EUROPA > DG Health and Consumer Protection > Public Health  Contact | Search | What's New? | Subscribe | Site Map | Index 
Biocides home
Source document:
SCENIHR (2009)

Summary & Details:
GreenFacts (2009)

Effects of Biocides on antibiotic resistance

3. Is there evidence that bacteria resistant to biocides are emerging?

3.1 How can bacterial resistance to biocides be determined?

The SCENIHR opinion states:

3.4. Resistance to biocides

3.4.1. Occurrence of resistance

Bacterial resistance to biocides has been reported since the 1950s, particularly with the contamination of cationic biocide formulations (Adair et al. 1971, Chapman 2003, Russell 2002b). In most instances bacterial resistance emerged following the improper use or storage of the formulations, resulting in a decrease in the effective concentration (Centers for Disease Control 1974, Prince and Ayliffe 1972, Russell 2002b, Sanford 1970). Bacterial resistance to all known preservatives has also been reported (Chapman 1998, Chapman et al. 1998).

In the health care setting, bacterial resistance to biocides has long been reported with compounds such as: chlorhexidine (Stickler 1974); quaternary ammonium compounds (Gillespie et al. 1986, Romao et al. 2005); bisphenol, triclosan (Bamber and Neal 1999, Heath et al. 1998, Sasatsu et al. 1993); iodophor (O’Rourke et al. 2003); parabens (Flores et al. 1997, Hutchinson et al. 2004); and more reactive biocides such as glutaraldehyde (Fraud et al. 2001, Griffiths et al. 1997, Manzoor et al. 1999, Nomura et al. 2004, Van Klingeren and Pullen 1993, Walsh et al. 2001) and peroxygens (Dukan and Touati 1996, Greenberg et al. 1990, Greenberg and Demple 1989). In a recent study, Smith and Hunter reported that although biocides may be effective against planktonic populations of bacteria such as methicillin-resistant Staphylococcus aureus (MRSA) and Pseudomonas aeruginosa, some biocides currently used in hospitals are ineffective against nosocomial pathogens growing as biofilms attached to surfaces and fail to control this reservoir for hospital-acquired infections (Smith and Hunter 2008). Concerning triclosan, Tabak and colleagues reported that the tolerance of Salmonella in the biofilm was attributed to low diffusion through the extracellular matrix, while changes of gene expression might provide further resistance both to triclosan and to other antimicrobials (Tabak et al. 2007).

However, most of the evidence on bacterial resistance to biocides comes from laboratory-based experiments which investigated a wide range of agents such as: cationic biocides (Tattawasart et al. 1999, Thomas et al. 2000); isothiazolones (Winder et al. 2000); phenolics (McMurry et al. 1998b, McMurry et al. 1999); hydrogen peroxide and peracetic acid (Dukan and Touati 1996) and other compounds (Walsh et al. 2003).

3.4.2. Biocide concentration and bacterial susceptibility

The concentration of a biocide has been deemed to be the most important factor that affects its efficacy (Russell and McDonnell 2000). In the case of bacterial biofilms, the biocide concentration and consequently the bacterial susceptibility, is strongly affected by the reduced diffusion of active molecules through the biofilm (Anderson and O'Toole 2008, Lewis 2008, Maillard 2007, Tart and Wozniak 2008). Concentration is also central to the definition of bacterial resistance in practice. Therefore, the measurement of bacterial lethality rather than the measurement of bacterial growth inhibition is paramount.

Many reports on emerging bacterial resistance to biocides are based on the determination of minimum inhibitory concentrations (MICs). Using MICs to measure bacterial resistance is arguable since much higher concentrations of biocides are used in practice and, therefore, failing to achieve a reduction of bacterial numbers (i.e. lethality) because of elevated MICs is unlikely (Russell and McDonnell 2000). Indeed, some studies have shown that bacterial strains showing a significant increase in MICs to some biocides were nevertheless susceptible to higher (in use) concentrations of the same biocides (Lear et al. 2006, Thomas et al. 2005).

Thus, the determination of minimum bactericidal concentrations (MBCs) is a more appropriate methodology that allows the comparison of lethality between a standard and the resistant strains. Here the standard strains represent the population of bacteria which is normally susceptible to the biocide.

