For example, in a collection of 132 vancomycin-resistant E. faecium strains from Belgium, eight transposon RFLP types were found [20].The two predominant types, one indistinguishable fromtn1546 (`type 1') and a more common close variant (`type 2'), were found invre from both animal and human feces. A statistically signi cant association was found between this type 2 and pig VRE isolates. These ndings are in agreement with results reported by Jensen et al [29],Woodford et al [30], and Willems et al [31]. Thus, VRE isolates from human and farm animals may share a common vancomycin resistance gene pool. These results would suggest that genetic diversity in Tn1546 can be successfully exploited for tracing the routes of transmission of vancomycin resistance genes. Further studies are therefore needed, and, for example,vre from di erent origins should be exchanged among research groups in

3 Acar et al Avoparcin and virginiamycin as animal growth promoters ãæñ several countries to compare glycopeptide-resistant clones and transposon types. In The Netherlands, the use of avoparcin in turkey ocks has been correlated with a high frequency of VRE compared with ocks not fed avoparcin, although details are scant [28]. More evidence of the role of avoparcin in the emergence of VRE in food animals was provided recently. Bager et al [33] showed a statistically signi cant decline of VRE isolated from broilers after the ban of avoparcin in 1995 in Denmark. Kruse et al [34] also demonstrated a strong and statistically signi cant association between the use of avoparcin in Norwegian poultry production and the occurrence of VRE. Pantosti et al [35] demonstrated, 18 months after the banning of avoparcin, a decline in the percentage of poultry meat samples containing vana gene-positive VRE from 14.6% to 8%. Thus, avoparcin consumption data correlate well with the prevalence of vana VRE in animals. Several US studies have correlated the acquisition of VRE by humans with the intensity and duration of antimicrobial therapy. Clonal dissemination has been associated with oral as well as intravenous vancomycin [36]. Consumption data in the USA compared with those of major European countries from 1984 to 1996 show an acceleration of vancomycin use in the early 1980s, and further acceleration when oral formulations became available in the mid-1980s. During 1996, for example, kg of vancomycin was used in the USA, compared with 60 kg for the whole of The Netherlands [8,37]. Its intravenous use may have increased because of the emergence of multiresistant coagulase-negative or - positive staphylococci, enterococci, and Streptococcus pneumoniae. The increase in oral use in the USA is possibly a result of its widespread prescription for antibiotic-associated colitis. The vancomycin consumption data correlate well with the greater prevalence of VRE in US hospitals compared with Europe. Moreover, its spread in the former in the 1990s may have been preceded by increased resistance of E. faecium to ampicillin and aminoglycosides. Indeed, high-level resistance to ampicillin and gentamicin among E. faecium isolated in one institution signi cantly increased over the 22 years from 1968 [38]. Given the mobility of the vana and vanb resistance genes, transfer of these genes to ampicillin- and gentamicin-resistant E. faecium may have occurred subsequently. Multiresistant vancomycin-resistant E. faecium strains may have been further selected by the use of b-lactam antibiotics and aminoglycosides. E. faecium is less susceptible to b-lactam antibiotics than E. faecalis because of low a nity or the overproduction of penicillin-binding proteins [39].This might partly explain the greater spread of E. faecium compared to other enterococci.the vana gene has been linked with gentamicin resistance [40] and the vanb gene with ampicillin resistance in enterococci [41]. Thus, the emergence of ampicillin and aminoglycoside resistance may be a prerequisite for the emergence and nosocomial spread of VRE. VIRGINIAMYCIN Used extensively as a growth promoter for 30 years, virginiamycin consists of a