Assessment of the possible effect of the four antimicrobial treatment substances on the emergence of antimicrobial resistance 1

Transcription

1 The EFSA Journal (2008) 659, 1-26 Assessment of the possible effect of the four antimicrobial treatment substances on the emergence of antimicrobial resistance 1 Scientific Opinion of the Panel on Biological Hazards (Question No EFSA-Q ) Adopted on 6 March 2008 PANEL MEMBERS Olivier Andreoletti, Herbert Budka, Sava Buncic, Pierre Colin, John D. Collins, Aline De Koeijer, John Griffin, Arie Havelaar, James Hope, Günter Klein, Hilde Kruse, Simone Magnino, Antonio Martinez López, James McLauchlin, Christophe Nguyen-Thé, Karsten Noeckler, Birgit Noerrung, Miguel Prieto Maradona, Terence Roberts, Ivar Vågsholm, Emmanuel Vanopdenbosch. SUMMARY Following a request from European Commission (DG SANCO), the Panel on Biological Hazards was asked to deliver a scientific opinion on the possible effect of four antimicrobial treatment substances on the emergence of antimicrobial resistance. The scope of this opinion was to assess the possible development of antimicrobial resistance when chlorine dioxide, acidified sodium chlorite, trisodium phosphate and peroxyacids are applied for poultry carcasses decontamination. For the purpose of this opinion, the terms acquired reduced susceptibility to the substances used for the removal of meat surface contamination and resistance to therapeutic antimicrobials were used. Therefore, acquired reduced susceptibility to the four substances used for the removal of meat surface contamination as well as to other substances including therapeutic antimicrobials has been considered. Abattoir was the end-point of the present scientific opinion. The BIOHAZ Panel concluded that despite a long history of use, there are currently no published data to conclude that the application of chlorine dioxide, acidified sodium chlorite, trisodium phosphate or peroxyacids to remove microbial contamination of poultry carcasses at the proposed conditions of use will lead to the occurrence of acquired reduced susceptibility to 1 For citation purposes: Scientific Opinion of the Panel on Biological Hazards on a request from DG SANCO on the assessment of the possible effect of the four antimicrobial treatment substances on the emergence of antimicrobial resistance. The EFSA Journal (2008) 659, 1-26 European Food Safety Authority, 2008

2 these substances. Similarly, there are currently no published data to conclude that the application of chlorine dioxide, acidified sodium chlorite, trisodium phosphate or peroxyacids to remove microbial contamination of poultry carcasses at the proposed conditions of use will lead to resistance to therapeutic antimicrobials. Uncertainties originate from the facts that acquired reduced susceptibility to some biocides other than those in question was found followed improper use of biocides. In addition, most of the evidence on acquired reduced susceptibility to some biocides other than those in question was derived from laboratory-based experiments. The Joint AFC/BIOHAZ guidance document on the submission of data for the evaluation of the efficacy of substances for the removal of microbial surface contamination of foods of animal origin should be amended. The BIOHAZ Panel recommended that any data on the potential of occurrence for acquired reduced susceptibility to biocides and/or resistance to therapeutic antimicrobials should be included. Research on the likelihood of emergence of acquired reduced susceptibility to substances used for the removal of the microbial surface contamination of foods of animal origins and other foods and resistance to therapeutic antimicrobials should be encouraged. Key words: chlorine dioxide, acidified sodium chlorite, trisodium phosphate, peroxyacids, antimicrobial resistance, poultry decontamination The EFSA Journal (2008) 659, 2-26

3 TABLE OF CONTENTS Panel Members...1 Summary...1 Table of Contents...3 Background as provided by DG SANCO...4 Terms of Reference as provided by DG SANCO...4 Acknowledgements...4 Assessment Introduction Decontamination of carcasses Antimicrobial resistance Acquired reduced susceptibility to substances used for removal of meat surface contamination and resistance to therapeutic antimicrobials Factors affecting the efficacy of biocides Factors affecting the selection for resistance Chlorine dioxide Mode of action Mode of application and production Potential selection of acquired reduced susceptibility and resistance to therapeutic antimicrobials Acidified sodium chlorite Mode of action Mode of application Potential selection of acquired reduced susceptibility and resistance to therapeutic antimicrobials Trisodium phosphate Mode of action Mode of application Potential selection of acquired reduced susceptibility and resistance to therapeutic antimicrobials Peroxyacids Mode of action Mode of application Potential selection of acquired reduced susceptibility and resistance to therapeutic antimicrobials...16 Answer to the ToR and Recommendations...17 Answer to the ToR...17 Recommendations...17 Documentation provided to EFSA...17 References...18 Glossary...23 Appendices...24 Appendix I Resistance to therapeutic antimicrobials Clinical resistance Inherent (intrinsic) resistance Acquired resistance Cross-resistance Co-resistance...25 Appendix II...26 The EFSA Journal (2008) 659, 3-26