Likewise, the determination of the lethality of the in-use concentration of a biocide will indicate whether a bacterial strains is insusceptible (i.e. naturally resistant) or resistant (by comparison to a standard strain). The determination of the inactivation kinetic following exposure to a biocide, and in particular the shape of the inactivation curve, will provide information as to the nature of resistance of a population of cells and/or the interaction of the biocide with the cell population.

The determination of the lethality of a biocide must involve the use of a neutralising agent or the removal of the biocide. Failure to do so will provide an over-estimation of the lethality of the biocide.

MIC determinations have been used in many studies as an indicator of bacterial sensitivity change to a biocide (Russell and Mcdonnell 2000, Walsh et al. 2003). Bacteria showing an increased low-level of resistance/tolerance to a biocide might be selected by a low concentration of a biocide. Their level of resistance can increase through selection, for example by repeated exposure to a low concentration of a biocide or to increasing concentrations of a biocide (Abdel Malek et al. 2002, Langsrud et al. 2003, Maillard 2007, Tattawasart et al. 1999, Thomas et al. 2000, Walsh et al. 2003).

The determination of bacterial growth kinetics in the presence of a low concentration of a biocide can also provide indications of a change in bacterial phenotype (Gomez-Escalada et al. 2005a, Maillard 2007, Thomas et al. 2005).

Table 7  highlights the methodologies that have been used to measure bacterial resistance to biocides

Table 7: Methodologies to measure bacterial resistance 

Source & ©: SCENHIR,  Assessment of the Antibiotic Resistance Effects of Biocides (2009),
Section Resistance to biocides used in consumer product, p. 29-31


3.2 Has resistance to biocides been observed in health care applications?

The SCENIHR opinion states:

3.4.4. Resistance to biocides in specific applications Resistance to biocides used in health care

As early as 1966, bacterial resistance in Gram-negative bacilli to silver used in compresses in burn wounds was reported (Bridges et al. 1977, Cason et al. 1966, Klasen 2000, Moyer et al. 1965). In 1968, complications associated with silver nitrate compresses led to the use of silver sulphadiazine (silver combined with a sulphonamide) (Klasen 2000). In the 1970s, there were several reports of outbreaks of burn wound infection or colonisation by Gram-negative isolates resistant to silver sulphadiazine (Enterobacter cloacae) (Gayle et al. 1978), Providencia stuartii (Wenzel et al. 1976), Pseudomonas aeruginosa (Klasen 2000) and to silver nitrate (Pseudomonas aeruginosa) (Bridges et al. 1979), Salmonella Typhimurium) (McHugh et al. 1975). However, Percival et al. (2005) questioned the possibility of increasing silver resistance linked to an increase in antibiotic resistance in wound care. The induction of bacterial resistance has been decribed in almost all biocides, particularly in the less reactive ones such as quaternary ammonium compounds, bisbiguanides and phenolics, but also in the more reactive ones such as glutaraldehyde.

However, unlike antibiotic resistance, the issues relating to biocide resistance are considered to have a very low profile and priority (Cookson 2005). Despite the widespread use of disinfectant and antiseptic in health care settings, acquired resistance to current disinfectants in bacteria isolated from clinical specimens or the environment has rarely been well characterised. Emerging bacterial resistance to biocides has been well decribed in vitro; but evidence in practice is lacking (Cookson 2005, Maillard 2007, Russell 2002a, Weber and Rutala 2006).

Isolates with reduced susceptibility remain susceptible to clinically used concentrations of the disinfectants (Lear et al. 2006); the concentrations of disinfectants and antiseptic used in practice are substantially higher than the MICs of strains with reduced susceptibility (Weber and Rutala 2006). This finding is in constrast with antibiotic resistance, which has emerged over time, rendering a number of antibiotics clinically unusable.

However, after initial findings that the use of mupirocin resulted in a decolonization of patients carrying methicillin-resistant Staphylococcus aureus (MRSA), further studies were performed. Not only did they describe the appearance of mupirocin resistance of certain MRSA strains but also showed that MRSA strains carried a quaternary ammonium resistance gene (qacA) located in a gentamicin resistance plasmid that encoded for an efflux mechanism resulting in low-level chlorhexidine resistance (Cookson et al. 1991a). Moreover, transferable triclosan resistance in MRSA has been described, occurring together with a high-level of mupirocin in a hospital environment (Cookson et al. 1991b).

