Long-term effects on the nasopharyngeal flora of children following antimicrobial therapy of acute otitis media with cefdinir or amoxycillin-clavulanate

The nasopharynx of normal children is colonized by non-pathogenic aerobic and anaerobic organisms (Mackowiak, 1982), some of which have the ability to interfere with the growth of potential pathogens (Brook, 2000). Interfering organisms are isolated less often from the nasopharynx of otitis-media-prone children than normal controls (Brook, 2000; Bernstein et al., 1994). These organisms include the aerobic alpha-haemolytic streptococci (AHS, mostly Streptococcus mitis and Streptococcus sanguinis), anaerobic streptococci (Peptostreptococcus anaerobius) and Prevotella melaninogenica (Murray & Rosenblatt, 1976). Conversely, colonization by potential respiratory pathogens such as Streptococcus pneumoniae, Haemophilus influenzae and Moraxella catarrhalis increases significantly in otitis-media-prone children and in the general population of young children during respiratory illness (Faden et al., 1991). Nasopharyngeal flora with interfering capability may therefore inhibit colonization or growth of potential pathogens in the host and may play a role in preventing ear infections (Mackowiak, 1982; Bernstein et al., 1994).

Administration of antimicrobial agents can affect the composition of the nasopharyngeal bacterial flora (Foote & Brook, 1989). Oral flora organisms with interfering capability are generally susceptible to amoxycillin. These include aerobic and anaerobic streptococci, as well as penicillin-susceptible Pre. melaninogenica. Amoxycillin-clavulanate is also effective against beta-lactamase-producing Pre. melaninogenica. In contrast, these organisms are relatively resistant to the second- and third-generation cephalosporins (Brook & Gilmore, 1993).

This study compared the effects on the nasopharyngeal flora of two antimicrobial therapy for acute otitis media in children. One antimicrobial was amoxycillin-clavulanate, which is a wide-spectrum antimicrobial that is effective also against potential interfering organisms, and the other agent was cefdinir, a third-generation cephalosporin, which is a lesser inhibitor of these organisms.

Children diagnosed with acute otitis media and treated with either amoxycillin-clavulanate or cefdinir were included in this retrospective study. Included in the analysis were the first 25 patients who received amoxycillin-clavulanate and the first 25 who received cefdinir, and who also completed their course of therapy and had cultures as outlined below. The choice of the antimicrobial was made at the discretion of the examining physician. Patient age was similar in both groups and ranged from 6 months to 6 years (mean 2 years and 4 months), and 32 were males. Excluded from the analysis were those who had received antimicrobial therapy in the previous 3 months or the 2 month follow-up period, attended a day care centre, and those with an underlying illness or facial anomalies. The study was granted Institutional Review Board approval.

Nasopharyngeal cultures were obtained prior to therapy and on a follow-up visit 24 days after completion of 10 days of antimicrobial therapy and at follow-up visits on days 3035 and days 6065. These were obtained with calcium alginate swabs that were immediately plated onto media supportive of the growth of aerobic and anaerobic bacteria. The microbiologists were blinded to the patients therapy. The cultures were processed for the recovery of potential pathogens, and three types of organisms known to possess inhibitory activity: AHS, Peptostreptococcus and Prevotella species. The inhibitory activity of the organisms known to possess inhibitory activity was tested against one strain each of a recent clinical isolate of S. pneumoniae, H. influenzae and M. catarrhalis. Processing of specimens, identification of organisms, determination of beta-lactamase production and determination of inhibitory activity were done as previously described (Brook & Gober, 1995).

Patients received 90 mg amoxycillin-clavulanate kg¹ day¹ divided into two doses or 14 mg cefdinir kg¹ day¹ given in one dose. Both drugs were administered for 10 days. Compliance with therapy was assessed by a dosage card and inspection of the unused medicine after completion of treatment. Patients who missed two or more doses or who failed to return their medicine bottles and dosage cards were excluded. Statistical significance was calculated using the chi-square test.

Before therapy, 17 potential pathogens (S. pneumoniae, H. influenzae and M. catarrhalis) were isolated from the nasopharynx of 14 (56 %) of those treated with amoxycillin-clavulanate, and 20 potential pathogens were recovered from 15 (60 %) of those treated with cefdinir. Following therapy, at days 1215, the number of potential pathogens was reduced to a similar extent with both therapies, to three in those treated with amoxycillin-clavulanate and two in those treated with cefdinir. However, the number of potential pathogens rebounded faster in those treated with amoxycillin-clavulanate as compared with cefdinir in the two subsequent specimens taken at days 3035 and 6065 (12 and 18 in the amoxycillin-clavulanate, and six and nine in the cefdinir group, P < 0.01 and P < 0.001, respectively) (Table 1, Fig. 1).

