INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

-Lactamases are enzymes that hydrolyze the -lactam ring of a -lactam drug, leading to the inactivation of the drug. This is the most common resistance mechanism, especially among the gram-negative bacteria (7). Until the latter half of the 1970s, penicillins and cephalosporins were in most cases still effective against oral gram-negative anaerobes. In 1977, Murray and Rosenblatt (16) reported a frequency of -lactamase production by Bacteroides melaninogenicus (now reclassified as Prevotella melaninogenica and several other pigmented gram-negative rods) exceeding 56%. During the 1980s and 1990s, the production of -lactamases by these bacteria steadily increased. Anaerobic gram-negative rods, including Prevotella, Porphyromonas, and Fusobacterium species, can be involved in pyogenic orofacial and upper respiratory tract infections. These organisms may also protect otherwise-susceptible pathogens from a -lactam agent, which in turn leads to a failure of penicillin treatment (6). Cases of clinical failure with penicillin treatment for human orofacial infections have been documented along with reports suggesting that previous penicillin therapy increases the incidence of penicillin-resistant bacteria in the oral cavity (2, 3, 8, 23).

Previous reports of -lactamase production have been based mainly on hospitalized adults and children. To our knowledge -lactamase production has not been studied for healthy infants in a nonhospital setting. Otitis media, among other upper respiratory tract infections, is one of the most common infections in childhood (22). This infection is often treated with antimicrobial agents, especially -lactam antibiotics. Therefore, if antibiotic use provokes -lactamase production (7), it would be reasonable to assume that young children easily harbor -lactamase-producing strains. Indeed, previous studies (12, 13) have shown that -lactamase production by certain groups of oral anaerobes in early childhood can be much more frequent than generally anticipated.

The aim of the present longitudinal study was to examine the age-related frequency of -lactamase-producing oral anaerobic gram-negative species in infants during their first year of life in relation to their exposure to antimicrobial agents.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Subjects. The original study population, from whom 50 consecutive healthy infants were chosen for a detailed bacteriologic investigation, comprised 329 children participating in the Finnish Otitis Media Cohort Study in Tampere, Finland, from 1994 to 1997. The infants attended scheduled visits at the study clinic, where bacterial samples were collected and a wide variety of background data was gathered by interviewing the parent(s). If an infant became sick between the visits, the infection was treated in the same clinic, which provided the information of the antimicrobial history. Forty-four Caucasian infants, 2, 6, and 12 months of age, with complete bacteriologic and demographic data were included in the present study.

Bacterial samples and culture. During the scheduled visits, unstimulated saliva samples (0.1 to 0.3 ml) were collected from the cheek area with a calibrated plastic pipette. The measured samples were placed in VMGA III transport medium (5), delivered to the laboratory by overnight mail, and processed within 24 h.

Testing of -lactamase activity. Out of 735 anaerobic gram-negative isolates, 195 isolates showing decreased susceptibility to penicillin G (1 µg/ml; inhibition zone 20 mm) (as determined with special potency antibiotic disks [Oxoid, Unipath Limited, Basingstoke, Hampshire, United Kingdom]) were included for further examination (12). -Lactamase production was tested by using the qualitative chromogenic cephalosporin disk test (nitrocefin; AB Biodisk, Solna, Sweden). The results were read after 15, 30, and 60 min.

Statistics. The statistical significance of the association between the amount of -lactamase-producing species or groups and the exposure to antimicrobial agents was tested by using Mantel-Haenzel chi-square statistics. The exposure to antimicrobial agents was counted as follows: mothers' antimicrobial courses (in 15 cases) during the 6-month period before the first sampling occasion, at 2 months of age (maternal exposure); infants' antimicrobial courses between 2 and 6 months of age (early exposure); and infants' antimicrobial courses between 6 and 12 months of age (late exposure).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