4 BACKGROUND AS PROVIDED BY DG SANCO In a report prepared by the Scientific Committee on Veterinary Measures relating to Public Health (SCVPH) issued on 30 October , it was stated that antimicrobial substances should only be permitted for use if a fully integrated control programme is applied throughout the entire food chain. The SCVPH opinion issued on April 2003 on the evaluation of antimicrobial treatments for poultry carcasses 3 concluded that decontamination can constitute a useful tool in further reducing the number of pathogens. Both documents stressed that antimicrobial substances shall be assessed thoroughly before their use is authorised. With the adoption of the hygiene package in 2004 and the introduction of the HACCP principles in the entire food chain, establishments will be obliged to improve their hygiene and processing procedures. In addition, Regulation (EC) No 2160/ will force Member States to initiate implementing salmonella control programmes for poultry and pigs at farm level. Under such conditions the use of substances for the removal of microbial surface contamination from food of animal origin could be considered. Article 3(2) of Regulation (EC) No 853/2004 of the European Parliament and of the Council laying down specific hygiene rules for food of animal origin, provides a legal basis to permit the use of a substance other than potable water to remove surface contamination from products of animal origin. In light of the preparation of implementing measures resulting from Regulation (EC) No 853/2004, permission for use should be preceded by a thorough scientific evaluation of all risks involved. A number of scientific evaluations have taken place on the general aspects of antimicrobial treatment of food of animal origin and on the safety and toxicological aspects of four specific substances that are considered for approval. A draft implementing measure has been proposed to allow the use of four specified substances (chlorine dioxide, trisodium phosphate, acidified sodium chlorite and peroxyacids) for the removal of surface contamination of poultry carcasses. The draft Commission Regulation will lay down detailed specifications for the use of the four substances including conditions of use. Recently the Commission has prepared a request to the Scientific Committee on Emerging and Newly Identified Health Risks - SCENIHR for an overall assessment of the antibiotic resistance effects of biocides. Furthermore, EFSA has initiated a self-tasking project on antimicrobial resistance. In the light of these initiatives it is necessary to perform an assessment of the impact on the environment of the four substances mentioned above. Moreover, it is necessary to investigate if it is possible that the use of the four substances could lead to antimicrobial resistance in the micro-organisms TERMS OF REFERENCE AS PROVIDED BY DG SANCO To assess the possible effect on the emergence of antimicrobial resistance in case chlorine dioxide, acidified sodium chlorite, trisodium phosphate and peroxyacids were applied according to the proposed conditions of use as a substance to remove microbial surface contamination from poultry carcasses. ACKNOWLEDGEMENTS The European Food Safety Authority wishes to thank the members of the Working Group for the preparation of this opinion: F. Aarestrup, B. Carpentier, P. Hartemann (SCENIHR), G. Klein (Chair & Rapporteur) & M. Prieto. 2 Report by SCVPH on "Benefits and limitations of antimicrobial treatments for poultry carcasses" (1998). 3 Opinion of the SCVPH on the "Evaluation of antimicrobial treatments for poultry carcasses" (2003) 4 OJ L 325, , p. 1. The EFSA Journal (2008) 659, 4-26

5 ASSESSMENT 1. Introduction 1.1. Decontamination of carcasses The existing European Community (EC) legislation includes a number of provisions that enhance food safety as well as control and prevent pathogenic microorganisms. With the adoption of the new Regulations on hygiene 5 and the introduction of the principles of Hazard Analysis Critical Control Points (HACCP) in the entire food chain, establishments are obliged to improve their hygiene and processing procedures including verification and validation of the systems used. Moreover, Regulation (EC) No 2160/ sets rules on the proper and effective measures that shall be taken to detect and control Salmonella and other specified zoonotic agents at all relevant stages of production, processing and distribution in order to reduce their prevalence and the risk they pose to public health. Decontamination or removal of microbial surface contamination of carcasses involves the application of a substance at a given step during the slaughter process. At present, there is no harmonised terminology of these substances or the treatment itself. Sanitizer, disinfectant, sterilant, antiseptics, biocides, antimicrobials and decontaminating agent are most widely used for substances intended to be used for the removal of microbial surface contamination of food of animal origin (see glossary). In regard to the application of these substances, antimicrobial treatment of carcasses and carcass decontamination are terms often used in the scientific literature. Antimicrobial treatments could be an effective measure for reducing the microbial contamination of carcasses but should never be used as the primary measure. Previous opinions of the Scientific Committee on Veterinary Measures relating to Public Health and of the Scientific Panel on Biological Hazards (SCVPH, 1998; SCVPH, 2003; EFSA, 2006a) concluded that antimicrobial treatments of carcasses should be considered as supplementary means of reducing the microbial load of foods of animal origin and should be a part of an integrated control programme throughout the whole food chain. Article 3(2) of Regulation 853/2004 of the European Parliament and Council, which lays down specific hygiene rules for food of animal origin, constitutes the legal basis for the use of substances other than potable water or clean water to remove surface contamination from foods 7 of animal origin intended for human consumption. The use of substance(s) for the removal of microbial surface contamination of foods of animal origin is authorised according to the legislative procedures of the EC. However, the EC shall consult EFSA on any matter within the scope of Regulation 853/2004 that could have a significant impact on public health. Indeed, EFSA has already assessed the safety and efficacy of substances intended to be used for the removal of microbial surface contamination of food of animal origin (EFSA, 2005a; EFSA, 2005b; EFSA, 2006b; EFSA, 2006c). Moreover, the BIOHAZ Panel in collaboration with AFC 8 Panel developed a guidance document on the submission of data for the evaluation of the safety and efficacy of substances for the removal of microbial surface contamination of foods of animal origin intended for human consumption (EFSA, 2006a). If antimicrobial treatments of carcasses are to play a role in control of foodborne pathogens, all interested parties on food safety must know more about the potential for development of resistance among target 5 Regulation 852/2004: OJ L226/ , p3; Regulation 853/2004: OJ L226/ , p22 6 OJ L 325, , p For the purpose of this document foods of animal origin are limited to fresh unprocessed products of animal origin 8 Panel on food additives, flavourings, processing aids and materials in contact with food The EFSA Journal (2008) 659, 5-26

