Integrating animal health surveillance and food safety: the issue of antimicrobial resistance.

Surveillance of antimicrobial resistance in commensal, zoonotic and pathogenic bacteria from humans, animals and food is an essential source of information when formulating measures to improve food safety. International organisations (the World Health Organization, the World Organisation for Animal Health, the Food and Agriculture Organization of the United Nations, and the Codex Alimentarius Commission) have developed a complete set of standards related to resistance surveillance programmes and are calling for the establishment of integrated surveillance programmes. The most important task in establishing an integrated surveillance programme for antimicrobial resistance should be the harmonisation of laboratory testing methodology and antimicrobial-use reporting. Overthe last decade, the integration of surveillance of antimicrobial resistance has been an important step toward addressing the global concern with antimicrobial resistance. However, very few systems are in place and there is still a lot to do before harmonised surveillance systems become the norm.

Antimicrobial resistance: a complex issue.

The discovery of antibiotics represented a turning point in human history. However, by the late 1950s infections that were difficult to treat, involving resistant bacteria, were being reported. Nowadays, multiresistant strains have become a major concern for public and animal health. Antimicrobial resistance is a complex issue, linked to the ability of bacteria to adapt quickly to their environment. Antibiotics, and antimicrobial-resistant bacteria and determinants, existed before the discovery and use of antibiotics by humans. Resistance to antimicrobial agents is a tool that allows bacteria to survive in the environment, and to develop. Resistance genes can be transferred between bacteria by horizontal transfer involving three mechanisms: conjugation, transduction and transformation. Resistant bacteria can emerge in any location when the appropriate conditions develop. Antibiotics represent a powerful selector for antimicrobial resistance in bacteria. Reducing the use of antimicrobial drugs is one way to control antimicrobial resistance; however, a full set of measures needs to be implemented to achieve this aim.

Antimicrobial resistance in animal and public health: introduction and classification of antimicrobial agents.

Bacteria have a remarkable ability to adapt, evolve and survive by developing resistance to therapeutic compounds. This ability is also shared by other pathogenic agents such as viruses, fungi, and parasites. Even when focusing on bacterial resistance only, this phenomenon is quite complex to analyse due to the diversity of animal species, the diversity of rearing environment, the number of antimicrobial classes available and the diversity of pathogenic bacteria involved. This introductory paper includes developments on the place of antiviral compounds in veterinary medicine and a classification of antimicrobials used in food-producing animals.

Antagonism of the antibacterial action of some penicillins by other penicillins and cephalosporins.

There are many examples of two penicillins acting synergistically, usually by one competitively inhibiting beta-lactamase, thus protecting the other from inactivation. There are few reports on penicillins antagonizing each other. Eight strains of three genera (Proteus, Escherichia, Pseudomonas) isolated at Boston City Hospital or Institut Pasteur, Paris, showed antagonism of carbenicillin or ampicillin by cephaloridine, cloxacillin, or 6-aminopenicillanic acid. Broth dilution tests showed that with seven of the eight strains the minimum inhibiting concentration (MIC) of the more active antibiotic was increased by 800-6,400% by low concentrations (often one-tenth the MIC) of the antagonist, whereas higher concentrations of "antagonist" were not as antagonistic, This suggested that two or more receptor sites of action for penicillins were present; the antagonist thus blocks the antibacterial action at the more sensitive site but acts additively with the antagonized antibiotic at the less sensitive site. The possibility that the mechanism of antagonism was induction of inactivating enzymes (beta-lactamase, penicillin acylase) was studied in two strains(one Escherichia coli and one Proteus rettgeri), and two antagonists were studied in detail. With E. coli cephaloridine was a poorer inducer of beta-lactamase than were the antagonized antibiotic and 6-aminopenicillanic acid. From these organisms, the good inducers of a beta-lactamase that acted on benzylpenicillin did not induce enzymes that inactivated carbenicillin. Thus, the mechanism of antagonism was not due to beta-lactamase induction.

Antimicrobial resistance at farm level.

