Antibiotic-mediated release of endotoxin and the pathogenesis of gram-negative sepsis.

Since the earliest days of antibiotic chemotherapy to treat infection with Gram-negative microbes, investigators have recognized that such treatments may result in the release of microbial constituents that might, in turn, exacerbate the pathophysiological manifestations of disease. Both in vitro studies and in vivo animal experiments have over the years provided evidence in support of this concept; however, the actual clinical importance of this phenomenon to patients with Gram-negative sepsis is unclear. Recently published reports from a number of laboratories have shown that cell wall-active antibiotics that differ in their fundamental mechanisms of action in disrupting microbial growth (via selective interactions with various penicillin binding proteins) also differ in their relative ability to induce the release of biologically active endotoxin both in vitro and in vivo. Further, quantitative differences in total endotoxin release correlate well with antibiotic-initiated morphological changes in the microbe. Of potential significance is the finding that these differences are also reflected in differential production of cytokines from endotoxin-stimulated mononuclear phagocytes and other host target cells, including 11-6 and TNF. Since these immunologic hormones have been strongly implicated as contributing factors to the pathogenesis of Gram-negative sepsis, interest in the potential use of this chemotherapeutic approach as a means of controlling the host immunopathologic response has increased. Carefully controlled clinical trials in which different antibiotic treatments are correlated with production of cytokines will be of significant potential value in evaluating the actual significance of this phenomenon in the Gram-negative septic patient.