6 microorganisms. Evidence of possible development of acquired reduced susceptibility of bacteria to the active substance is one of the criteria recommended on the Joint AFC/BIOHAZ guidance document on the submission of data for the evaluation of the safety and efficacy of substances for the removal of microbial surface contamination of foods of animal origin intended for human consumption. Meanwhile, the SCENIHR has received a request on the assessment of the antibiotic resistance effects of biocides. In light of recent scientific evidence, clarification is sought as to whether cross resistance to antibiotics should be an additional criterion to consider in the common principles for the evaluation of dossiers for biocidal products as laid out in Annex VI of the Directive 98/8/EC 9 of the European Parliament and of the Council on the placing of biocidal products on the market. The scope of this opinion is to assess the possible development of antimicrobial resistance when chlorine dioxide, acidified sodium chlorite, trisodium phosphate and peroxyacids are applied for poultry carcasses decontamination. For the purpose of this opinion, the terms acquired reduced susceptibility to the substances used for the removal of meat surface contamination (see section 2 for definition) and resistance to therapeutic antimicrobials (see Appendix I for definition) will be used. Therefore, acquired reduced susceptibility to the four substances used for the removal of meat surface contamination as well as to other substances including therapeutic antimicrobials will be considered. Abattoir is the end-point of the present scientific opinion. The impact of releasing the biocides to the effluents at abattoir is out of the scope of this opinion and will not be considered Antimicrobial resistance During the last decade, concern about resistance to therapeutic antimicrobials by various mechanisms has increased world wide after occasional treatment failures in human and animal infectious diseases (EARSS, 2005; Harbarth and Samore, 2005). Concern has been also raised about the potential for microorganism(s) to develop resistance to substances used in food manufacture such as biocides. Biocides (for definition see glossary), and as opposed to therapeutic antimicrobials, have multiple target sites against the microbial cells. Thus the emergence of acquired reduced susceptibility is unlikely to be caused by the modification of a target site or the bypass of a metabolic process. Other mechanisms for antimicrobial resistance such as impermeability, efflux and degradation have been described. It is likely that these mechanisms operate synergistically although there have been very few studies investigating multiple bacterial mechanisms of acquired reduced susceptibility following exposure to a biocide. Acquired reduced susceptibility to biocides has been reported since the 1950s, particularly with the use of cationic biocide formulations (Russell, 2002a; Chapman, 2003). In most instances acquired reduced susceptibility emerged following the improper use or storage of the formulations, resulting in a decrease in the effective concentration (Sanford, 1970; Prince and Ayliffe, 1972; Chapman, 2003). Moreover, it has to be noted that most of the evidence on acquired reduced susceptibility to biocides remains with laboratory-based experiments investigating some agents such as cationic biocides (Tattawasart et al., 1999; Thomas et al., 2000), isothiazolones (Winder et al., 2000), phenolics (McMurry et al., 1998; McMurry et al., 1999), hydrogen peroxide (Dukan and Touati, 1996) and other compounds (Walsh et al., 2003). Some of the resistance mechanisms are intrinsic (or innate) to the microorganism while others have been acquired through selection of mutations or through the acquisition of genetic elements (Poole, 2002). Innate mechanisms can confer high-level resistance to biocides. The 9 Directive 98/8 of the European parliament and of the Council concerning the placing of biocidal products on the market. OJ L 123, p-63. The EFSA Journal (2008) 659, 6-26

7 most common intrinsic resistance mechanism is the impermeability barrier, which is found in spores (Russell, 1990; Russell et al., 1997; Cloete, 2003), but also in vegetative bacteria such as mycobacteria and to some extent Gram-negative bacteria. The impermeability barrier limits the amount of a biocide that penetrates within the cell, thus decreasing the effective biocide concentration (Denyer and Maillard, 2002; Lambert, 2002; Champlin et al., 2005). Another mechanism that has been described in the literature and has increased in importance as a resistance mechanism over the past decade is the presence of efflux pumps, which decreases the intracellular concentration of toxic compounds, including biocides (Paulsen et al., 1996; Brown et al., 1999; Levy, 2002; Poole, 2002; McKeegan et al., 2003; Piddock, 2006b). The importance of efflux pump in terms of acquired reduced susceptibility to biocides might be considered as modest and is usually measured as an increase in MICs (minimum inhibitory concentrations). The enzymatic degradation of biocides has also been described as a resistance mechanism in bacteria, notably to heavy metals (e.g. silver and copper; enzymatic reduction of the cation to the metal) (Cloete, 2003); parabens (Valkova et al., 2001), aldehydes (formaldehyde dehydrogenase) (Kummerle et al., 1996), peroxygens (catalase, super oxide dismutase and alkyl hydroperoxidases mopping up free radicals) (Demple, 1996). Environmental degradation of various compounds has been described notably among pseudomonads and complex microbial communities. It is however unclear as to the importance of these mechanisms when biocides are used in high concentrations. As for efflux, increase resistance following degradation of biocides has been measured as an increase in MICs but not necessarily as a decrease in lethal activity. The modification of target sites has been described on rare occasions but does not seem to be widespread among bacteria, although there is a paucity of information on this subject. There is little information on the effect of biocides on the transfer of genetic determinants. Exposure of Salmonella Enteritidis to chlorine (25 mg/l) has been shown to induce the mar operon and confer cross-resistance to therapeutic antimicrobials such as chloramphenicol, tetracycline, ciprofloxacin and nalidixic acid (Potenski et al., 2003). One study highlighted that while some biocides at a sub-inhibitory (residual) concentration could inhibit genetic transfer, other increased genetic transfer efficiency (Pearce et al., 1999). The induction of acquired reduced susceptibility, following exposure to a low concentration of a biocide has been reported in a number of studies. The mechanisms involved include the overexpression of efflux pumps (Maira-Litran et al., 2000; Gilbert and McBain, 2003; Randall et al., 2007), over expression of multigene systems such as soxrs and oxyr (Dukan and Touati, 1996), production of guanosine 5 -diphosphate 3 -diphosphate (ppgpp) (Greenway and England, 1999). These mechanisms are parts of the stress-response systems in bacteria, for which more evidence is available in the literature. In some occasions, a specific mechanism has not been established and just a phenotypic change leading to the emergence of resistance to several unrelated compounds in vitro has been reported (Chapman, 2003; Walsh et al., 2003; Thomas et al., 2005). It has become clear that attached cells express properties distinct from planktonic cells, such as an increased resistance to biocides. Recent work has indicated that slow growth and/or induction of an Rpos mediated stress response could contribute to biocide resistance (Mah and O'Toole, 2001). The EFSA Journal (2008) 659, 7-26