synergistic mixture of type A and type B streptogramins. In humans, oral pristinamycin, also a mixture of streptogramin A and B, has been used for 30 years for staphylococcal infections, particularly in France, where topical virginiamycin was also used. Recently, dalfopristin/quinupristin (Synercid), a mixture of type A and B semisynthetic streptogramins, has been developed for human intravenous use [42]. Synercid has promising activity against methicillinresistant Staphylococcus aureus (MRSA), and against VRE, excluding E. faecalis [43]. The recent controversy has arisen because animal-derived virginiamycin-resistant organisms (or their genetic determinants) are cross-resistant with Synercid and might cause human infections. There is anecdotal evidence of clinical infection with streptogramin-resistant enterococci in humans, but the emergence of resistance was associated with therapeutic use of Synercid. Resistance to streptogramin A, but not streptogramin B, confers resistance to the combination. The mechanism of streptogramin resistance in one strain of human E. faecium is inactivation of the streptogramin A component by an acetyltransferase encoded by the sata gene [44] and the satg gene [45]. In human staphylococcal isolates, type A streptogramins are resisted by acetyltransferase enzymes (encoded by vat, vatb, vatc) or active e ux (encoded by vga, vgab). All these mechanisms are reported to be plasmid-mediated. Little is known about the genetic mechanisms in animal isolates. After 30 years' use of virginiamycin in animals, streptogramin-resistant staphylococcal or streptogramin-resistant enterococcal infections in humans are still rare, as con rmed by Jones et al [46]. Of clinical isolates of Gram-positive cocci from 200 centers in the USA and Canada, 99.8% of 1011 E. faecium isolates were sensitive to Synercid, as were 99.8% of 519 VRE isolates. More than 99% of the isolates of Staphylococcus aureus were sensitive to Synercid. Even in France, with the additional selective pressure of pristinamycin, there is no evidence of increasing streptogramin resistance. Of 895 isolates of MRSA from 95 non-university hospitals, 98.5% were sensitive to both pristinamycin and Synercid. In 1997, van den Bogaard found pristinamycin-resistant enterococci in the feces of Dutch pigs [47]. However, these were not speciated, which is important, as E. faecalis is intrinsically resistant to streptogramins. The Danish monitoring program also reported virginiamycin resistance in about half of the isolates of E. faecium from pigs and broilers [48,49]. Although a control group was not included, Welton et al showed the emergence of streptogramin-resistant E. faecium from cloacal cultures of turkeys fed virginiamycin [16].

4 ãðò Clinical Microbiology and Infection, Volume 6 Number 9, September 2000 From The Netherlands, a letter reported enterococci (again, non-speciated) resistant to pristinamycin in 30% of healthy suburban residents [47]. Nevertheless, confusingly, the 1997 Danish Danmap Report found [50] only 2% of healthy humans with stool samples containing streptogramin-resistant E. faecium. Hence, rigorous studies are needed to determine the prevalence of streptogramin-resistant enterococci and to determine whether these organisms re ect transitory excretion or colonization or represent a threat to human health. DISCUSSION Understandable anxieties about antibiotic resistance, with its overtones of human health impairment, can engender evangelical calls for premature action. Nevertheless, the importance of therapeutic use in humans to the development of such resistance should not be underestimated. Our review of the data provides no rigorous evidence that the two growth promoters avoparcin and virginiamycin used in animals create antibiotic-resistant bacterial strains in humans that cause clinical infections, and there are insu cient data for the other antibiotics used in this way. However, for avoparcin there are recent studies suggesting that vana-positive VRE in feces of healthy humans may be derived from animals. Fortunately, these VRE have not been associated