During the infants' (n = 44) first year of life, the occurrence of -lactamase-producing gram-negative anaerobic species or groups increased with age, from 11% (5 infants) at 2 months to 55% (24 infants) at 6 months to 89% (39 infants) at 12 months. The mean number of -lactamase-producing species or groups per subject increased from 0.1 (range, 0 to 2) at 2 months to 1.8 (range, 0 to 4) at 12 months (Table 1). The age-related frequency and distribution of -lactamase-producing bacteria by species or groups are presented in Table 2. The most pronounced changes were observed in the colonization of the Prevotella melaninogenica group (Prevotella denticola, Prevotella loescheii, and P. melaninogenica), which increased from 11% at 2 months to 73% at 12 months. Other -lactamase-producing species with a frequency exceeding 20% at 12 months consisted of nonpigmented Prevotella species (including pentose fermenters and nonfermenters), Porphyromonas catoniae, and Fusobacterium nucleatum. The decreased susceptibility (inhibition zone  20 mm) to penicillin (1 µg/ml), as determined with a penicillin G-impregnated disk, coincided well with the -lactamase production by a qualitative chromogenic cephalosporin disk test, except for two isolates. Follow-up colonization by -lactamase-producing species was assessed by comparing the positive findings obtained at 6 months with those (positive or negative) obtained at 12 months. Fourteen of the 16 infants who at 6 months of age harbored -lactamase-producing P. melaninogenica group isolates still harbored them at 12 months of age. Two of three infants still harbored nonpigmented Prevotella group isolates at 12 months, and one of four infants still harbored -lactamase-producing P. catoniae isolates at 12 months.

View this table: [in this window] [in a new window]   TABLE 1.   Number of infants at various ages harboring -lactamase-producing gram-negative anaerobic species or groups

View this table: [in this window] [in a new window]   TABLE 2.   Distribution and age-related frequency of -lactamase-producing bacterial isolates by species or groups

Thirteen infants had not been exposed to antimicrobial agents through maternal or personal courses during their first year of life. They harbored, on average, 1.31 -lactamase-producing species or groups each at 12 months. Seventeen children who had had one or two exposures to antimicrobial agents through maternal courses (antimicrobial preparation not known) or personal courses (including -lactams and/or trimethoprim-sulfamethoxazole combinations intermittently) harbored, on average, 1.82 -lactamase-producing species or groups each. Children with three or more exposures to antimicrobial courses (including maternal courses and -lactams and/or trimethoprim-sulfamethoxazole combinations intermittently) harbored, on average, 2.21 -lactamase-producing species or groups each at 12 months. Twenty-six of the 31 children exposed to antimicrobial agents were exposed to -lactams. The most common indication for medication was otitis media, and the most common antimicrobial agents used were amoxicillin (38 courses) and a trimethoprim-sulfamethoxazole combination (28 courses). The number of antimicrobial treatments given to the infants during their first year of life varied between 0 and 9 per infant. Six infants were on antimicrobials during one or more sampling occasions.

A correlation between maternal antimicrobial exposure and the number of -lactamase-producing species or groups observed in infants at the age of 12 months was detected (P = 0.026). No positive correlation between early exposure and -lactamase production at 12 months of age could be found. A clear correlation between late exposure and -lactamase-producing species or groups in infants at 12 months of age was detected (P = 0.006) (data not shown). The Mantel-Haenzel chi-square analyses showed a significant association between the exposure to antimicrobial agents during the first year of life and the amount of -lactamase-producing species at 12 months (Table 3).

View this table: [in this window] [in a new window]   TABLE 3.   No. of antibiotic exposures during the first year of infants' lives and the number of -lactamase-producing species at 12 months

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The frequency of -lactamase production by anaerobic gram-negative rods increased considerably during the first year of life of the study infants. In addition, the number of simultaneously isolated -lactamase-producing species also increased with age. Previous antimicrobial exposures had a significant effect on the development of resistance.

