Tel.+49 30 838 62450 Fax.+49 30 838 62522
With regard to causes of morbidity and mortality worldwide, community-acquired pneumonia holds the highest significance and Streptococcus pneumoniae (S. pneumoniae) is the most prevalent causative pathogen of community-acquired pneumonia. Despite adequate antibi-otic therapy and constant endeavour to improve the clinical outcome of high risk patients, existing treatment and drug regimens are still insufficient, considering the high mortality (above 20 %) for immunocompetent patients in intensive care units hospitalized for severe community-acquired pneumonia. The main contributing factor to an adverse outcome of pa-tients suffering from severe pneumonia is an excessive activation of the innate immune sys-tem, which leads to the development of acute respiratory failure, sepsis and multiple organ failure. Thus, limitation of exaggerated inflammatory reactions, combined with an effective antibiotic treatment might improve the outcome of affected patients. Efficient therapeutic protocols of various respiratory infections utilize Quinolones. Apart from their well described antimicrobial characteristics, Quinolones were suggested to exert anti-inflammatory effects. Particularly moxifloxacin has been shown to inhibit the release of proin-flammatory cytokines in in vitro experiments, using human monocytes and human alveolar epithelial cells stimulated with heat-inactivated pneumococci or lipopolysaccharides. Howev-er, clinical studies demonstrated a less clear picture. In a retrospective study, patients with severe community-acquired pneumonia were treated with moxifloxacin, which resulted in improved outcome as compared to monotherapy with -lactam antibiotics. It was hypothe-sized that potential immunomodulatory features of moxifloxacin were responsible for this improvement of patient survival. In contrast, a prospective study showed no clinical benefit for patients with sepsis, who were treated with moxifloxacin in addition to carbapenem antibi-otics. This study aimed at analysing the impact of moxifloxacin on the inflammatory host re-sponse in ex vivo cultured intact human lung tissue and in a murine model of severe pneu-mococcal pneumonia. Therefore, ex vivo cultured human lung tissue was stimulated with TNF- or infected with S. pneumoniae, treated with ampicillin, moxifloxacin or solvent and proinflammatory cytokines were quantified in the supernatant.
For ampicillin which is one standard treatment in community-acquired pneumonia, anti-inflammatory effects have not been described so far. Therefore, it was used as reference antibiotic in this study. Upon stimulation with TNF- human lung tissue produced IL-6 and IL-8, which was reduced by moxifloxacin but not by ampicillin. Infection with S. pneumoniae also induced IL-6 and IL-8 production in human lung tissue which was not influenced by treatment with moxifloxacin or ampicillin. Additionally, a murine model was established, in which antibiotic treatment with ampicillin or moxifloxacin was applied in severe pneumococcal pneumonia to mimic the clini-cal treatment initiated in the emergency department. Mice infected with a lethal dosage of S. pneumoniae were treated in 12-hour intervals with ampicillin, moxifloxacin or solvent, starting 24 hours post infection. In addition, sham infected and solvent treated animals were investi-gated. At specified time points post infection, the bacterial burden in lung, blood and spleen was analysed. Cytokine levels were measured in bronchoalveolar lavage fluid and blood plasma. Furthermore, proinflammatory gene regulation in murine lung tissue was investigat-ed by real-time polymerase chain reaction. Bronchoalveolar lavage fluid and blood leuko-cytes were differentiated and pulmonary vascular permeability was determined. Additionally, lung tissue from infected mice was histologically analysed. Upon infection with S. pneumoni-ae, mice developed severe pneumonia within 24 hours post infection. Antibiotic treated mice were rescued from developing acute respiratory distress syndrome and sepsis, and recov-ered from clinical signs of pneumonia within the observation period. Pulmonary bacterial bur-den was lower in moxifloxacin treated mice in contrast to ampicillin from 36 hours post infec-tion on, and fewer cases with bacteraemia were observed when compared with ampicillin treated mice. At 36 hours post infection, ampicillin treated mice showed increased pulmonary vascular permeability, which could not be attributed to the lower bacterial burden in moxiflox-acin treated mice. This was shown by an additionally investigated group of mice treated with moxifloxacin and ampicillin simultaneously. Animals in this group displayed equally high permeability as mice with ampicillin monotherapy. Bronchoalveolar lavage fluid and blood cytokine levels, proinflammatory gene expression, blood leukocytes and morphologic lung injury scores were similar in mice treated with moxifloxacin or ampicillin. To estimate a poten-tial effect of moxifloxacin on the integrity of the cellular monolayer, transcellular electrical resistance of human umbilical vein endothelial cell monolayers was measured in the pres-ence of moxifloxacin or ampicillin with or without thrombin stimulation. Moxifloxacin exerted no direct stabilising effects on the endothelial barrier, and it also did not reduce thrombin-induced disruption of integrity. The same applied to ampicillin.
In summary, in human lung tissue moxifloxacin treatment reduced TNF-α induced inflamma-tion in contrast to ampicillin. However, ampicillin and moxifloxacin showed similar effects in pneumococci-induced inflammation of human lung tissue and in severe murine pneumococ-cal pneumonia. Nevertheless, treatment with moxifloxacin reduced pulmonary bacterial load and bacteraemia in contrast to ampicillin. Pulmonary vascular permeability in the acute phase of murine pneumococcal pneumonia was enhanced by ampicillin in contrast to moxi-floxacin treatment, which was independent from effects of moxifloxacin.