Bacteria that are resistant to antimicrobials are widespread. This article reviews the distribution of resistant bacteria in farm environments. Humans, animals, and environmental sites are all reservoirs of bacterial communities that contain some bacteria that are susceptible to antimicrobials and others that are resistant. Farm ecosystems provide an environment in which resistant bacteria and genes can emerge, amplify and spread. Dissemination occurs via the food chain and via several other pathways. Ecological, epidemiological, molecular and mathematical approaches are being used to study the origin and expansion of the resistance problem and its relationship to antibiotic usage. The prudent and responsible use of antibiotics is an essential part of an ethical approach to improving animal health and food safety. The responsible use of antibiotics during research is vital, but to fully contribute to the containment of antimicrobial resistance 'prudent use' must also be part of good management practices at all levels of farm life.

Strategies in aminoglycoside use and impact upon resistance.

Assessment of amikacin resistance over a 10-year period at our institution revealed that the number of resistant strains remained stable. Qualitatively, amikacin-resistant Enterobacteriaceae and Pseudomonas were fairly stable. There was a slight increase in amikacin-resistant Acinetobacter and staphylococci. Different factors influencing the emergence and spread of resistant hospital bacteria have been studied at different periods and compared with similar data on gentamicin-resistant strains. Transmissible plasmids, multiple mechanisms of resistance, and high levels of resistance were more frequent in gentamicin-resistant strains. In the amikacin-resistant strains, the level of resistance was 16 to 32 mg/liter with few autotransferable plasmids. A synergistic or additive effect with cephalosporins, which may be a factor in decreasing the risk of selection of the resistant strains since there is no plasmid-mediated resistance to cephalosporins, was demonstrated in Enterobacteriaceae. To control the development of aminoglycoside resistance in hospitals, it may be necessary to restrict the use of more than the one drug to which resistance is developing; to use the antibiotic at the right dosage and, when necessary, in a combination that may prevent the emergence of resistant organisms and plasmids; and to develop measures to control bacterial and R factor transmission.

Synergism and antagonism in double beta-lactam antibiotic combinations.

Combinations of two beta-lactam antibiotics may be advantageous in certain clinical situations, providing a synergistic activity against specific organisms or a broad spectrum of antibiotic coverage. Depending on the combination and the bacteria, synergism, indifference, or antagonism can be observed. Synergism may occur when two beta-lactam antibiotics, acting on different penicillin-binding proteins, are combined or when a penicillinase-susceptible beta-lactam antibiotic is protected by another beta-lactam antibiotic acting as a beta-lactamase inhibitor in strains producing a penicillinase (chromosomal or plasmidic). With different species, such as Enterobacter, Citrobacter, indole-positive Proteus, Serratia, Aeromonas, and Pseudomonas, which produce an inducible chromosome-encoded cephalosporinase, antagonism will appear if one of the two combined antibiotics causes induction of the beta-lactamase and the other becomes inactivated by the increased amount of the enzyme. Although most combinations of new beta-lactam antibiotics (ureido-penicillins, third-generation cephalosporins, monobactams) appear to be indifferent, antagonism and possible selection of resistant mutants are the drawbacks of such combinations. Nevertheless, highly active compounds, if used at doses above the minimal inhibitory concentrations, especially in the case of potential cephalosporinase-inducers, may be safe in vivo as far as avoiding antagonism is concerned, but not necessarily with respect to the selection of resistant mutants.

Epidemiology of quinolone resistance: Europe and North and South America.

After nearly 10 years of fluoroquinolone usage for a wide range of bacterial infections, a striking difference has been observed in the incidence of bacterial resistance to fluoroquinolones between bacteria responsible for community- and hospital-acquired infections, respectively. Resistance is only rarely encountered among common pathogens. In most studies, 97 to 100% of all pathogens are fully susceptible to fluoroquinolones. In contrast, resistance to fluoroquinolones has emerged and increased among bacteria responsible for nosocomial infections. The incidence of resistance to fluoroquinolones varies between bacterial species, clinical settings and countries, and is related to local epidemic spread of a few clones. The highest incidence of resistance is observed in Pseudomonas aeruginosa, Acinetobacter spp., Serratia marcescens and, particularly, methicillin-resistant Staphylococcus aureus (MRSA): some investigators have reported 95 to 100% fluoroquinolone resistance among MRSA. Follow-up of trends in the resistance to fluoroquinolones based upon surveillance programmes are needed.

Beta-lactamases in clinical isolates. Spectrum implications of sulbactam/ampicillin.

