Bactericidal antibiotics temporarily increase inflammation and worsen acute kidney injury in experimental sepsis.

OBJECTIVE: To explore the relationships among bactericidal antimicrobial treatment of sepsis, inflammatory response, severity of acute kidney injury, and outcomes.

DESIGN: Controlled laboratory experiment.

SETTING: University laboratory.

INTERVENTIONS: Sepsis was induced by cecal ligation and puncture in 52 rats and was treated with either bactericidal antibiotics (ampicillin/sulbactam) or placebo (saline). Serial blood specimens were obtained after cecal ligation and puncture for serum creatinine, interleukin-6, and neutrophil gelatinase-associated lipocalin concentrations. RIFLE (Risk, Injury, Failure, Loss, End-stage kidney disease) criteria were used to assess severity of acute kidney injury. All animals were observed for survival up to 1 wk. In a separate experiment, six healthy animals were given antibiotics and renal function was assessed. Another 12 animals were euthanized 2 days after laparotomy for kidney histology.

MEASUREMENTS AND MAIN RESULTS: Survival in the placebo group was 50% compared with 81.8% in the antibiotic group (p < .05). Most animals (93%) without antibiotics developed acute kidney injury, of which 39% exhibited greater than a threefold rise in serum creatinine (RIFLE-F). Furthermore, survival decreased as acute kidney injury severity increased. Surprisingly, all antibiotic-treated animals developed acute kidney injury, of which 68.6% reached RIFLE-F. However, renal dysfunction was less persistent in these animals. Patterns of plasma interleukin-6 were similar to creatinine with higher concentrations seen earlier in antibiotic-treated animals but with faster resolution. Interleukin-6 concentration at 24 hrs was independently associated with the development of RIFLE-F. Histologic findings were consistent with functional parameters showing that antibiotics worsened acute kidney injury.

CONCLUSION: In polymicrobial sepsis, bactericidal antibiotics resulted in more inflammation and more severe acute kidney injury. However, resolution of inflammation and acute kidney injury was faster with antibiotics and correlated best with survival. These results suggest that transient worsening of renal function may be an expected consequence of sepsis therapy. These findings also question the value of peak severity of acute kidney injury as a primary end point and suggest that resolution of acute kidney injury may be more appropriate.