Antimicrobial-induced endotoxin and cytokine activity in an in vitro model of septicemia in foals

OBJECTIVE: To determine which antimicrobials that are used to treat neonatal foals with septicemia attributable to Escherichia coli will minimize endotoxin release from bacteria and subsequent activity of inflammatory mediators while maintaining bactericidal efficacy. SAMPLE POPULATION: Blood samples from 10 healthy foals. PROCEDURE: Escherichia coli isolates A and B were isolated from 2 septicemic foals, and minimal inhibitory concentrations (MIC) were determined for 9 antimicrobials. Five of these antimicrobials were tested in vitro at 2 and 20 times their respective MIC. Whole blood or mononuclear cells grown in tissue-culture media were incubated with 105 colony-forming units of E. coli and each antimicrobial or saline (0.9% NaCl) solution. After 6 hours, number of viable bacteria remaining was determined, and supernatant was tested for endotoxin and tumor necrosis activity. RESULTS: Testing in whole blood was compromised by bactericidal effects of the blood itself. In mononuclear cell suspensions, each antimicrobial significantly reduced the number of viable bacteria to low or undetectable amounts. Antimicrobials did not differ significantly in efficacy of bacterial killing. Amikacin used alone or in combination with ampicillin resulted in significantly less endotoxin activity than did ampicillin, imipenem, or ceftiofur alone. There was a correlation between TNF-alpha and endotoxin activity. CONCLUSIONS AND CLINICAL RELEVANCE: Aminoglycosides appear less likely to induce endotoxemia and TNF-alpha synthesis during bactericidal treatment of E. coli septicemia, compared with beta-lactam antimicrobials. Use of ampicillin, imipenem, or ceftiofur in the treatment of septicemic neonatal foals should be accompanied by appropriate treatment for endotoxemia.