Serotype Distribution and Antibiotic Susceptibility of Streptococcus Pneumoniae Isolates Causing Systemic Infections Among Children in Germany, 1992 to 1996

Streptococcus pneumoniae is a frequent cause of pneumonia, otitis media and bacteremia in children. The recent introduction of the Haemophilus influenzae type b vaccine in Germany has increased the relative importance of S. pneumoniae as a major cause of acute bacterial meningitis in children. Conjugation of pneumococcal polysaccharide antigens to carrier proteins may provide the means of preventing invasive S. pneumoniae infections during early childhood.1 An obstacle to this approach is that there are more than 80 serotypes of S. pneumoniae, and currently it does not appear possible to include more than a limited number of antigens in a conjugated formulation.2 Therefore a knowledge of the regional distribution of pneumococcal capsular types and antibiotic susceptibility in children is essential for the development of effective vaccine strategies and treatment protocols. Methods. In the present investigation 129 pneumococcal isolates from sterile body sites were obtained from patient specimens collected between December, 1992, and March, 1996, by 40 laboratories throughout Germany. Pneumococci were collected consecutively in each laboratory and sent to the Institute of Medical Microbiology in Aachen, where each isolate was confirmed as S. pneumoniae by optochin sensitivity and bile solubility. Only one isolate from each patient <14 years was included in the study. Capsular typing was carried out by the Quelling reaction using sera provided by the Statens Seruminstitut (Copenhagen, Denmark). Typing of pneumococcal strains belonging to serogroups was kindly done by Dr. J. Henrichsen, Statens Seruminstitut. MICs of penicillin, cefotaxime, erythromycin and chloramphenicol were determined for each isolate in Mueller-Hinton broth (Difco Laboratories, Detroit, MI) supplemented with lysed horse blood in a microdilution format (Merlin Diagnostika, Bornheim, Germany) as recommended by the National Committee for Clinical Laboratory Standards (NCCLS).3 Results. The mean patient age was 2.8 years. Seventy-five pneumococcal infections (58.1%) were seen in children <24 months old. Of 124 patients (96%) for whom the sex was reported, 73 (59%) were male. Pneumococcus was isolated from blood in 80 cases (62.0%), cerebrospinal fluid in 41 (31.8%) and other materials from normally sterile body sites in 8 (6.2%). The specimens from other normally sterile body sites (n = 8) included 5 strains cultured from pleural effusions and 1 strain each from joint fluid, peritoneal fluid and bone aspirate. Table 1 shows the proportion of serotypes from each body site for the 10 most common serotypes. Twenty-six capsular types were identified altogether, but the most common types in children, in order of decreasing frequency, were 14, 18C, 9V, 6A, 19F, 4, 3, 6B, 7F and 23F. Compared with serotypes of isolates from blood, types 18C and 7F were more frequently isolated from cerebrospinal fluid. Types 9V and 4 were relatively more often seen among the isolates from blood. Isolates from 116 patients (90.0%) were of serotypes included in the currently available 23-valent vaccine.TABLE 1: Serotype frequency and isolation site for pneumococcal isolates from 129 patients <14 years of age in Germany Table 2 shows antibiotic sensitivity and MIC90 values of all 129 S. pneumoniae isolates. The rates of resistance to cefotaxime, erythromycin and chloramphenicol were 0%, 7.0% and 5.1%, respectively. Only one strain was found to be intermediately susceptible to penicillin. This strain (serotype 6B) showed an elevated MIC value for cefotaxime (MIC 0.125 μg/l). Penicillin-resistant strains (MIC ≥ 2 mg/l) were not encountered.TABLE 2: Antibiotic susceptibilities of 129 strains of Streptococcus pneumoniae from systemic infections in children (Germany, 1992 to 1996) Our data are in accordance with those of a recent review of published and unpublished pneumococcal serogroup and serotype data of sterile site pneumococcal isolates from 16 countries. The authors showed, in order of decreasing frequency, types/groups 14, 6, 19, 18, 9, 23, 7, 4, 1 and 15 to be the most common serotypes in the developed countries.4 Type 3, which was shown to be responsible for 6.2% of pneumococcal infections in the present investigation, was not among the 10 most common types in that investigation.4 In the present investigation isolates from 116 patients (90.0%) were of serotypes included in the currently available 23-valent pneumococcal vaccine, but the level of protection offered by the usage of this vaccine in children would be much lower, because 75 pneumococcal infections (58.1%) were seen in children <24 months old, known to respond poorly to T cell-independent antigens. Among others a pneumococcal conjugate vaccine containing polysaccharide type antigens 4, 6B, 9V, 14, 18C, 19F and 23F for use in young children has been described.5 Because types 6A and 6B are likely to be cross-protective,6 this formulation would provide a level of coverage for 69.0% of children <14 years old in the current study. The addition of at least type 7F, which has been shown by the present study to be responsible for 12.2% of pneumococcal meningitis in Germany, to the vaccine formulation is therefore strongly suggested. Our observations highlight potential geographic differences in serotype distribution and underscore the need for widened and intensified international surveillance. As far as we know, in comparison with other countries7 the prevalence in Germany of pneumococcal isolates showing reduced susceptibility to penicillin in children and in adults8 is one of the lowest published to date worldwide. In contrast the rate of resistance to erythromycin is higher in children than in adults, possibly because of wide usage of this antibiotic in pediatric patients. Acknowledgments. We thank the staff of the participating institutions for providing the pneumococcal isolates and M. Lemperle for technical assistance. Our thanks are due to Professor Henrichsen and Dr. Konradsen, Statens Seruminstitut, Copenhagen, Denmark, for typing the pneumococci, belonging to groups. We thank Rhône-Poulenc Rorer Clinical Research for financial support of the study. Ralf René Reinert, M.D., D.T.M.&H. Achim Kaufhold, M.D. Johannes Josef Schlaeger Vinod Mechery Rudolf Lütticken, M.D. Institute of Medical Microbiology; Technical University (RWTH) Aachen; Aachen, Germany