ABSTRACT
We examined factors associated with penicillinase production by nasal carriage Staphylococcus aureus strains in 648 children aged 3 to 6 years attending 20 randomly sampled playschools. The children were prospectively monitored for drug use and medical events for 6 months and were then screened for S. aureus carriage. Isolates were tested for their susceptibility to penicillin G and methicillin, and penicillinase production by methicillin-susceptible, penicillin-resistant strains was quantified. S. aureus was isolated from 166 children (25.6%). Exposure to amoxicillin-clavulanate during the previous 3 months was associated with higher penicillinase production by penicillin-resistant, methicillin-susceptible strains (odds ratio, 3.6; P = 0.03). These results suggest that use of the amoxicillin-clavulanate combination could induce a herd selection process of S. aureus strains producing higher levels of penicillinase.
Staphylococcus aureus is one of the most frequent bacterial pathogens of humans. It causes skin infections, osteoarthritis, and respiratory tract infections in the community, as well as postoperative and catheter-related infections in hospitals. These infections can lead to life-threatening bacteremia and septic metastases. Shortly after the advent of penicillin G, a penicillin-resistant S. aureus strain was found to produce a penicillinase that inactivated the antibiotic (14, 29). Penicillinase confers resistance to all penicillins except for penicillin M (3, 27) and is inactivated by beta-lactamase inhibitors (16, 23). Penicillinase-producing S. aureus strains spread rapidly in hospitals and, several years later, within the community (22). A few years after the availability of methicillin, methicillin-resistant S. aureus (MRSA) strains were described. Unlike penicillinase, which is plasmid encoded, methicillin resistance is mediated by penicillin-binding protein 2a (PBP2a) (6), an enzyme with very low affinity for beta-lactams that is encoded by the chromosomal mecA gene. The frequency of MRSA strains has increased gradually in hospitals over the past 2 decades, and they now account for more than 50% of nosocomial isolates.
Besides PBP2a (the most frequent mechanism of methicillin resistance), several authors have reported the isolation of borderline oxacillin-susceptible S. aureus strains in the community (10, 13, 20, 24, 30). These strains are characterized by oxacillin MICs close to the breakpoint distinguishing between methicillin-susceptible and methicillin-resistant strains, whereas the oxacillin MICs for most MRSA strains are high (18). The main mechanism believed to account for this phenotype is beta-lactamase hyperproduction (2, 18). Although borderline oxacillin-susceptible S. aureus strains were first described in 1986 (18), risk factors for colonization by high-level beta-lactamase-producing S. aureus strains in the community, and particularly prior antibiotic exposure, remain to be identified.
The aim of this study was to determine whether the level of penicillinase production by nasal carriage methicillin-susceptible S. aureus strains in healthy children is associated with prior antibiotic exposure.
(These results were presented in part at the 40th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Ontario, Canada, September 2000.)
MATERIALS AND METHODS
Survey design.In December 1995, we randomly selected 20 French playschools in an administrative department of central France with 600,000 inhabitants, where more than 99.5% of 3- to 6-year-old children attend playschools. This department was chosen as typical of French departments in terms of sociodemographic characteristics, medical activity of practitioners not affiliated with hospitals, and social security coverage. The 3- to 6-year age group was chosen to enhance the feasibility of recruiting a random population-based sample; their inclusion was based on registration at the beginning of the school year. The parents gave informed consent to the study, and the children gave oral consent to bacterial screening.
At the beginning of the study, the parents were given a questionnaire on which to record sociodemographic characteristics, together with their child's drug consumption (trade name, unit dose, daily frequency, and duration of treatment). The study was conducted over a 6-month period, from December 1995 to May 1996, with the help of the teachers. Because the ecological niche of S. aureus is the anterior nares (15), the children were screened for anterior nasal carriage of S. aureus at the end of the 6-month follow-up period. Children whose parents could not remember treatment details were excluded.
Ethical considerations.This project was sponsored by the Institut National de la Santé et de la Recherche Médicale (INSERM). It was examined by the institutional review board of the Creteil teaching hospital and approved by the French Ministry of Health. It was also approved by the appropriate French computer watchdog committees (Comité Consultatif sur le Traitement de l'Information en Matière de Recherche dans le Domaine de la Santé and Commission Nationale de l'Informatique et des Libertés). The legally required scientific informed-consent forms were signed by the parents of enrolled children.
Sample collection and microbiological methods.After sampling, swabs were immediately placed in TVG-AER transport medium (Bio-Rad, Marnes-La-Coquette, France) and transported within 12 h to the central laboratory (BIO VSM Lab). On reception, the swabs were plated on selective mannitol salt agar (bioMérieux, Marcy-l'Etoile, France) and incubated for 24 and 48 h at 37°C. Mannitol-fermenting colonies were Gram stained and tested for catalase and free coagulase by using rehydrated rabbit plasma (Bio-Rad) according to the currently recommended procedure (11). S. aureus strain ATCC 25923 was used as a control for culture and identification procedures. Mannitol-fermenting staphylococci that tested positive for free coagulase were identified as S. aureus and were stored at −70°C in brain heart infusion (BHI) with 15% glycerol. They were then transported to the Microbiology Department of Robert-Debré Hospital for further studies.