8 2. Acquired reduced susceptibility to substances used for removal of meat surface contamination and resistance to therapeutic antimicrobials For the purpose of this opinion, acquired reduced susceptibility to the substances in question is defined as the situation when a bacterium develops tolerance to higher bacteriostatic or bacteriocidal concentrations than phenotypically related bacteria of the original or wild type strain. The most frequently used methods aim to determine acquired reduced susceptibility are minimal inhibitory concentration (MIC) and minimal bactericidal concentration (MBC). MIC is defined as the lowest concentration of one of the substances in question that will inhibit the growth of a given microorganism under standardised growth conditions. MBC is determined as the lowest concentration achieving a set number of decimal reductions in a given time at a given temperature using standardized methods (NF EN 1276). Such determination does not take into account the shape of the survival curve, i.e. existence of a shoulder and/or a tail is ignored. Such tails can be observed when attached cells or biofilms from pure culture are challenged against either biocides or therapeutic antimicrobials. The surviving cells are called persister cells (Lewis, 2007). It should be noted that MBC and MIC are not related i.e. the same bacterial strain may have a low MIC and a high MBC (Soumet, 2006). For the purpose of this opinion, resistance to therapeutic antimicrobials is related to clinical resistance, inherent resistance, acquired resistance, cross-resistance and co-resistance. The description of these terms is presented in the Appendix I. 3. Factors affecting the efficacy of biocides The efficacy 10 of biocides depends on a range of factors, (Reuter, 1984; Reuter, 1989; Reuter, 1994; EFSA, 2006a), both intrinsic and extrinsic factors. Intrinsic factors are characteristics of the biocidal agent and its application. For example, the concentration itself is crucial. If not prepared correctly the necessary concentrations are not reached. Therefore compliance by the user manufacturer s instructions is important. Contact time on the target surface, e.g. the carcass, must be also defined prior to use for a given efficacy. Furthermore, the combination of contact time and concentration determines the effectiveness of the application. The stability of the active compounds in the formulation of the biocide also influences the efficacy. Storage times or unsuitable environment for storage may reduce the efficacy. Some of these factors have also been observed in relation to the four substances in question, e.g. trisodium phosphate (Capita et al., 2002b). Extrinsic factors derive from the environment during application. Temperature of the environment and the target surface are of importance, as most substances have a lower efficacy at cooling temperatures. Protein fractions on the surface reduce the efficacy as they interact with the substance. The application on carcasses is a scenario, where proteins always are present, either on the surface itself and/or in the chilling water. The mode of application, such as spraying or dipping, also influences the efficacy, as the contact time of the solution is of importance as well as mechanical effects (both by spraying and dipping) and also the water replacement rate at dipping. ph is another important factor. The concentration of the microorganisms on the carcass, the age of the attached bacterial community can be also considered as extrinsic factors influencing efficacy. 10 The use, under defined conditions, of substance(s) for the removal of microbial surface contamination will be regarded efficacious when any reduction of the prevalence and/or numbers of pathogenic target bacteria is statistically significant when compared to the control (e.g. water) and this reduction at the same time is of human relevance (EFSA, 2006a). The EFSA Journal (2008) 659, 8-26