These few data indicate a need for further investigations on the long-term use of biocides in hospital environments and the relation to resistance against antimicrobial agents (Cookson et al. 2005). Stickler and Jones (2008) described the possibility of emerging triclosan resistance in Proteus mirabilis and suggested that urinary flora of catheterized patients should be monitored for Proteus mirabilis strains with reduced susceptibility to triclosan in any clinical trial or subsequent clinical use of triclosan for the prevention of urinary catheter encrustation and blockage

Source & ©: SCENHIR,  Assessment of the Antibiotic Resistance Effects of Biocides (2009),
Section Resistance to biocides used in consumer product, p. 34-35


3.3 Has resistance to biocides been observed in consumer products?

The SCENIHR opinion states: Resistance to biocides used in consumer products

Flores et al. (1997) isolated several bacterial strains resistant to a number of commonly used preservatives/biocides in cosmetic products. The bacterial strains isolated from the contaminated cosmetic products and their resistance to specific biocides are described in Table 9 . It was also demonstrated that safe preservation of cosmetic products requires a mixture of biocides. The effect of these resistant bacteria has been only investigated for the deterioration of the cosmetics, but not for pathogenicity.

In another study, biocide resistant strains of Enterobacter gergoviae (Davin-Regli et al. 2006) and Pseudomonas aeruginosa were isolated from contaminated cosmetics and from the floors of the washing area of industrial plants for the manufacture of cosmetics (Ferrarese et al. 2003). It appeared that the extensive use of some biocides for preservation (parabens; formaldehyde; formaldehyde releasers, imidazolidinyl urea and 1,3-Dimethylol-5,5-dimethyl (DMDM) hydantoin; and phenoxy ethanol) had lead to the development of the resistant strains. These strains were responsible for the deterioration of the cosmetics. Pseudomonas aeruginosa is also isolated from different aqueous solutions including cosmetics, disinfectants, ointments, soaps, vaginal irrigations, eye drops and dialysis equipment and fluids (Morrison and Wenzel 1984, Na’was and Alkofahi 1994). As a result of the development of bacteria resistant to specific biocides, a mixture of biocides is commonly used for the safe preservation of cosmetics. This means that the consumer is exposed to more biocides, both qualitatively and quantitatively. It was shown that Pseudonomas aeruginosa isolated from cosmetics and several other types of products is pathogenic and resistant to several types of antibiotics (Scully et al. 1986).

On the other hand, Cole et al. (2003) claimed after a study on 1238 isolates collected from the homes of antibacterial product users and non-users, that the results showed a lack of cross-resistance to antibiotic and antibacterial agents in target bacteria, as well as increased prevalence of potential pathogens in the homes of non-users. It should be noted that in this study, the isolates were selected based on their antibiotic resistance and then tested for their biocide insusceptibility. With our current state of knowledge, it is generally accepted that antibiotic resistance in clinical isolates is not necessarily associated with resistance to biocides.

Meanwhile the large use of triclosan in many home and personal-care products including deodorants, soaps, oral rinces, toothpaste and cutting boards may be associated with the decreased susceptibility to triclosan in clinical specimens of S. aureus (Bamber and Neal 1999, Suller and Russel 2000). Investigators have also reported increased tolerance to triclosan due to mutations in efflux pumps of E. coli and P. aeruginosa, or in M. smegmatis (see review of Weber and Rutala 2006). Bacillus, Micrococcus and Staphylococcus were able to contaminate cosmetics protected with preservatives like parabens and phenoxy-ethanol (Flores and al. 1997).

In the laboratory, it has been possible to develop bacterial mutants with reduced susceptibility to disinfectants that also demonstrate decreased susceptibility to antibiotics. Similarly, wild-type strains with reduced susceptibility to disinfectants (principally quaternary ammonium compounds and triclosan) have been reported.