Detection of penicillin and methicillin resistance and measurement of penicillinase production.Penicillin resistance was detected by the penicillin disk susceptibility test and the nitrocefin-based test (7). MICs of oxacillin were determined by the standard agar dilution method, as recommended by the French Microbiology Society (7).
All S. aureus isolates were screened for the mecA gene (encoding methicillin resistance) by PCR, as reported by Vannuffel et al. (31).
Penicillinase production was measured for penicillin-resistant, methicillin-susceptible S. aureus isolates. Strains were grown overnight in 10 ml of BHI medium, collected by centrifugation, and stored at −20°C. Cell pellets were resuspended in 0.5 ml of 50 mM phosphate buffer, pH 7.4, containing lysostaphin at a final concentration of 0.1 mg/ml. The suspensions were incubated at 37°C for 15 to 20 min and then sonicated for 1 min on ice. The disrupted cells were centrifuged at 17,000 × g rpm for 10 min at 4°C. Penicillinase activity in cleared extracts was determined by using nitrocefin as the substrate, and the protein concentration was determined by using the Bio-Rad Bradford dye-binding assay. For penicillinase assays, 20 to 50 μl of extract was added to a cuvette containing 0.1 mM nitrocefin in 50 mM phosphate buffer, pH 7.4, and hydrolysis was monitored at 482 nM for 5 min. Activity was linearly dependent on the amount of extract added, and specific penicillinase activity in the extracts was recorded as moles of nitrocefin hydrolyzed per minute per milligram of protein. All activity assays were run in duplicate.
PFGE.The 20 S. aureus isolates with the highest penicillinase production (>0.237 μmol/min/mg) were analyzed by pulsed-field gel electrophoresis (PFGE) of SmaI-digested total DNA, as described previously (9).
Descriptive and analytical methods.Only exposure to systemically administered antibiotics was taken into account. We first examined factors possibly associated with nasal carriage of S. aureus. The analysis of factors associated with penicillinase production was restricted to penicillin-resistant, methicillin-susceptible strains. After the homogeneity of variance was tested, a t test was used to compare penicillinase production according to the oxacillin MIC of the corresponding strain. Penicillinase values were dichotomized around the median for univariate and multivariate analyses.
The chi-square test was used to compare binary variables. Multivariate analyses used logistic regression models which were constructed with variables with P values of <0.2 in univariate analysis, followed by backward stepwise regression. P values of <0.05 were considered statistically significant. Stata SE 8 software was used.
RESULTS
Participation rate and population characteristics.Based on school registers, 832 children were eligible. Nonparticipation was due to lack of questionnaire completion (7.6%), refusal by parents (4.7%), refusal by children (3.9%), usual absence of children such as for an afternoon nap at home (2.3%), absence of children from the school because of parents' holidays (1.4%), and other reasons (1.4%). The parents of 648 children (77.9%) completed the questionnaire, and these children were sampled for S. aureus carriage. The mean age was 4.9 years (95% confidence interval [95% CI], 4.8 to 5.0), and the mean body weight was 18.3 kg (95% CI, 18.0 to 18.5].
Antibiotic exposure during the previous 3 months.At least one course of antibiotics was prescribed to 42.8% of the children during the 3 months before nasal sampling. All the antibiotics were given orally. The patterns of different antibiotic classes are presented in Table 1. Among the cephalosporins, narrow-spectrum drugs were most frequently used (9.3% of children); expanded- and broad-spectrum drugs accounted for 0.9 and 4.5% of prescriptions, respectively.
Description of the population and analysis of factors associated with S. aureus nasal carriage
Risk factors for S. aureus carriage.One hundred sixty-six children (25.6%) were colonized by S. aureus. The carriage rate was not influenced by gender or by amoxicillin or macrolide use. Age was the only factor associated with S. aureus colonization: the risk of colonization increased with age. Cephalosporin use during the previous 3 months was associated with a lower risk of S. aureus colonization in univariate analysis but not in multivariate analysis (Table 1).
Beta-lactam susceptibility.Seven strains (4.2%) were susceptible to penicillin G. One strain (0.6%) was resistant to methicillin and harbored the mecA gene. This strain was susceptible to erythromycin, clindamycin, trimethoprim-sulfamethoxazole, rifampin, and gentamicin. The carrier was a 4.7-year-old girl with no predisposing factors such as hospitalization during the previous 6 months or parents close to sick persons.
Level of penicillinase production and antibiotic use.Among penicillin-resistant, methicillin-susceptible strains (n = 158), the pattern of penicillinase production (Fig. 1) was clearly associated with the oxacillin MICs for the strains (P < 0.001) (Fig. 2). Upon analysis of penicillinase production dichotomized around the median, two factors were associated with increased penicillinase production: an age of ≥6 years (odds ratio [OR], 2.7; P = 0.02) and amoxicillin-clavulanate exposure (OR, 3.7; P = 0.03). Exposure to amoxicillin alone was not associated with the level of penicillinase production; neither was exposure to cephalosporins or macrolides (Table 2).