9 Some of these factors have been observed also in relation to the four substances in question, e.g. trisodium phosphate (Capita et al., 2002a; Capita et al., 2002b; Sampathkumar et al., 2004). For instance, the natural microflora of poultry carcasses is far less susceptible and/or better protected through the presence of organic material to bactericidal concentrations than bacterial cells that have been deposited on the poultry skin under experimental conditions before the biocide application or bacterial cells on equipment surfaces such as steel. 4. Factors affecting the selection for resistance The use of biocides can select for resistance to therapeutic antimicrobials in four different ways: 1) Cross-resistance: (i) selection for genes encoding resistance to both the biocidal substance and one or more therapeutic antimicrobial classes or (ii) change the physiological response of the bacterium to become less susceptible to both biocidal substance and therapeutic antimicrobial agents. 2) Co-resistance: selection for clones or mobile elements also carrying antimicrobial resistance. 3) Indirectly select for clones that are resistant to therapeutic antimicrobials. 4) Enhance DNA uptake by e.g. activating a SOS 11 response in bacteria. For a full evaluation of the potential selection for antimicrobial resistance by the use of biocidal products is necessary to study all four potential ways of selection. Limiting a study to one or two could potentially miss important knowledge. In only a few instances has cross-resistance been observed between biocides other than the four substances in question and therapeutic antimicrobials. This has mainly been caused by efflux pumps mediating reduced susceptibility to both classes of antimicrobial agents (Piddock, 2006a; Piddock, 2006b). Co-resistance can occur when mechanisms encoding resistance or reduced susceptibility are genetically linked. This is for example the case for tolerance to quaternary ammonium compounds in Gram-negative bacteria. The qac-genes are often together with sul1 genes encoding sulphonamide resistance located as part of class I integrons which also can harbour various other resistance genes. Class I integrons can be located on plasmids or in the chromosome. The use of the biocidal substances might also select indirectly for resistance to therapeutic antimicrobials by causing a clonal drift in the bacterial population towards isolates that are more resistant. Clonal drift has commonly been observed as the cause of changes in the overall occurrence of resistance to therapeutic antimicrobials among food borne pathogens. As an example the emergence of multi-drug resistant Salmonella Typhimurium DT104 caused an overall increase in the occurrence of resistance to therapeutic antimicrobials among Salmonella from food animals and humans in several countries. This change was most likely not associated to changes in the usage of therapeutic antimicrobials, but simply caused by a change in the clones found in the different reservoirs. Exposure to a biocide is a major stress factor. Thus, it must be expected that a biocide can initiate a SOS response in a bacterium which can promote horizontal gene transfer of both therapeutic antimicrobial and virulence genes (Beaber et al., 2004; Ubeda et al., 2005). 11 SOS response is an inducible DNA repair system that allows bacteria to survive sudden increases in DNA damage. The EFSA Journal (2008) 659, 9-26

10 Sub-lethal treatment concentration may result from the inappropriate consideration of the factors that affect the efficacy and therefore may induce the development of less susceptible microorganisms. 5. Chlorine dioxide 5.1. Mode of action Synonym: Chloroperoxyl, Chlorine (IV) oxide Chemical name: Chlorine peroxide CAS Registry Number: Chemical formula: ClO 2 Description: Greenish yellow to orange gas with a pungent odour Chlorine dioxide is an oxidizing biocide. Chlorine dioxide kills microorganisms by direct action on the cellular membrane and through disruption of fundamental cellular processes (USDA, 2002a). Chlorine dioxide is a synthetic yellowish green gas with chlorine-like odor. ClO 2 is unstable as a gas and will undergo decomposition into chlorine gas (Cl 2 ), oxygen gas (O 2 ), and heat. However, ClO 2 is stable and soluble in an aqueous solution and does not form hypochlorous acid or react with ammonia. It functions independent of ph and can provide excellent control at a fraction of the chlorine dosage because it can be used at much lower doses (Keener et al., 2004). At high concentrations, chlorine dioxide deactivates microorganisms by breaking the cell wall. This disinfecting action occurs immediately upon exposure. Chlorine dioxide is relatively inert in reacting with individual amino acids and reacts very quickly with peptides and proteins. Upon reaching the cell wall, chlorine dioxide can react with certain membrane proteins and alter the permeability of the cell membrane, or in higher concentrations, rupture the cell wall. At lower concentrations, chlorine dioxide alters the outer membrane proteins and lipids sufficiently to disrupt the permeability of the outer membrane. Chlorine dioxide is then able to penetrate bacterial cell walls and disrupt protein synthesis by virtue of its reaction with amino acids and nucleotides. This reaction with amino acids in protoplasm prevents the cell from producing proteins, thereby killing the cell. The primary physiological mode of inactivation of bacteria by chlorine dioxide has been attributed to a disruption of protein synthesis. The mode of action of chlorine dioxide on Escherichia coli was assessed by Berg et al. (1986) by studying outer membrane permeability to macromolecules and potassium, and observing effects on respiration. The results indicate that gross cellular damage involving significant leakage of intracellular macromolecules does not occur. There was a substantial efflux of potassium, however, and respiration was inhibited even at sublethal doses. It was concluded that the inhibition of respiration, which could be due to the damage to the cell envelope, was not the primary lethal event. Observations of the efflux of K+ strongly implicate the loss of permeability control as the primary lethal event at the physiological level, with nonspecific oxidative damage to the outer membrane leading to the destruction of the trans-membrane ionic gradient Mode of application and production According to the United States regulations, chlorine dioxide may be used as an antimicrobial agent in water used in poultry processing in an amount not to exceed 3 ppm residual chlorine The EFSA Journal (2008) 659, 10-26