Therefore, there is accumulating evidence that biocide resistant bacteria can be found in consumer products, but to date there are no studies to indicate that they are linked to antibiotic resistance and/or the emergence of pathogenic microorganisms

Table 9: Bacteria isolated from contaminated cosmetic products and their resistance to biocides 

Source & ©: SCENHIR,  Assessment of the Antibiotic Resistance Effects of Biocides (2009),
Section Resistance to biocides used in consumer product, p. 35-36


3.4 Has resistance to biocides been observed in the food production chain?

The SCENIHR opinion states: Resistance to biocides used in food production

Despite the widespread use of biocides in food production, data on resistance to biocides in microorganisms isolated in the plant or in the finished product are scarce. Meanwhile, there is some evidence of acquisition of a tolerance (if not resistance) for food-borne pathogens. Mokgatla et al. (1998, 2002) described a Salmonella strain growing in the presence of 28 mg/L- HOCl that was protecting itself by decreasing the level of species that could react with HOCl to generate toxic reactive oxygen radicals and by improving DNA damage repair mechanisms. These results are in agreement with the data of Aarestrup and Hasman (2004) who found that the use of chlorine might select resistant Salmonella bacteria.

Potenski et al. (2003) described mutants of Salmonella enteritidis selected following exposure to chlorine or sodium nitrite, sodium benzoate or acetic acid showing resistance to multiple antibiotics (tetracycline, chloramphenicol, nalidixic acid, ciprofloxacine), suggesting the mar operon mutation was responsible for resistance.

A recent study carried out by by Capita (2007) demonstrated that the use of acidified sodium chlorite may induce the selection in different serotypes of Salmonella a resistance against this biocide and a cross resistance to various antibiotics. This is also in accordance with the study of Oren-Gradel (2005) on the possible association between Salmonella persistence in poultry houses and resistance to commonly used disinfectants and a mutative role of the mar operon. Resistance to biocides used in animal husbandry

Given the increasing use of biocides in animal facilities, there are more and more concerns that they may select for resistant pathogens. However, while numerous investigations addressed the appearance of antimicrobial resistance following the use of antibiotics in farm animals (EFSA 2007), data relating the occurrence of resistance to the use of disinfectants are limited.

Gradel et al. (2005) tested MIC values against five commercial disinfectants (formaldehyde, glutaraldehyde/benzalkonium chloride, an oxidizing compound (non specified), tar oil phenol, and an iodophor) commonly used in poultry premises in Denmark on nine different serotypes of Salmonella isolated from different poultry farms. No significant differences could be established between MICs from flocks using or not using a certain disinfectant. Adaptation and de-adaptation studies revealed mutants highly resistant to triclosan (mar-type resistance) but comparable results were not obtained for the five used disinfectants. The authors concluded that even the adaptation and de-adaptation experiments could not demonstrate altered MICs to the five disinfectants regularly used on poultry farms.

Comparable investigations were conducted by Randall et al. (2007). They studied particularly the susceptibility of Salmonella enteritica var Typhimurium isolates comprising wild-type and laboratory mutants that were exposed to a tar oil phenol, an oxidising compound or a dairy steriliser disinfectant (quaternary ammonium biocide). They could show that exposure to these disinfectants could induce the expression of AcrAB and TolC efflux pumps, but that a single exposure was insufficient to select for mar-strains, associated with a reduced susceptibility to antibiotic such as ß-lactams, chloramphenicol, fluoroquinolones and tetracyclines, and increased tolerance to organic solvents and decreased susceptibility to disinfectants such as pine oil (Baucheron et al. 2005, Randall and Woodward 2002).

Earlier studies of Oethinger et al. (1998) had shown an association between cyclohexane tolerance and fluoroquinolones resistance in clinical isolates of E.coli. An association between cyclohexane resistance in Salmonella of different serovars isolated from animal facilities (as well as from human hospitals) and an increased resistance to multiple antibiotics, disinfectants (ethidium bromide, cetrimide, cyclohexane, triclosan) and dyes (acridine orange) was also described by Randall et al. (2001). Ninety-five percent of the cyclohexane-resistant strains isolated originated from poultry, but originated from only one turkey-rearing company, and hence might not be representative. The cyclohexane-resistant strains were also significantly more resistant to triclosan and cetrimide than the cyclohexane-susceptible strains. An overall finding was that that the resistance to antibiotics and disinfectants is consistent with the over-expression of AcrAB, as described by other authors for E. coli (Ma et al. 1996, Moken et al. 1997).