Pattern of penicillinase production by penicillin-resistant, methicillin-susceptible strains of S. aureus (n = 158). The median value was 0.05 μmol/min/mg.
Penicillinase production according to the oxacillin MIC (n = 158).
Analysis of factors associated with the level of penicillinase production by penicillin-resistant, methicillin-susceptible strains of S. aureusa
Total DNAs from the 20 strains with the highest levels of penicillinase production were analyzed by PFGE. Only two pairs of isolates (isolates 52 and 70 and isolates 463 and 481) were indistinguishable (Fig. 3).
PFGE patterns of the 20 S. aureus strains with the highest levels of penicillinase production. The identification numbers of the strains are shown above the gel.
DISCUSSION
The prevalence of S. aureus nasal carriage found in this prospective community-based study (25.6%) is in keeping with values reported elsewhere (25 to 40%) (8, 12, 20, 28). As reported for unselected populations in other countries, we found that multiresistant MRSA was still rare among nasally carried strains in the community children (1, 25, 26, 28).
Our most striking result is the relationship between amoxicillin-clavulanate exposure during the previous 3 months and the level of penicillinase production by nasal S. aureus isolates. No such relationship was found with single-agent amoxicillin therapy.
We did not cross-validate the parents' records of drug exposure with physicians' records, but such validation would likely have had a minimal impact on the validity of our results. Indeed, this was a prospective study, the questionnaires were distributed 6 months before bacteriological screening, and parents were unaware of whether their child was an S. aureus carrier. This method had the advantage of recording actual drugs administered, not only prescribed antibiotics.
The penicillinase production levels observed may have been biased by different nitrocefin hydrolysis rates of different type of penicillinases (34). Although we did not determine the types of penicillinases produced by our strains, the very strong correlation observed between the strains' intrinsic penicillinase activities and the oxacillin MICs supports the potential clinical relevance of our penicillinase data. We also observed a correlation between clavulanate exposure and oxacillin MICs (data not shown). Furthermore, PFGE analysis indicated that the isolates producing higher levels of penicillinase were not clonal.
Combinations of beta-lactam agents with beta-lactamase inhibitors (sulbactam, clavulanic acid, or tazobactam) are effective for infections due to bacteria producing beta-lactamases. However, it has been suggested that the recent emergence of beta-lactamase-overproducing S. aureus strains (18) and of other bacterial species producing inhibitor-resistant enzymes could be related to the selective impact of frequent clavulanate use (4, 5, 32). Our findings strongly support this hypothesis, because they indicate that the impact of amoxicillin-clavulanate on S. aureus survival in its ecological niche depends on the level of penicillinase production. This can be interpreted as due to a selection process which might be the result of a higher probability of survival for strains producing higher levels of beta-lactamase, while strains with lower levels of penicillinase are eradicated, when carriers are exposed to amoxicillin-clavulanate with tissue clavulanate concentrations resulting from amoxicillin-clavulanate doses currently used at the site of the ecological niche of S. aureus. To confirm this hypothesis, in vitro experiments comparing the impact of several clavulanate concentrations on mixed population of staphylococci should be performed.
Apart from the specific case of S. aureus type A beta-lactamase, which efficiently inactivates cefazolin and can lead to cefazolin treatment failure (21), the clinical significance of high-level penicillinase-producing S. aureus strains is not clear. It has been suggested that such strains could be involved in surgical wound infections (17, 19, 33). For none of our isolates were the oxacillin MICs higher than 2 μg/ml, suggesting that any clinical consequences would be minimal. However, amoxicillin-clavulanate selection pressure might, in the future, lead to the selection of strains producing enough penicillinase to confer clinical resistance. Thus, our findings argue strongly for increasing attention in the community (i) to surveillance of the emergence and spread of clinically relevant high-level penicillinase-producing S. aureus strains and (ii) to surveillance and monitoring of the population exposure to amoxicillin-clavulanate.
ACKNOWLEDGMENTS
We are grateful for the cooperation of the Conseil Départemental de l'Ordre des Médecins du Loiret, Caisse Primaire d'Assurance Maladie du Loiret, Service de Protection Maternelle et Infantile—Conseil Général du Loiret, Service de Médecine Scolaire et Inspection Académique de l'Education Nationale (Ministère de l'Education Nationale) du Loiret, Société des Pharmaciens du Centre, and Syndicat des Pharmaciens du Centre. We are also indebted to the schoolteachers, the children, and their parents for their invaluable contributions to this project and to P. Lavoine (INSERM U258, Villejuif, France) for secretarial assistance. We also thank C. Bernède (Institut Pasteur, Paris, France) for statistical assistance.
This study was funded in part by grants from the Institut National de la Santé et de la Recherche Médicale, from GlaxoSmithKline, and from the Delegation à la Recherche Clinique, Assistance Publique, Hôpitaux de Paris.
FOOTNOTES
- Received 2 March 2004.
- Returned for modification 24 May 2004.
- Accepted 25 July 2004.
- Copyright © 2004 American Society for Microbiology