11 dioxide (21 CFR ). Chlorine dioxide is typically used in poultry processing in the United States either as sprays or washes for on-line reprocessing, or added to the chiller water (chiller bath applications) to limit the potential for microbial cross-contamination (more information is given in Appendix II). For use as an antimicrobial agent it is added to water in a concentration up to 50 mg/l in order to maintain a residual concentration of 2.5 mg/l (USDA, 2002a). In the process used for this application chlorine dioxide is described as rapidly transformed to chlorite and chlorate ions in a ratio of 7:3. Thus, the concentrations of chlorite and chlorate would be 33 and 14 mg/l, respectively. Only 2.5 mg/l (about 5% of the initial content) remains as chlorine dioxide. They are different ways of producing chlorine dioxide: use of sodium chlorite (NAClO 2 ) and an acid: this can be achieved by using hydrochloric acid (HCl), but also any other acids. H + may be also generated by electrolytic way. However, weak acids partly need significantly longer reaction duration and a higher surplus has to be added in order to achieve a complete reaction to ClO 2 ; use of sodium chlorite and chlorine gas (chlorine chlorite procedure) and the reaction balance is not 100% on the side of ClO 2. This is important because the traces of NaClO 2 and Cl 2 have different reactional properties. For example for the formation of chlorophenols (which gave bad flavour) the transfer of one chlorine atom on one phenol atom is necessary. In cases of pure chlorine dioxide, phenols are oxidised without transfer of chlorine, but phenol reverts by releasing an electron and chlorine dioxide becomes chlorite. Hypochlorous acid (HOCl) can also chlorinate a phenol ring. For a good reaction ph must be 3.5 or less; use of sodium chlorate and acid, but this way of production seems to be limited to the paper factories, because of the toxicity of residual chlorate. Thus it is crucial to know that there are procedures using ready-made dilutions that are free of active chlorine (acid-chlorite procedure) and those that are not free of active chlorine. In aqueous solution photolytic disintegration of ClO 2 begins to happen at wave lengths less than 432 nm into chlorine, chlorite an especially into chlorate and chlorite; with increasing temperature (starting at approx 40 C) also rapid thermal degradation takes place. The ph value is also very important. Thus in essence the stability of chlorine dioxide solutions depends on: mode of production; temperature (the higher, the faster disintegration takes place); storage (the more aerial contact and light irradiation, the faster disintegration takes place); application conditions (presence of wasting matters like organic matter or for instance iron or manganese ions); container material (diffusion of chlorine dioxide by plastic material of containers). For these reasons it is very important to manage properly the conditions of application, because concentrations of Cl 2, chlorites and chlorates as impurities are of great interest. The EFSA Journal (2008) 659, 11-26

12 5.3. Potential selection of acquired reduced susceptibility and resistance to therapeutic antimicrobials At this moment there is no published information indicating the occurrence of cross-resistance or co-resistance between chlorine dioxide and therapeutic antimicrobials. No studies have been published on the selection for resistance to therapeutic antimicrobials by enhancing DNA uptake due to exposure to ClO 2. Mechanisms which could allow ClO 2 exposure to influence resistance to therapeutic antimicrobials have not been reported. There is also no indication that the use of ClO 2 could support the spread of resistance to therapeutic antimicrobials by direct selection, although it may be possible by indirect selection. Strains with reduced susceptibility to hydrochlorous acid (HOCl) were reported for Salmonella and other species (Escherichia coli) (Chesney et al., 1996; Mokgatla et al., 2002). This phenomenon seems not to have been observed for ClO 2, even if the mode of action seems to be identical. Chlorine compounds are highly active oxidating compounds and probably exhibit their action by destroying the activity of various proteins (Bloomfield, 1996). The activity of chlorine compounds is dependent on ph and has highest activity at low ph (McDonnell and Russell, 1999). It is well known that bacteria have a natural protection against oxidants such as superoxide and hydrogen peroxide. In contrast there is limited knowledge on protection against other oxidants. For investigating the possible protective mechanisms involved in the reduced susceptibility to HOCl, Mokgata et al. (2002) have studied one resistant Salmonella strain and a sensitive one at a final active concentration of 28 mg/l of HOCl. The resistant Salmonella isolate differed from the sensitive in production of catalase and decrease of the activity levels of hydroxyl radicals and singlet oxygen, moieties thought to be integrally involved in the antibacterial action of HOCl. Furthermore, the resistant strain did not display the same degree of DNA damage as the sensitive. A single study has indicated that the production of glutathion can protect Escherichia coli against the activity of chlorine (Chesney et al., 1996). There are, however, no reports of naturally occurring chlorine resistant bacteria or cross-resistance. 6. Acidified sodium chlorite 6.1. Mode of action Definition: Acidified sodium chlorite is a combination of sodium chlorite and any acid generally approved in food Synonym: Acidified chlorite Chemical name: Sodium chlorite (Chlorous acid, sodium salt) CAS Registry Number: Chemical formula: NaClO 2 Description: Clear, colourless, liquid The antimicrobial action of acidified sodium chlorite is derived from chlorous acid. The level of chlorous acid in an acidified sodium chlorite solution is determined by the ph of the solution. Chlorous acid and chlorine dioxide, both uncharged, are able to disrupt the permeability of the outer membrane of bacterial cell walls and penetrate them to disrupt protein The EFSA Journal (2008) 659, 12-26