Table 1. Potential pathogens and interfering organisms recovered from the nasopharynx of children treated with amoxycillin-clavulanate and cefdinir

(15K): Fig. 1. Potential pathogens recovered from the nasopharynx of children treated with amoxycillin-clavulanate (white bars) and cefdinir (black bars).

Following therapy, differences between the groups were also noted in the recovery of organisms with interfering capability. Immediately following amoxycillin-clavulanate therapy, the number of interfering organisms declined from 54 to 13, while following cefdinir treatment their number was reduced from 59 to 39 (P < 0.001) (Table 1, Fig. 2). The differences between the two therapy groups persisted in the two later specimens taken at days 3035 and days 6065 (25 and 38 in the amoxycillin-clavulanate group, and 52 and 51 in the cefdinir group, P < 0.001 and P < 0.05 respectively).

(16K): Fig. 2. Interfering organisms recovered from the nasopharynx of children treated with amoxycillin-clavulanate (white bars) and cefdinir (black bars).

Fifty-four of the 111 (49 %) interfering Prevotella species isolates produced beta-lactamase.

This study compared the effects of two types of antimicrobial therapies, amoxycillin-clavulanate and cefdinir, on the nasopharyngeal flora in children. Both agents are effective against penicillin-susceptible and -resistant potential pathogens (S. pneumoniae, H. influenzae and M. catarrhalis) but have selective activity against members of the nasopharyngeal bacterial flora. We found that at the end of amoxycillin-clavulanate therapy the oral flora is more depleted of organisms with interfering capability than at the end of cefdinir therapy. The differences between the two therapy groups persisted in the two other specimens taken 1 month and 2 months later and correlated with the faster reacquisition of potential pathogenic bacteria that took place in those treated with amoxycillin-clavulanate.

Amoxycillin-clavulanate is an antimicrobial with a broader spectrum of efficacy; it is active against AHS, anaerobic streptococci and penicillin-resistant Prevotella species (almost half of the Prevotella species isolates recovered in this study were beta-lactamase-producers). Cefdinir, in contrast, is less effective in vitro against AHS and beta-lactamase-producing anaerobic bacteria (Guay, 2000). Another possible mechanism for the increased survival of interfering aerobic and anaerobic streptococci following cefdinir treatment is the survival of beta-lactamase-producing Gram-negative anaerobic bacilli (including Prevotella species) which are resistant to cefdinir. The beta-lactamase produced by these organisms can shield the aerobic and anaerobic streptococci from cefdinir (Brook, 2001).

The presence of organisms with interfering potential may play a role in the prevention of respiratory infections. Bernstein et al. (1994) found a significantly greater number of colonies of AHS in the adenoids of non-otitis-prone children as compared to otitis-prone children. In contrast, they concomitantly recovered a higher number of non-type b H. influenzae in the otitis-prone group as compared to the non-otitis-prone group. These findings suggest the potential protective nature of AHS with inhibitory activity in preventing otitis media. The ability of the indigenous normal nasopharyngeal flora to inhibit colonization with potential pathogens has been studied. AHS were found to inhibit the colonization in patients and the in vitro growth of a variety of pathogenic bacteria, including S. pneumoniae, group A beta haemolytic streptococci and Staphylococcus aureus (Brook & Gober, 1995; Brook, 1999). The production of bacteriocin and other inhibitory substances that suppress some bacterial growth, or utilization of nutrients in the nasopharyngeal environment essential for the potential pathogens, may explain this phenomenon (Brook, 2001). Organisms other than AHS, such as Pre. melaninogenica and Pep. anaerobius, may also interfere with the growth of potential pathogens (Murray & Rosenblatt, 1976; Brook, 1999).

This study suggests a potential beneficial effect of using a narrow-spectrum antimicrobial that selectively spares the interfering organisms while eliminating pathogenic bacteria. The benefit of such therapy is the prevention of reacquisition of pathogenic bacteria in the nasopharynx. In contrast, utilization of a broad-spectrum antimicrobial is associated with prolonged absence of inhibitory organisms and rapid recolonization with pathogens. Further prospective studies are warranted to explore the clinical implications of these findings and find out if the recolonization with potential pathogens is associated with recurrence of ear and other infections.

The authors acknowledge the laboratory support of D. Giraldo.