Macrophage cyclooxygenase expression, immunosuppression, and cardiopulmonary dysfunction after blunt chest trauma

Background: Two series of experiments were performed in swine who received severe blunt chest trauma. The goals were to determine the time course of c onstitutive and inducible cyclooxygenase (COX) isozyme expression in pulmon ary macrophages (Mos), and to determine whether COX expression and cardiopu lmonary dysfunction were altered when neutrophils (PMNs) were pharmacologic ally depleted with cyclophosphamide (CYC). Methods. In series 1 (n = 17), anesthetized, mechanically ventilated swine were subjected to right chest trauma via captive bolt gun, hemorrhage, and a 60-minute shock period. In series 2 (n = 41), CYC (50 mg/kg intravenously ) was administered 4 days before trauma, and the shock period was shortened to 30 minutes. In both series, hemodynamic support and supplemental oxygen were provided for an additional 60 to 90 minutes after shock. Mos were iso lated from serial bilateral bronchoalveolar lavages (BALs) and COX protein expression was measured with Western blots. Results. In series 1, death occurred in 11 of 17. In survivors, Mo COX-1 pe aked at > 100 times baseline in both right BAL and left BAL by 60 minutes ( before resuscitation). Changes in Mo COX-2 were minimal. In series 2, befor e trauma, CYC (n = 16) reduced circulating and BAL PMNs by > 90% relative t o control (n = 25, both p < 0.05) with no complicating side effects. After trauma, death occurred in 11 of 25 controls versus 9 of 16 with CYC. In sur vivors, Pao(2)/Fio(2) was < 250 and PaCO2 was 25% higher on constant minute ventilation, indicating mismatched ventilation/perfusion; both changes wer e reduced with CYC (p < 0.05). In controls, bilateral histologic damage inc luded edema, alveolar hemorrhage, and interstitial infiltrates. These chang es were reduced by one third with CYC (p = 0.08). Trauma-induced changes in BAL protein, BAI, elastase, or Mip COX expression were not lessened by CYC . Conclusion., After unilateral chest trauma, M phi COX-1, not COX-2, is indu ced bilaterally and before fluid resuscitation; CYC prevented PMN infiltrat ion and attenuated structural and functional changes after resuscitation, w hich suggests that PMNs have a role in the pathogenic mechanism of secondar y lung injury; M phi COX expression and other injury markers were not alter ed by CYC; and since M phis continued to express proinflammatory COX protei n even after pretreatment with a powerful nonspecific immunosuppressant, an d since there is residual alveolar capillary damage even in the absence of PMNs, it is logical to conclude that no single cell type or mediator is a p ractical therapeutic target and that novel resuscitation strategies must ad dress multiple elements in the inflammatory cascade.