Streptococcus pneumoniae : molecular epidemiology of isolates causing invasive disease and characterization of tolerance responses to lytic antibiotics

Abstract

Streptococcus pneumoniae is a major contributor to community-acquired pneumonia and invasive disease. Despite adequate treatment, the morbidity and mortality remain high. The increasing antibiotic resistance makes the scenario even worse. In this thesis the increase in invasive disease observed in Sweden during recent years was investigated, by studying isolates collected from the Swedish Microbiology Laboratories during 1987, 1992 and 1997 as well as other national and international isolates. These isolates were characterized by antibiotic susceptibility testing, and serotyping. Clonal relationships were studied by molecular fingerprinting methods. A major part of the increase in invasive pneumococcal disease was attributed to a rise of type 14 isolates from 1987 to 1992, and an increase of type 1 isolates during 1992 to 1997. The majority of the type 14 isolates belonged to one single penicillin susceptible clone, assigned S 14:1. This clone was also observed to spread globally. Also, among type 1 isolates a single clone, assigned S1:1, predominated. Rapid fluctuations in time and geographic localization of clones causing pneumococcal infections have implications on vaccine strategies. The current serotype distribution in Sweden implies a low coverage rate in children for the 7-valent conjugated vaccine to be introduced in Sweden. In contrast to bacteremic isolates, specific clones were not observed to spread among isolates from patients with meningitis.

To investigate prognostic factors in pneumococcal bacteremia and to study why there is a difference in mortality rates in pneumococcal bacteremia in patients from the United States and from Sweden, a prospective study of invasive disease was conducted in 5 countries. Older age, residence in a nursing home, chronic pulmonary disease and high APS scores were found to be independent predictors for death. APACHE II scores could be used as a sole prognostic factor at admission of patients with pneumococcal bacteremia. Type 14 was most common among all pneumococcal isolates, but type 3 dominated in the countries with the highest case- fatality rates, the United States and Spain. Molecular typing of the 5 most common serotypes showed clonal spread both in and between countries. Fewer patterns were found among type 3 isolates, suggesting a closer relationship than among isolates belonging to other serotypes. Pneumococci susceptible to penicillin did not cluster with penicillin non-susceptible isolates.

Treatment of invasive disease caused by antibiotic resistant or antibiotic tolerant bacteria may lead to therapeutic failures. Mechanisms by which bacteria become resistant/tolerant are important to study to be able to prevent further development of resistance. A mutant in the putative histidine kinase vncS was shown to be tolerant to penicillin and vancomycin. This mutant could more easily be transformed to high-level penicillin resistance than the parental strain R6, implying that tolerance might be a preferred background for importation of resistance determinants. In a rabbit model, a vncS mutant derivative of the encapsulated strain D39, could not be eradicated in contrast to the wildtype D39, indicating treatment failure. Penicillin tolerance was found in 8% of clinical isolates, though with differences between serotypes. Isolates with a reduced susceptibility to penicillin were three times more tolerant than susceptible isolates. Vancomycin tolerance was found in 3% of the isolates, all with a reduced susceptibility to penicillin. They were all of type 9V and were related by a molecular fingerprinting method. No resistance has yet been reported to vancomycin in pneumococci, the drug of last resort in patients with meningitis caused by multidrug resistant strains. However, tolerance may represent a favored background for future resistance development.