Combining ampicillin with an irreversible beta-lactamase inhibitor such as sulbactam is a promising technique for controlling infections due to resistant organisms. The combination not only restores the effectiveness of ampicillin against species that have acquired resistance to it, but it can extend the antimicrobial spectrum to species that have never exhibited susceptibility to ampicillin. However, the inhibition of beta-lactamases and the evaluation of inhibitors are still complicated by numerous factors known collectively as 'the epidemiology of beta-lactamases'. This refers to the distribution of enzymes in different bacterial species and in different geographical locations, the numerous types of enzymes, the variable number and amount of beta-lactamases that may exist in the same cell, and, of course, the potential for transfer of beta-lactamase resistance from one species to another.

The epidemiology of bacterial resistance to quinolones.

The new fluoroquinolones have been in use for nearly 10 years in the treatment of community- and nosocomially-acquired infections. Resistant clones may be selected during therapy and disseminate if favourable epidemiological conditions prevail. Resistance to the fluoroquinolones is still rare in common pathogens with 97 to 100% of strains remaining susceptible. Resistance has been reported in methicillin-susceptible Staphylococcus aureus, Campylobacter jejuni/coli, Salmonella, Shigella and Escherichia coli. Among nosocomial pathogens, the incidence of fluoroquinolone resistance varies between bacterial species, countries and periods of study, and is dependent on local epidemiological factors and antibiotic policies. The highest incidence of resistance is observed in Serratia and Acinetobacter spp., and particularly in methicillin-resistant S. aureus. Surveillance programmes are needed to follow up trends in resistance to the fluoroquinolones and their possible association with clinical failures.

[beta-Lactamases of Branhamella catarrhalis and their phenotypic implications]. / beta-Lactamases de Branhamella catarrhalis et leurs implications phénotypiques.

Of the 50 strains of beta-lactamase-producing Branhamella catarrhalis isolated at Saint Joseph's Hospital (Paris) that were studied, 94% produced BRO-1 type beta-lactamase and 6% produced the BRO-2 type. We examined the transfer of BRO-1 and BRO-2 genes and found that, among 7 donor strains producing BRO-1, all were able to transfer the gene for BRO-1 production by conjugation. Of the 4 donor strains producing BRO-2, 2 were able to transfer the gene for BRO-2 production by conjugation. Three BRO-1 beta-lactamase-producing transformants were obtained from total DNA extracted from 3 strains producing BRO-1. Plasmid bands were demonstrated in strains of B. catarrhalis, but no change in plasmid profiles was seen in beta-lactamase-positive recombinants, supporting previous studies that suggested the beta-lactamases are chromosomal. In vitro activity of oral beta-lactams was tested for 67 strains of B. catarrhalis (56 beta-lactamase-producing strains). Cefixime, cefpodoxime and the combination ampicillin-clavulanic acid were very active against the beta-lactamase-producing strains. BRO-1 beta-lactamase appears to affect the activity of cefaclor, cefuroxime and loracarbef. BRO-2 beta-lactamases have no effect on the activity of these cephalosporins. Cefixime and cefpodoxime seemed the least affected by beta-lactamase production.

Epidemiology of antibiotic resistance in Haemophilus influenzae.

During the last 10 years, a continuous increase in the incidence of beta-lactamase-producing Haemophilus influenzae has been observed; in addition, beta-lactamase-negative ampicillin-resistant strains are better identified and more frequently isolated. During the same period, resistance to tetracyclines and chloramphenicol decreased to a very low level. The incidence of resistant bacteria is highly variable according to the site of infection, patient's age, type of Haemophilus, and country or region, according to local epidemiological factors. Follow-up multicenter studies are needed to monitor the evolution of resistance to these antibiotics and also emergence and spread of resistance to other antibiotics, such as new fluoroquinolones, new beta-lactams, and new macrolides.

A seven-year survey of susceptibility to marbofloxacin of pathogenic strains isolated from pets.

This study, Vetoquinol S.A. epidemiosurveillance, was conducted from 1994 to 2001 in order to determine the susceptibility (by MIC determination) to marbofloxacin (a third generation fluoroquinolone used only in individual administration for animals). Strains from infected pets originated from six European countries. Isolates were collected from urinary infections (Escherichia coli), respiratory infections (Pasteurella multocida), dermatological infections (Staphylococcus intermedius, Pseudomonas aeruginosa) and otitis (S. intermedius, P. aeruginosa). The MIC distribution for each species was the same both before and after the launch of marbofloxacin in 1995. In E. coli, a resistant population was present before the use of marbofloxacin; this resistance was induced by co- or cross-resistance to other antibiotics used previously. Over this period, there was no significant evolution of MIC(90) for any bacterial species studied and no development of resistance was observed. Marbofloxacin was the most active antibiotic against P. multocida isolates and had the lowest MIC. No difference in MIC distribution was seen between the S. intermedius (unimodal distribution) isolated from dermatological infections and those from otitis. This was also true for P. aeruginosa. The use of marbofloxacin was not found to have induced a significant increase or spread of resistant bacteria.