with an emergence of VRE infections in humans in Europe. Direct evidence for such a link will be di cult to demonstrate, given the molecular complexity of the problem. Such a sequence might comprise: presence of vana genes in enterococci in animals; selection of thesevre in animals following the use of avoparcin or other antibiotics and horizontal gene transfer; ingestion of food contaminated with vana-positive VRE by humans; selection of thesevre in humans following the use of glycopeptides or other antibiotics and horizontal gene transfer; clinical infection with vana-positivevre in humans. Our understanding is supported by a recent comprehensive review of the literature by the Heidelberg Appeal Nederland (HAN) Foundationöan independent alliance of almost 200 senior academicsöwhich also concluded [51] that `documented in-vivo cases showing spread of antimicrobial resistant Gram-positive bacteria from livestock to humans are in essence non-existent'. CONCLUSIONS We need to clarify the issue of gene transfer and bacterial population dynamics. Speci cally, the following questions need to be answered: To what extent do growth promoters select for resistant organisms in animals? What is the prevalence of resistant enterococci in human populations? If they are in the gut transitorily, can they transfer resistance to other organisms? If so, to what extent? Do animal enterococci colonize humans? If so, do they cause disease? Answers to these questions need well-designed studies and would take years of research, but would enable a proper risk analysis. Banning antibiotics as growth promoters in animals will not solve the problem of antibiotic resistance in hospitals. ACKNOWLEDGMENT We thank Dr Stephen Lock CBE for his invaluable help in the preparation of this manuscript. REFERENCES 1. Friis C. Consumption of antimicrobials by food animals and the resistance of pathogens in food animals. J Hosp Infect 1999; in press. 2. Phillips I. Assessing the evidence that antibiotic growth promoters in uence human infections. J Hosp Infect1999; 43: 173^8. 3. Leclercq R, Derlot E, Duval J, Courvalin P. Plasmid-mediated resistance to vancomycin and teicoplanin in Enterococcus faecium. N EnglJ Med1988; 319: 157^ Uttley AHC, Collins CH, Naidoo J, George RC. Vancomycinresistant enterococci. Lancet1988; 1: 57^8. 5. Anonymous. Nosocomial enterococci resistant to vancomycin in the United States, 1989^1993. MMWR1993; 42: 597^9. 6. Leclercq R, Courvalin P. Resistance to glycopeptides in enterococci. Clin Infect Dis1997; 24: 545^ Noble WC,Virani Z, Cree RGA. Co-transfer of vancomycin and other resistance genes from Enterococcus faecalis NCTC12201to Staphylococcus aureus. FEMS Microbiol Lett1992; 93: 195^8. 8. Wegener HC. Historical yearly usage of glycopeptides for animals and humans: the American^European paradox revisited. Antimicrob Agents Chemother1998; 42: Vandamme P,Vercauteren E, Lammens C et al. A survey of enterococcal susceptibility patterns in Belgium. J Clin Microbiol 1996; 34: 2572^ Goossens H. Spread of vancomycin-resistant enterococci: di erences between the United States and Europe. Infect Control Hosp Epidemiol1998; 19: 546^ Clark NC, Cooksey RC, Hill BC, Swenson JM, Tenover FC. Characterization of glycopeptide-resistant enterococci from US hospitals. Antimicrob Agents Chemother1993; 37: 2311^ Eliopoulos GM,Wennersten CB, Gold HS et al. Characterization of vancomycin-resistant Enterococcus faecium isolates from the United States and their susceptibility in vitro to dalfopristin^quinupristin. Antimicrob Agents Chemother1998; 42: 1088^ Coque TM, Tomayko JF, Ricke SC, Okhyusen PC, Murray BE. Vancomycin-resistant enterococci from nosocomial, community and animal sources in the United States. Antimicrob Agents Chermother1996; 40: 2605^ Silverman J,Thal LA, Perri MB, Bostic G, Zervos MJ. Epidemiologic evaluation of antimicrobial resistance in communityacquired enterococci. J Clin Microbiol1998; 36: 830^ Thal LA, Chow JW, Mahayni R et al. Characterization of antimicrobial resistance in enterococci of animal origin. Antimicrob Agents Chemother1995; 39: 2112^15.

