Thoracic epidural anesthesia time-dependently modulates pulmonary endothelial dysfunction in septic rats

INTRODUCTION: Increasing evidence indicates that epidural anesthesia improves postoperative pulmonary function. The underlying mechanisms, however, remain to be determined. Because pulmonary nitric oxide has been identified to play a critical role in pulmonary dysfunction in sepsis, we hypothesized that thoracic epidural anesthesia (TEA) modulates endothelial dysfunction via a nitric oxide-dependent pathway. METHODS: Thirty-six Sprague-Dawley rats underwent sham laparotomy or induction of peritoneal sepsis caused by cecal ligation and puncture (CLP). Septic animals were then treated with either bupivacaine 0.5% or normal saline epidurally (15 μl/h(-1)) for 6 hours or 24 hours after injury. Previous experiments demonstrated that these time points correspond with a hyperdynamic (at 6 hours) and hypodynamic circulation (at 24 hours), respectively. In addition, two sham control groups received either bupivacaine 0.5% or normal saline epidurally (15 μl/h(-1)). Six and 24 hours after injury, hemodynamic measurements and arterial blood gas analyses were performed in awake, spontaneously breathing rats. Exhaled nitric oxide, bradykinin-induced pulmonary vasoconstriction (a surrogate marker of endothelial dysfunction), pulmonary wet/dry-weight ratio (an estimate of pulmonary edema), and myeloperoxidase activity (MPO, a surrogate marker of neutrophil infiltration into lung tisssue) were investigated at 6 and 24 hours by using an established model of isolated and perfused lungs. RESULTS: In hyperdynamic sepsis, treatment with TEA resulted in reduced bradykinin-induced pulmonary vasoconstriction (P < 0.05) and lower levels of exhaled NO as compared with those in untreated septic rats (P < 0.05). However, the development of pulmonary edema or MPO activity in the lungs was not alleviated by sympathetic blockade in this phase of sepsis. Conversely, TEA led to an increased bradykinin-induced pulmonary vasoconstriction and pulmonary edema despite reduced exNO levels and pulmonary MPO activity in hypodynamic sepsis (each P < 0.05 versus CLP 24 h). Pulmonary gas exchange was only marginally affected under the influence of TEA in hypodynamic sepsis. Mean arterial pressure and heart rate were not affected beyond the changes caused by sepsis itself. CONCLUSIONS: The results of the present study suggest that TEA modulates the NO pathway and exerts positive effects on pulmonary endothelial integrity only in hyperdynamic sepsis. Whether the negative effects on endothelial function in hypodynamic sepsis have an impact on overall morbidity and mortality remains to be determined in future studies.