Seven years survey of susceptibility to marbofloxacin of bovine pathogenic strains from eight European countries.

This study was conducted from 1994 to 2001 to determine the susceptibility of bovine pathogenic bacteria to marbofloxacin (a third generation fluoroquinolone used only in individual administration for animals). Strains originated in bovine diseases from eight European countries. They were isolated from gut infections (Escherichia coli, salmonellae), mastitis (E. coli, Staphylococcus aureus, Streptococcus uberis, S. agalactiae, S. dysgalactiae) and respiratory diseases (Mannheimia haemolytica, Pasteurella multocida, Haemophilus somnus). There was no change in the MIC distributions for each species after the launch of marbofloxacin in 1997. In E. coli, a resistant population was present before the use of marbofloxacin having been induced by co- or cross-resistance to other antibiotics used previously. Over this period the only a significant change seen was an increase in MIC(90) of E. coli from the gut (1.275 microg/ml in 1994/1995 to 5.098 microg/ml in 2001). All the salmonellae were susceptible to marbofloxacin with a MIC(90) = 0.073 microg/ml in 2001 without development of high level resistance. The use of marbofloxacin seems not to have favoured a significant increase and spreading of resistant bacteria.

Epidemiological study of Staphylococcus aureus resistance to new quinolones in a university hospital.

During a 14-month period, from December 1984 to February 1986, 630 Staphylococcus aureus isolates were identified at Broussais Hospital. Thirty-eight isolates (6%), from 35 patients, were found to be pefloxacin-resistant S. aureus (PRSA) with minimal inhibitory concentrations greater than or equal to 8 mg l-1. PRSA isolates were tested for susceptibility to 35 antibiotics, including nine quinolones, and heavy metal ions. Phage-type was determined. Out of the 38 PRSA isolates, 35 (92%) were methicillin- and multiply-resistant; however, all PRSA isolates were sensitive to vancomycin and coumermycin. Fifteen isolates (39%) had similar phage-type and identical antibiotic susceptibility pattern with high level resistance to pefloxacin (MICs equal to 64 mg l-1); they were isolated from the same surgical unit. The 23 remaining PRSA isolates differed by their phage and susceptibility patterns. Pefloxacin MICs ranged from 8 to 512 mg l-1 with a bimodal distribution; cross-resistance was observed with the eight other quinolones tested. Only nine PRSA isolates (24%), including four 'epidemic' isolates, were obtained from patients who had been treated with quinolones. From these data there is apparently no direct relationship between quinolone administration and selection of PRSA in infected patients.

Antibiotic synergy and antagonism.

The field of synergistic combinations of antibiotics is extremely broad and mostly has been explored in vitro. Some fixed combinations were successfully developed commercially. A few combinations were tested in animal models, and a smaller number was studied in human patients. Any practitioner in infectious diseases has some individual cases, published or unpublished, which add some evidence to the role of synergistic combinations in difficult therapy problems--either on the side of the patient when immunosuppressed or on the side of the bacterial strain, when multiply resistant. MRSA, VISA, E. faecium resistant to penicillin G and highly resistant to aminoglysocides and to vancomycin, P. aeruginosa resistant to ceftazidime and imipenem, and Acinetobacter baumani resistant to imipenem are some of the bacterial strains dangerous for the patient and the hospital, which trigger the imagination of the microbiologist and physician to find a satisfactory treatment. On the side of the drug industry, the increasing knowledge of resistance mechanisms and of synergistic mechanisms may open some new approach, such as efflux inhibitors, a membrane-active compound that can be combined with a partner antibiotic. Antagonism between antibiotics would be worthwhile to study because it likely contributes to the disadvantages of the inappropriate use of antimicrobial combinations.

Male fertility and positive Chlamydial serology. A study of 61 fertile and 82 subfertile men.

