Recombinant antithrombin attenuates pulmonary inflammation following smoke inhalation and pneumonia in sheep

Objective: The interaction between coagulation and inflammation has become one of the major topics in critical care medicine. In the present study, we investigated the effect of posttreatment of sepsis with recombinant human antithrombin. Design: Experimental laboratory in a university hospital. Setting: University laboratory. Subjects: Female merino ewes (n = 16). Interventions: After 1 wk of recovery from the surgical preparation, a tracheotomy was performed followed by insufflation of 48 breaths of cotton smoke (<40°C). Afterward, a stock solution of live Pseudomonas aeruginosa (5 × 10¹¹ colony-forming units) was instilled in the both lung lobes through a bronchoscope. All sheep were mechanically ventilated employing 100% oxygen. An infusion of recombinant human antithrombin (100 units·kg^-1·24 hrs^-1, intravenously; n = 6) or saline (n = 6) was started 1 hr after injury. Sham control animals (n = 4) were surgically prepared but not insufflated with smoke and bacteria. Lung histologic changes were evaluated by a scoring system. Measurements and Main Results: The infusion of recombinant human antithrombin maintained the baseline antithrombin activity throughout the study; in the saline-treated group, antithrombin activity decreased significantly. The lung wet/dry weight ratio and the histology score (combined scores for congestion, edema, inflammation, and hemorrhage) were significantly increased by the insult, but recombinant human antithrombin attenuated these responses. More than 30% of both bronchi and bronchioles were obstructed by cast formation after smoke inhalation and pneumonia. The cast was composed of epithelial cells, neutrophils, mucus, and fibrin. The obstruction was significantly improved by recombinant human antithrombin infusion. Arterial pressure and urine output were also attenuated in recombinant human antithrombin-treated animals. The increases in plasma nitrate/nitrite concentrations and pulmonary shunt fraction after the injury were not attenuated by recombinant human antithrombin. Conclusion: Posttreatment by recombinant human antithrombin was effective in treating acute lung injury after smoke inhalation and pneumonia in sheep. We hypothesize that the decrease in antithrombin activity during sepsis might induce severe airway obstruction and that supplementation with antithrombin inhibits this decrease.