13 synthesis by virtue of reactions with sulfhydryl, sulfide, and disulfide containing amino acids and nucleotides (USDA, 2002b). Sodium chlorite, at a concentration of mg/l, is activated with any acid approved for use in foods at levels sufficient to provide solutions with ph values in the range for either a 15 second spraying or 5-8 second dipping. In the case of immersion in chilling water, the concentration is up to 150 mg/l at ph between 2.8 and 3.2. The main active ingredient of acidified sodium chlorite solution is chlorous acid which is a very strong oxidizing agent, stronger than either chlorine dioxide or chlorine. The level of chlorous acid depends on the ph of the solution. So, 31% is formed at ph 2.3, near 10% at ph 2.9 and only 6% at ph 3.2. The potential formation of chlorine dioxide is limited, not exceeding 1-3 mg/l (EFSA, 2005a). Chlorous acid is also thought to facilitate proton leakage into cells and thereby increase the energy output of the cells to maintain their normal internal ph. This effect also adversely affects amino acid transport. The primary physiological mode of inactivation of bacteria by both chlorous acid and chlorine dioxide has been attributed to a disruption of protein synthesis (USDA, 2002b). It is considered a broad-spectrum oxidative antimicrobial, effective on pathogenic bacteria as well as viruses, fungi, yeast, molds, and some protozoa Mode of application Acidified sodium chlorite is typically used in poultry processing in the United States either as sprays or washes for on-line reprocessing, or added to the chiller water (chiller bath applications) to limit the potential for microbial cross-contamination (more information is given in Appendix II). According to the United States regulations, acidified sodium chlorite may be used as an antimicrobial agent in poultry processing waters applied as a pre-chiller or chiller solution at levels that result in sodium chlorite concentrations between 50 and 150 ppm in combination with any acid considered GRAS in the United States at levels sufficient to achieve a solution ph of 2.8 to 3.2 (21 CFR ). Acidified sodium chlorite may be also used as an antimicrobial agent in poultry processing waters applied as a spray or dip solution at levels that result in sodium chlorite concentrations between 500 and 1,200 ppm, in combination with any acid considered GRAS in the United States at levels sufficient to achieve a solution ph of 2.3 to 2.9 (21 CFR ) Potential selection of acquired reduced susceptibility and resistance to therapeutic antimicrobials At this moment there is no published information indicating the occurrence of cross-resistance or co-resistance between acidified sodium chlorite and therapeutic antimicrobials. No studies have been published on the selection for resistance to therapeutic antimicrobials by enhancing DNA uptake due to exposure to acidified sodium chlorite. Mechanisms which could allow acidified sodium chlorite exposure to influence resistance to therapeutic antimicrobials have not been reported. At this moment, there is no indication that the use of acidified sodium chlorite could support the spread of resistance to therapeutic antimicrobials by direct selection, although it may be possible by indirect selection. The EFSA Journal (2008) 659, 13-26

14 7. Trisodium phosphate 7.1. Mode of action The chemistry of trisodium phosphate is as follows: Synonym: Trisodium monophosphate Chemical name: Trisodium orthophosphate CAS Registry Number: Chemical formula: Na PO 3 4 Description: Colourless or white crystals Trisodium phosphate is typically used in aqueous solutions containing 8 to 12% with a high ph value (ph 12). The solution is kept at a temperature between 7 and 13ºC. The mechanism of action of trisodium phosphate as an antimicrobial agent is reported to result from its high alkalinity in solution. A 1% solution of trisodium phosphate in water has a ph of This high alkalinity can help remove fat films to allow the chemical to contact more bacteria as well as disrupt fatty molecules in the cell membrane causing the cells to leak intracellular fluid. It is also thought that trisodium phosphate may help prevent attachment of bacteria to poultry skin and may act as a surfactant to help facilitate the removal of bacteria from carcasses (USDA, 2002c). Trisodium phosphate exerts a destructive effect on pathogens and a detergent effect that allows the removal of bacteria by the washing process. The lowest effective concentration for microbial control is 8%. Trisodium phosphate is ionised in water generating Na + 3- and PO ions 4 (EFSA, 2005a). The mechanisms of the trisodium phosphate mode of action include (i) surfactant properties, (ii) destruction due to high ph, (iii) removal of loosely associated bacteria from the skin, (iv) removal of carcass surface fat, invariably resulting in removal of bacteria attached to the fat, and (v) destruction of the bacterial cell wall (Capita et al., 2002b; Keener et al., 2004). It is hypothesised that the increased wetting ability of hot water and trisodium phosphate physically remove bacteria in addition to killing them Mode of application Trisodium phosphate is typically used in poultry processing in the United States either as sprays or washes for on-line reprocessing, or added to the chiller water (chiller bath applications) to limit the potential for microbial cross-contamination (more information is given in Appendix II). Trisodium phosphate, has been approved by regulatory agencies in the United States for use in poultry process water within a concentration range of 8 to 12%. The solution must be maintained at a temperature of 7.2 to 12.8 o C and applied by spraying or dipping carcasses for up to 15 seconds (9 CFR (c) and FDA regulation 21 CFR ). Since 1994, interim approval has been granted to use trisodium phosphate as a processing aid for the purpose of reducing microorganisms when: (1) applied as a spray or dip to raw, unchilled carcasses for up to 15 seconds in an 8 to 12 % solution of trisodium phosphate maintained within a temperature range of 18.3 to 29.4 o C; and (2) applied as a spray or dip to raw, unchilled poultry giblets for The EFSA Journal (2008) 659, 14-26