The predisposing factors involved in the origin and evolution of -lactamase production remain unclear. Almost all gram-negative rods produce small amounts of chromosomal -lactamases (21). However, -lactamases encoded by genes that are transferred extrachromosomally, as well as strong induction of chromosomal -lactamases, have caused great clinical problems during recent years. As reported earlier, the use of antimicrobial drugs has been known to increase the frequency of -lactamase production for both anaerobes and aerobes (17, 19). Acute otitis media episodes are especially common during the first year of life (1, 22), and they are the main reason for antimicrobial treatment, often recurrent, in early infancy. It is generally accepted that to limit the spread of bacterial resistance, the use of antimicrobial agents should be controlled. To combat bacterial infections, it is necessary that the dosage to be administered is proper, as subinhibitory concentrations may lead to gradual development of plasmid-mediated resistance under selection pressure or acquisition of resistance genes from other bacteria (17). In the present study, the infants who had no personal antibiotic history during the surveillance year or maternal exposure before the age of 2 months harbored, on average, 1.31 -lactamase-producing species at 12 months of age. When the infants had been exposed to antimicrobial agents (in most cases alone but in some cases in combination with another antimicrobial agent[s]), the number of -lactamase-producing species increased significantly. Our observation is well in line with that of Brook and Gober (4), who demonstrated an increase in the number of penicillin-resistant oropharyngeal bacteria due to the use of amoxicillin as preventive medication against recurrent otitis media in children. In addition to an infant's exposure to antibiotics, we also estimated the possible influence of the mother's antimicrobial use on the colonization of -lactamase-producing bacteria in infants. Indeed, bacterial transmission from mothers to infants could explain why infants at 2 months of age whose mothers were on antimicrobial medication harbored more -lactamase-producing species than infants with no exposure at all. Selection of -lactamase-producing subpopulations and proliferation of them could explain the high incidence of resistant species at 12 months.

Plasmid-mediated resistance in the bacteria of both humans and animals has been previously described (24). However, plasmid analysis performed for P. melaninogenica in a previous study (12) showed no correlation between the production of -lactamase and the presence of plasmids. It has been suggested that genes that encode resistance to tetracyclines and penicillin from anaerobes from different families and species may be transferred together to the host (18, 24). However, according to a recent study (14), -lactamase-producing F. nucleatum isolates from young children were resistant to penicillin G (MICs, 2 to 256 µg/ml) but susceptible to tetracycline hydrochloride. The factors behind -lactamase production need to be investigated further. Aspects of special interest would be the composition of the oral microflora and the antimicrobial drug history of the family members. Special attention should be paid to children in day care, as they are more likely to be exposed to and transmit early acute otitis media to their infant siblings (15).

According to a previous study (11), strain turnover may easily occur in young children with a developing oral ecosystem. In this study, some of the infants who harbored -lactamase-producing isolates at 6 months did not harbor them at 12 months. It is possible that a turnover of the bacterial population in the biological environment, from -lactamase-positive strains to -lactamase-negative strains, had occurred or that methodologic biases included adverse factors inherent in the sampling efficacy or merely in the ability to isolate the index species from mixed cultures. However, to evaluate the stability of a strain, genetic methods (e.g., ribotyping, arbitrarily primed PCR, or an equivalent method) should be used to document the degree of clonal similarity.

One of the explanations for the high frequency of -lactamase production is certainly the fact that P. melaninogenica is a genotypically heterogeneous species (11), which means that one individual can simultaneously harbor several strains with different characteristics, in this case both -lactamase-positive and -lactamase-negative strains, and with indistinguishable genotypes (12). Therefore, in this study several isolates per infant, if available, were tested for -lactamase production, because if only one isolate per infant had been tested, the true rate of colonization by -lactamase-producing strains might have been underestimated.

The child's immediate environment and close contacts have a great influence on the development of the early oral microflora. On the basis of the present investigation, we conclude that the high frequency of -lactamase-producing bacterial strains reflects the net effect of the child's and the family members', especially the mother's, antibiotic exposure.