5 Acar et al Avoparcin and virginiamycin as animal growth promoters ãð" 16. Welton LA, Thal LA, Perri MB et al. Antimicrobial resistance in enterococci isolated from turkey ocks fed virginiamycin. Antimicrob Agents Chemother1998; 42: 705^ Ieven M, Vercauteren E, van Descheemaker P, Laer F, Goossens H. Comparison of direct plating and broth enrichment culture for the detection of intestinal colonization by glycopeptide-resistant enterococci among hospitalized patients. J Clin Microb 1999; 37: 1436^ Bates J. Epidemiology of vancomycin-resistant enterococci in the community and the relevance of farm animals to human infection. J Hosp Infect1997; 37: 89^ Devriese LA, Ieven M, Goossens H et al. Presence of vancomycin-resistant enterococci in farm and pet animals. Antimicrob Agents Chemother1996; 40: 2285^ Descheemaeker PRM, Chapelle S, Devriese LA, Butaye P, Vandamme P, Goossens H. Comparison of glycopeptide resistant Enterococcus faecium isolates and glycopeptide resistance genes of human and animal origins. Antimicrob Agents Chemother 1999; 43(8): 2032^ Wendt C, Krause C, Xander LU, Lo er D, Floss H. Prevalence of colonization with vancomycin-resistant enterococci in various population groups in Germany. J Hosp Infect1999; 42: 193^ Commonwealth Department of Health and Aged Care, Commonwealth Department of Agriculture, Fisheries and Forestry. The use of antibiotics in food-producing animals: antibiotic resistant bacteria in animals and humans. Report of thejoint ExpertTechnical Advisory Committee on antibiotic resistance (JETACAR). Commonwealth Department of Health and Aged Care, Commonwealth Department of Agriculture, Fisheries and Forestry, Wade JJ, Rolando N, Williams R, Casewell MW. Serious infections caused by multiply-resistant Enterococcus faecium. Microb Drug Res1995; 1(3): 241^ Wade JJ. The emergence of Enterococcus faecium resistant to glycopeptides and other standard agentsöa preliminary report. J Hosp Infect1995; 30: 483^ Chadwick PR, Woodford N, Kaczmarski EB, Gray S, Barrell RA, Oppenheim BA. Glycopeptide-resistant enterococci isolated from uncooked meat. JAntimicrob Chemother1996; 38: 908^ Berchieri Jr A. Intestinal colonization of a human subject by vancomycin-resistant Enterococcus faecium. Clin Microb Infect 1999; 5: 97^ Goossens H. Vancomycin-resistant Enterococcus faecium derived from animals colonizing the human gut: the missing evidence? Clin Microbiol Infect1999; 5: 64^ Van den Bogaard AE, Jensen LB, Stobberingh EF. Vancomycinresistant enterococci in turkeys and farmers. N Engl J Med 1997; 337: 1558^ Jensen LB, Ahrens P, Dons L, Jones RN, Hammerum AM, Aarestrup FM. Molecular analysis of Tn1546 in Enterococcus faecium isolated from animals and humans. J Clin Microbiol 1998; 36: 437^ Woodford N, Adebiyl A-MA, Palepou M-FI, Cookson BD. Diversity of vana glycopeptide resistance elements in enterococci from human and nonhuman sources. Antimicrob Agents Chemother 1998; 42: 502^ Willems RJL, Top J, van den Braak N et al. Molecular diversity and evolutionary relationships of Tn1546-like elements in enterococci from humans and animals. Antimicrob Agents Chemother 1999; 43: 483^ Van den Braak N, Belkum A, Keulen M,Vliegenthart J,Verbrugh HA, Endtz HP. Molecular characterization of vancomycinresistant enterococci from hospitalized patients and poultry products inthe Netherlands. J Clin Microbiol1998; 36: 1927^ Bager F, Aarestrup FM, Madsen M, Wegener F. Glycopeptide resistance in Enterococcus faecium from broilers and pigs following discontinued use of avoparcin. Microb