Background: Antibiotics are frequently used to treat critically ill patients, and its use is often accompanied by intestinal dysbiosis that might further lead to bacterial translocation (BT). Nevertheless, studies on the relationship between antibiotic therapy and BT are rare. In the present study, we investigated the effect of broad-spectrum antibiotics on BT in an experimental rat model of burn or sepsis injury.
Methods: The septic rat model was simulated by a second insult with lipopolysaccharides after burn injury. Ninety-two male Sprague-Dawley rats were randomly divided into control, burn, and sepsis groups (n = 8 or 9, each group), and the latter two groups were then treated with imipenem or ceftriaxone for 3 or 9 days. The mesenteric lymph nodes, liver, lungs, and blood were collected at each time point under sterile conditions for quantitative bacterial culture and strain identification. The differences between the groups were compared by Fisher exact test or Mann-Whitney U test.
Results: Only minimal Escherichia coli translocation to the mesenteric lymph nodes was observed in the normal control group, in which the BT rate was 12.5%. Burn injury did not affect the BT rate (Burn group vs. Control group, 12.5% vs. 12.5%, P = 1.000), whereas the BT rate showed an increased trend after the second insult with lipopolysaccharide (Sepsis group vs. Control group, 44.4% vs. 12.5%, P = 0.294), and many strains of Enterobacteria spp. were detected in distant organs (liver, lung, and blood) [Sepsis group vs. Control group, 0 (0,3) vs. 0 (0,0), U = 20, P = 0.045]. After the antibiotic treatment, BT to the distant organs was increased in burned rats [Burn IT3 group vs. Burn group, 0 (0,2) vs. 0 (0,0); Burn IT9 group vs. Burn group, 0 (0,1) vs. 0 (0,0); Burn CT9 group vs. Burn group, 0 (0,2) vs. 0 (0,0); all U = 20 and P = 0.076] but decreased in septic rats [Sepsis CT3 group vs. Sepsis group, 0 (0,0) vs. 0 (0,3), U = 20, P = 0.045]. The total amount of translocated bacteria, regardless of which antibiotic was used, was increased in burned rats [Burn IT9 group vs. Burn group, 2.389 (0,2.845) vs. 0 (0,2.301) Log10 colony-forming units (CFU)/g, U = 14, P = 0.034; Burn CT3 group vs. Burn group, 2.602 (0,3.633) vs. 0 (0,2.301) Log10 CFU/g, U = 10.5, P = 0.009], but there was a slightly decreased trend in septic rats [Sepsis IT9 group vs. Sepsis group, 2.301 (2,3.146) vs. 0 (0,4.185) Log10 CFU/g, U = 36, P = 0.721; Sepsis CT9 group vs. Sepsis group, 2 (0,3.279) vs. 0 (0,4.185) Log10 CFU/g, U = 32.5, P = 0.760]. Remarkably, the quantity of Enterococci spp. dramatically increased after broad-spectrum antibiotic treatment in both the burned and septic groups [Burn IT3 group vs. Burn group, 1 (0,5.164) vs. 0 (0,0) Log10 CFU/g, U = 16; Burn IT9 group vs. Burn group, 1 (0,2.845) vs. 0 (0,0) Log10 CFU/g, U = 16; Burn CT3 group vs. Burn group, 2.602 (0,3.633) vs. 0 (0,0) Log10 CFU/g, U = 8; Burn CT9 group vs. Burn group, 1 (0,4.326) vs. 0 (0,0) Log10 CFU/g, U = 16; Sepsis IT3 group vs. Sepsis group, 2.477 (0,2.903) vs. 0 (0,0) Log10 CFU/g, U = 4.5; Sepsis IT9 group vs. Sepsis group, 2 (0,3.146) vs. 0 (0,0) Log10 CFU/g, U = 9; Sepsis CT3 group vs. Sepsis group, 1.151 (0,2.477) vs. 0 (0,0) Log10 CFU/g, U = 18; Sepsis CT9 group vs. Sepsis group, 2 (0,3) vs. 0 (0,0) Log10 CFU/g, U = 13.5; all P < 0.05].
Conclusions: Broad-spectrum antibiotics promote BT in burned rats but prevent BT in septic rats, especially preventing BT to distant organs, such as the liver and lung. Moreover, Enterococci spp. with high drug resistance and high pathogenicity translocated most after antibiotic treatment.