Influence of in-vivo endotoxin liberation on anti-anaerobic antimicrobial efficacy

The ability of cefoxitin, clindamycin, imipenem, meropenem, metronidazole a nd piperacillin-tazobactam to cause Gram-negative anaerobic bacteria to rel ease endotoxin and the influence of such liberated endotoxin on antibiotic efficacy were investigated in in-vivo experiments in animal models. Experim ental infections in various animal models (mice, hamster and infant rats) w ith cultures of wild and reference strains of Bacteroides fragilis group an d Fusobacterium spp. were carried out by injecting these animals with diffe rent inocula (10(6), 10(7) and 10(8) cfu/ml) of the bacterial suspension. A ppropriate doses of the test antibiotics were then injected and the plasma lipopolysaccharide (endotoxin) release measured by the Limulus Amoebocyte L ysate (LAL) Assay. Evidence of worsening of the outcome of the infections p ost-therapy was assessed, including histopathological changes in the intern al organs. Infection with generalized septicemia was established with F. nu cleatum in the mice and hamster models while with the B. fragilis group, in fections only led to intra-abdominal abscess formation. Plasma endotoxin re lease was higher in animals infected with F. nucleatum than B. fragilis and was unrelated to the bacterial inoculum. Imipenem, meropenem and cefoxitin , in that order, induced the highest levels of endotoxin activities in the animal model, particularly following F. nucleatum infection. Histological e xamination of the internal organs of various animals showed variation in th e pattern of histopathological changes; grades 3-4 inflammatory changes in the liver were observed in the Fusobacterium-infected animals that were tre ated with the carbapenems and cefoxitin. Therapy with the other antibiotics did not exacerbate anaerobic sepsis. In this study, bacteremia did not lea d to massive endotoxin release and antibiotic therapy appeared not to have negatively influenced the outcome of most of the Gram-negative anaerobic in fections, except for infections caused by Fusobacterium spp. However, it is conceivable that if the gastrointestinal tract is the source of the endoto xin in patients with systemic inflammatory response syndrome, then the obli gate anaerobes like Bacteroides and Fusobacterium species, which are member s of the gut flora, may play a major role in the unfavorable outcome of ant ibiotic therapy in some of these infections.