Drug Res1999; 5: 53^ Kruse H, Johansen BK, Rorvik LM, Schaller G. The use of avoparcin as a growth promotor and the occurrence of vancomycinresistant Enterococcus species in Norwegian poultry and swine production. Microb Drug Res1999; 5: 135^ Pantosti A, Del Grosso M, Tagliabue S, Macri A, Caprioli A. Decrease of vancomycin-resistant enterococci in poultry meat after avoparcin ban. Lancet1999; 354: 741^ RaoGG,OjoF,KolokithasD.Vancomycin-resistantGram-positive cocci:riskfactorsforfecalcarriage.jhospinfect1997;35: 63^ Kirst HA,Thompson DG, NicasTI. Historical yearly usage of vancomycin. Antimicrob Agents Chemother1998; 42: 1303^ Grayson ML, Eliopoulos GM, Wennersten CB et al. Increasing resistance to beta-lactam antibiotics among clinical isolates of Enterococcus faecium: a 22-year review at one institution. Antimicrob Agents Chemother1991; 35: 2180^ Fontana R, Aldegheri M, Ligozzi M, Lopez H, Sucari A, Satta G. Overproduction of a low-a nity penicillin-binding protein and high-level ampicillin resistance in Enterococcus faecium. J Clin Microbiol1984; 38: 1980^ Cercenado E, Unal S, Eliopoulos CT et al. Characterization of vancomycin-resistance in Enterococcus durans. JAntimicrob Chemother 1995; 36: 821^ Rice LB, Carias LL, Rudin SB, Donskey CJ. Plasmid-mediated vanb-type resistance in Enterococcus faecium. In 37th Interscience Conference on Antimicrobial Agents and Chemotherapy [abstract C-157] Zervos MJ. Vancomycin-resistant Enteroccocus faecium infection in the ICU. New Horizons1996; 4: 385^ Solh N, Allignet J. Staphylococcal resistance to streptogramins and related antibiotics. Drug Res Updates1998; 1: 169^ Ronde-Fournier R, LeClercq R, Galimand M, Duval J, Courvalin P. Identi cation of the sata gene encoding a streptogramin A acetyltransferase in Enterococcus faecium BM4145. Antimicrob Agents Chemother1993; 37: Werner G, Witte W. Characterization of a new enterococcal gene satg, encoding a putitive acetyltransferase conferring resistance to streptogramin A compound. Antimicrob Agents Chemother 1999; 43: 1813^ Jones RN, Ballow CH, Biedenbach DJ, Deinhart JA, Schentag JJ. Antimicrobial activity of quinupristin^dalupristin (RP 59500, Synercid) tested against over 2800 recent clinical isolates from 200 medical centers in the United States and Canada. Diagn Microbiol Infect Dis1998; 30: 437^ Van den Bogaard AE, Mertens P, London NH, Stobberingh EE. High prevalence of colonization with vancomycin- and pristinamycin-resistant enterococci in healthy humans and pigs in The Netherlands: is the addition of antibiotics to animal feeds to blame? JAntimicrob Chemother1997; 40: 454^ Aarestrup FM, Bager F, Jensen NE, Madsen M, Meyling A, Wegener HC. Surveillance of antimicrobial resistance in bacteria isolated from food animals to antimicrobial growth promoters and related therapeutic agents in Denmark. APMIS 1998; 105: 606^ DANMAP98. Consumptionofantimicrobialagentsandoccurrenceof antibiotic resistance in bacteria, food and humans in Denmark. Copenhagen: Statens Veterinaere Serumlaboratorium, Statens Serum Institut, 1999.

6 ãðá Clinical Microbiology and Infection, Volume 6 Number 9, September DANMAP 97. Consumption of antimicrobial agents and occurrence of antibiotic resistance in bacteria, food and humans in Denmark. Copenhagen: StatensVeterinaere Serumlaboratorium, Statens Serum Institut, Bezoen A, Haren W, Hanekamp JC. Emergence of a debate: AGPs and public health. Human health and growth promoters (AGPs): Re-assessing the risk. Amsterdam: Heidelberg Appeal Nederland Foundation, 1999.