Recent investigations from Thailand (Chuanchuen et al. 2008) showed a high prevalence of antibiotic resistance in Salmonella enterica isolated from poultry and swine, but only very few variations of MICs to all disinfectants tested. Only 1.9% of the isolates were tolerant to cyclohexane. A recent study investigating the effect of cleaning and disinfection procedures in poultry slaughterhouses on the development of, or selection for biocide and antibiotic resistance in Campylobacter jejuni and C. coli showed that a very low number (1-2) of genotypes were recovered after cleaning and disinfection and that there was no increase in antibiotic resistance before and after exposure to the disinfection procedures (Peyrat et al. 2008). In two recent studies, Salmonella exposed to a range of common farm disinfectants was found to develop a low, but statistically significant, increased risk of selection of mutants with reduced susceptibility to ampicillin, ciprofloxacin and tetracycline. Some of the mutants selected were of the MDR phenotype (Karatzas et al. 2007; Karatzas et al. 2008).

In conclusion, there is understandable concern that the improper use of biocides in primary animal production could select for antibiotic-resistant bacteria. Indeed, laboratory-based studies have shown that this can occur, particularly when exposure to sub-optimal biocide concentrations is either prolonged or repeated. However, so far, these observations are not largely supported by field studies. There is a need to establish whether current cleaning and disinfection regimes in use in food animal production in the EU represent a real hazard with respect to the selection of antibiotic-resistant human and/or animal pathogens. Resistance to used in foods of animal origin

As mentioned above (see section 3.3.5), biocides may be used (and are already used in many third countries) for the disinfection and decontamination of foods of animal origin. There is a need to generate more data on the occurrence of biocidal-resistant bacteria on carcass surfaces and on foods of animal origin

Source & ©: SCENHIR,  Assessment of the Antibiotic Resistance Effects of Biocides (2009),
Sections Resistance to biocides used in food production & Resistance to biocides used in animal husbandry, p. 37-38


3.5 Has resistance to biocides due to discharges to the environment been observed?

The SCENIHR opinion states: Resistance to biocides that occur in the environment

Laboratory experiments have demonstrated that biocides, present at low concentrations in the environment after use and discharge, may lead to an increased selective pressure towards disinfectant and antibiotic resistance. Thus, the study of Randall et al. (2004) performed with triclosan and a phenolic farm disinfectant illustrated that Salmonella enterica was able to tolerate relatively high concentrations of disinfectants and to develop cross-resistance to certain antibiotics.

The study from McBain et al. (2003b) on the microbial population dynamics and antimicrobial susceptibility during exposure of sink drains microcosms to triclosan, clearly demonstrated that triclosan exposure did not significantly lower total counts of drain biofilm bacteria but dynamically altered the bacterial composition. This change in population was caused by innate resistance or insusceptibility of some species able to degrade triclosan. Most importantly, the authors noted that the antibiotic susceptibility profile was not affected.

Lear et al. (2002) isolated many intrinsically resistant bacteria from factory settings where triclosan and chloroxylenol were produced. A small number of non Pseudomonads isolates (Acinetobacter and Citrobacter) from the same samples demonstrated an increased insusceptibility to triclosan but remained susceptible to its in-use concentration. However, these environmental bacterial isolates exposed to the biocide showed resistance to some unrelated antibiotics (Lear et al. 2006).

A number of papers have investigated antibiotic resistant bacterial strains in hospital wastewater (Baquero et al. 2008, Kümmerer 2004), where high concentrations of antibiotics and disinfectants are found. However, to date, no study seems to have focused on the emergence of biocide resistant bacteria in hospital environments other than wastewater

Source & ©: SCENHIR,  Assessment of the Antibiotic Resistance Effects of Biocides (2009),
Section Resistance to biocides that occur in the environment, p. 38-39

The Three-Level Structure used to communicate this SCENIHR Opinion is copyrighted by GreenFacts asbl/vzw.