In this study, the frequency and titers of serum antibodies to Chlamydia trachomatis in fertile and subfertile subjects and possible correlations between a positive serology and the values of conventional semen parameters were investigated. The subfertile subjects were divided into those without and those with a history of genitourinary infection. No difference was noted between the proportion of seropositive patients in the fertile group (26.2%) as compared with those in the subfertile group with no history of genitourinary infection (27.7%). The percentage of positive tests was significantly increased only in the subfertile group with past genitourinary infection (53.5%). No difference was found between the antibody titers of the fertile and subfertile groups. Only the fertile group had any semen parameter, namely, the volume of the ejaculate, that was significantly different between the seropositive and seronegative patients. In conclusion, three main points should be stressed; the high frequency of serum antibodies to Chlamydia trachomatis in the control subjects and the apparent absence of effects of a positive serology on their fertility; the significant correlation between a past genitourinary infection and the presence of serum antibodies to Chlamydia trachomatis; and the need for a test to assess Chlamydia in semen.

Low rate of Clostridium difficile colonization in ambulatory and hospitalized HIV-infected patients in a hospital unit: a prospective survey.

OBJECTIVE: To determine the frequency of Clostridium difficile carriage in HIV-infected in- and out-patients, and to assess the role of this carriage in nosocomial transmission of C. difficile. PATIENTS AND METHODS: Prospective study in a university hospital. Forty-five consecutive HIV-infected out-patients and 120 hospitalized patients (52 HIV and 68 non HIV-infected-patients) were studied. During the period of hospitalization, 44 patients (24 HIV and 20 non-HIV-infected patients) with a negative culture within 48 h of admission were followed weekly for fecal carriage. Clostridium difficile culture and latex agglutination were performed on the fecal samples of each patient. In the case of positive culture and/or latex agglutination, C. difficile toxin assays were performed by microtitre cytotoxicity method. RESULTS: Out-patients: one patient was a carrier and one patient with diarrhoea was infected with a toxigenic strain (2/45, 4.5%, 95% CI = 1-17). Eighty percent of the HIV-infected out-patients had received antimicrobial agents previously. In-patients: in the first 48 h, five asymptomatic patients were carriers (three non-HIV and two HIV-infected patients). Among 20 patients who complained of diarrhoea, one HIV-infected patient had only a positive latex agglutination and one HIV-infected patient was infected with a toxigenic strain. Overall, 7/120 (5.8%, 95% CI = 2-10) patients were infected or colonized with C. difficile. During the hospitalization (743 patient-days), none of the 44 patients acquired C. difficile. CONCLUSION: This study suggests that in this given unit, C. difficile carriage is low, at least with single room accommodation, and in the absence of clusters of cases. This carriage is not different in HIV and non-HIV infected patients despite treatment with multiple antibiotics, and is not different in patients managed in different care environments. The systematic identification of C. difficile carriers for isolation and prophylactic treatment is not useful under these circumstances.

Resistance patterns of Haemophilus influenzae.

Haemophilus influenzae is an important respiratory tract pathogen, and the prevalence of strains resistant to the antibiotics used to treat it has increased since the 1970s. Data on H. influenzae have been gathered in the Alexander Project, an ongoing, international surveillance study. They reveal that beta-lactamase production is the primary mechanism for H. influenzae resistance and that the resistance has a wide geographical variation, reaching critical levels in some countries. Unlike many techniques used in the past, Alexander Project methods provide accurate, reproducible susceptibility data, allowing the comparison of resistance prevalence over time and between areas. Traditional antimicrobial breakpoints are being superseded by more accurate and clinically relevant methods of predicting drug efficacy, such as time above the MIC, AUC:MIC ratios and pharmacodynamic breakpoints. Continued surveillance for resistance and susceptibility testing of H. influenzae is vital to maximize the benefits of antimicrobial therapy and to contain the spread of infection.

Epidemiology of resistance to diaminopyrimidines.

Resistance to trimethoprim emerged in Enterobacteriaceae and later in other Gram-negative and Gram-positive bacteria within two years of the clinical introduction of the drug. Resistance is borne in many different replicons often present in multiply-resistant epidemic bacteria. The incidence of trimethoprim resistance is highly variable, depending upon methodology, type of patients, local epidemiology: this can be illustrated by the high variation of trimethoprim resistance among Salmonella, Shigella or MRSA in various countries and by the incidence of resistance in penicillin-resistant Streptococcus pneumoniae.