15 up to 30 seconds with an 8 to 12% solution of trisodium phosphate, until rulemaking approving its use can be finalised. Carcass exposure time is controlled by line speed and length of the application cabinet. A typical application is approximately 15 seconds at full line speed. If the line is stopped for more than 5 minutes, carcasses in the application cabinet are segregated and condemned Potential selection of acquired reduced susceptibility and resistance to therapeutic antimicrobials Some bacterial strains sometimes show a different spectrum of intrinsic susceptibility to trisodium phosphate (Fratamico et al., 1996; Capita-González et al., 2000; Sampathkumar et al., 2004). This has been shown for Salmonella, Listeria monocytogenes and E. coli. However differences were small and do not allow differentiation between resistant and susceptible isolates. Nothing is known regarding potential resistance mechanisms or genetic location. At this moment there is no published information indicating the occurrence of cross-resistance between trisodium phosphate and therapeutic antimicrobials. No studies have been published on the selection for resistance to therapeutic antimicrobials by enhancing DNA uptake due to exposure to trisodium phosphate. Mechanisms which could allow trisodium phosphate exposure to influence resistance to therapeutic antimicrobials have not been reported. At this moment, there is no indication that the use of a trisodium phosphate could support the spread of resistance to therapeutic antimicrobials by direct selection, although it may be possible by indirect selection. 8. Peroxyacids 8.1. Mode of action Definition: Synonym: Chemical name: CAS Registry Number: Chemical formula: Formulation of peroxyacetic acid (<15%), peroxyoctanoic acid (<2%) and Hydrogen Peroxide <10%) Peroxyacids, acetyl peroxide, acetyl hydroperoxide Ethaneperoxoic acid, octaneperoxoic acid and hydrogen dioxide , and , respectively C 2 H 4 O 3, C 8 H 16 O 3 and H 2 O 2, respectively Description: Clear, colourless, liquid 1-hydroxyethylidene-1,1-diphosphonic acid (HEDP) is usually added as stabilizer because of its metal chelating activity. Acetic and octanoic acids are also present in the peroxyacids solution. Acetic acid acts as an acidifier and octanoic acid as a surfactant. The peroxyacid solution is a mixture of peroxyacetic acid, peroxyoctanoic acid, acetic acid, octanoic acid, hydrogen peroxide and HEDP (EFSA, 2005a). The mixture is used at a maximum concentration of total peroxyacid, expressed as peroxyacetic acid, of 220 mg peroxyacetic acid per L, a maximum concentration of hydrogen peroxide of 110 mg per L, and a maximum concentration of HEDP of 13 mg per L (USDA, 2002d). A combined amount of peroxyacids, expressed as peroxyacetic acid, is usually given due to the difficulties in the analytical differentiation between peroxyacetic and peroxyoctanoic acids. Oxidation/reduction reactions take place in water generating water, acetic acid and octanoic The EFSA Journal (2008) 659, 15-26

16 acid. Upon application of the peroxyacids solution to the carcasses, acetic acid, octanoic acid, water and oxygen are generated as natural breakdown products (USDA, 2002d; EFSA, 2005a). The antimicrobial action of a peroxyacid mixture derives from the oxidative capacity of the constituents. Strong oxidizing agents are able to disrupt the permeability of the cell membrane and penetrate bacterial cell walls and disrupt protein synthesis by virtue of reactions with sulfhydryl, sulfide, and disulfide containing amino acids and nucleotides (USDA, 2002d). The active ingredients of peroxyacids are peroxyacetic acid and octanoic acid, and to a lesser extent peroxyoctanoic acid and hydrogen peroxide. Peroxyacids are a class of strongly oxidizing compounds. They kill microorganisms by direct action on the cellular membrane and through disruption of fundamental cellular processes. A secondary mechanism could be the acidification of the carcass surface (SCVPH, 2003) Mode of application Peroxyacids is typically used in poultry processing in the United States either as sprays or washes for on-line reprocessing, or added to the chiller water (chiller bath applications) to limit the potential for microbial cross-contamination (more information is given in Appendix II). According to the United States regulations, a mixture of peroxyacetic acid, octanoic acid, acetic acid, hydrogen peroxide, peroxyoctanoic acid, and HEDP may be used as an antimicrobial agent on poultry carcasses, poultry parts, and organs in accordance with current industry standards of good manufacturing practice where the maximum concentration of peroxyacids is 220 ppm as peroxyacetic acid, the maximum concentration of hydrogen peroxide is 110 ppm, and the maximum concentration of HEDP is 13 ppm (21 CFR ) Potential selection of acquired reduced susceptibility and resistance to therapeutic antimicrobials Sub-inhibitory concentrations of acids have previously been shown to lead to acid tolerance response for many food poisoning microorganisms (Hill et al., 1995). Adaptation to oxidative stress is also well known (Farr and Kogoma, 1991). No specific studies on both areas have been conducted for peroxyacids. Goni-Urriza et al. (2000) found no correlation between high MBCs of peroxyacids and resistance to therapeutic antimicrobials among 138 strains of Aeromonas isolated from European rivers. Arturo-Schaan et al. (1996) showed that although peroxyacids disinfection effectively reduced the number of E. coli strains isolated from sewage, the percentage of E. coli strains containing plasmids was not statistically different before and after peracetic acid disinfection. At this moment there is no published information indicating the occurrence of cross-resistance between peroxyacids and therapeutic antimicrobials. No studies have been published on the selection for resistance to therapeutic antimicrobials by enhancing DNA uptake due to exposure to peroxyacids. Mechanisms which could allow peroxyacids exposure to influence resistance to therapeutic antimicrobials have not been reported. At this moment, there is no indication that the use of peroxyacids could support the spread of resistance to therapeutic antimicrobials by direct selection, although it may be possible by indirect selection. The EFSA Journal (2008) 659, 16-26