Improved Detection of Streptococcus pneumoniae in Middle-Ear Fluid Cultures by Use of a Gentamicin-Containing Medium

Abstract

The performance of Columbia agar medium with added sheep blood and gentamicin (CAG) for isolation of Streptococcus pneumoniae from middle-ear fluid cultures was compared to that of routine blood agar medium (BA). Of 238 pneumococcal isolates recovered, CAG plates detected 233 (97.9%) but BA plates detected only 208 (87.4%) (P < 0.001).

Increasing resistance to antimicrobial drugs among pathogens causing otitis media and especially Streptococcus pneumoniae has resulted in a renewed interest in middle-ear fluid (MEF) cultures (3–6). Performance of a diagnostic tympanocentesis confirms the diagnosis and enables identification of its etiology and determination of antibiotic susceptibility of the isolate. Despite use of appropriate culture techniques, between 16 and 26% of MEF cultures obtained from children with acute otitis media are sterile, yield organisms of dubious significance, or are contaminated by members of the normal flora of the ear canal, such as Pseudomonas aeruginosa (1, 8).

In a prospective study, the use of a selective medium for the recovery of S. pneumoniae from MEF cultures was evaluated and compared to conventional culture media.

Children presenting at the pediatric emergency room of the Soroka Medical Center in southern Israel with symptoms and signs of acute otitis media were included in the study. Although no precise record of the reasons for referral of these patients was kept, severe cases and failure of antibiotic treatment instituted by primary-care physicians were overrepresented in the population. Pus draining from spontaneous perforations of the tympanic membranes was collected on sterile cotton-tipped swabs. If the eardrums were intact, a tympanocentesis was performed by an otolaryngologist after parents signed a written informed consent. Following a strict sterile technique, the anteroinferior portion of the tympanic membrane was punctured and MEF was aspirated and applied onto a similar swab. Inoculated swabs were sent to the Clinical Microbiology Laboratory for bacteriological culture in transport medium (MW 173 Amies medium; Transwab; Medical Wire and Equipment, Potley, United Kingdom).

On arrival, swabs were plated in random order on Trypticase soy agar medium containing 5% sheep blood (BA), Columbia agar supplemented with 5% sheep blood and 5 μg of gentamicin (CAG) per ml, and chocolate agar and were inoculated into a thioglycolate broth tube supplied by a commercial source (Hylabs, Rehovot, Israel). Plates were incubated aerobically at 35°C in a 5% CO₂-enriched atmosphere for 48 h. Presumptive identification of S. pneumoniae was based on colony morphology, presence of alpha-hemolysis on BA and CAG plates, and inhibition by optochin, and was confirmed by a positive slide agglutination test (Phadebact; Pharmacia Diagnostics, Uppsala, Sweden). Identification of other organisms was performed by using routine bacteriological procedures. Antibiotic susceptibility of pneumococcal isolates to erythromycin, tetracycline, clindamycin, chloramphenicol, and trimethoprim-sulfamethoxazole was determined by the disk diffusion method following the National Committee for Clinical Laboratory Standards’ recommendations (7). Pneumococci exhibiting an inhibition zone diameter of <20 mm around a 1-μg oxacillin disk were further tested for penicillin susceptibility by the E test (PDM Epsilometer; AB Biodisk, Solna, Sweden), as recommended by the manufacturer. Organisms exhibiting decreased susceptibility to penicillin (MIC, >0.1 μg/ml) were tested for ceftriaxone susceptibility by the E-test method.

The performances of the BA and CAG media for the recovery of S. pneumoniae were compared by using the chi-square test. A P value of <0.05 was considered statistically significant.

During a 6-month period, a total of 823 MEF specimens, obtained from 651 patients, were processed. S. pneumoniae was recovered from at least one culture medium in 240 MEF specimens. Other significant organisms isolated included Haemophilus influenzae (278 isolates), Moraxella catarrhalis (31 isolates), and Streptococcus pyogenes (30 isolates). Two pneumococcal isolates were recovered from chocolate agar plates only and, therefore, were not included in the data analysis. No pneumococcal isolates were recovered from the thioglycolate tube only. Of the remaining 238 cultures positive for S. pneumoniae, the organism was isolated in the BA medium only in 5 (2.1%) cultures, in the CAG medium only in 30 (12.6%) cultures, and in both media in the remaining 203 (85.3%) cultures. Overall, BA plates detected 208 (87.4%) pneumococcal isolates and CAG plates detected 233 (97.9%) isolates (P < 0.001). In general, BA plates in which pneumococci could not be detected were covered by P. aeruginosa, Enterobacteriaceae, or staphylococci, whereas the CAG plates usually grew pneumococci in pure culture or showed less overgrowth of other organisms. In 14 of the 203 (6.9%) cultures positive by both media, isolation of S. pneumoniae from BA required performance of one or more subcultures, a step which was not necessary in the parallel CAG plate.

A total of 146 of 240 (60.8%) pneumococcal isolates exhibited decreased penicillin susceptibility, 5 (2.1%) were intermediately susceptible to ceftriaxone, 40 (16.7%) were resistant to erythromycin, 11 (4.6%) were resistant to clindamycin, 34 (14.2%) were resistant to tetracycline, 4 (1.7%) were resistant to chloramphenicol, 109 (45.4%) were resistant to trimethoprim-sulfamethoxazole, and only 75 (31.3%) were fully susceptible to all antimicrobial drugs tested. Among the 30 isolates recovered from CAG plates only, the following number (fraction) of isolates showed decreased susceptibility to the drugs indicated: 22 (73.3%), penicillin; 1 (0.3%), ceftriaxone; 5 (16.7%), erythromycin; 2 (6.7%), clindamycin; 3 (10.0%), tetracycline; 0 (0.0%), chloramphenicol; and 14 (46.7%), trimethoprim-sulfamethoxazole. Only 6 isolates (20.0%) were fully susceptible.

Culture media containing gentamicin were developed to enhance recovery of S. pneumoniae from respiratory carriers (2). The inhibitory effect exerted by the aminoglycoside reduces growth of other members of the normal respiratory flora and facilitates isolation of pneumococci which are intrinsically resistant to the drug.

Although antimicrobial therapy for patients with otitis media is usually prescribed on an empirical basis, isolation of the causative organism enables detection of antibiotic resistance and guides selection of appropriate treatment. The results of the present study show that the CAG medium significantly improves detection of the organism from MEF cultures. If cultures on CAG medium had been omitted, 12.6% of all pneumococci would have been missed and recovery of the additional 5.9% would have been delayed for at least 24 h. The clinical importance of this observation is emphasized by the high prevalence of antimicrobial resistance found in the present study. Although this observation may be partially explained by the bias resulting from inclusion of patients who failed to respond to antibiotic therapy, high rates of resistance to a wide array of antimicrobial drugs were also found among pneumococci in previous studies conducted in southern Israel (3–6, 8).

It is concluded that use of the CAG medium, in addition to conventional plating, may improve the management of patients with otitis media in an era of increasing antimicrobial drug resistance.