Pneumothorax-associated pleural eosinophilia is tumour necrosis factor-alpha-dependent and attenuated by steroids

Background and objectives: The pathogenesis and the optimal treatment of eosinophilic pleural effusions are unknown. We aimed to examine whether pneumothorax‐associated pleural eosinophilia in mice is dependent on tumour necrosis factor (TNF)‐alpha, and whether it is affected by systemic administration of corticosteroids. Methods: Mice were injected intrapleurally with 0.4 mL air to create pneumothoraces. Animals were sacrificed 24 or 48 h later, and pleural lavage (PL) was performed. In the first experiment, comparisons were made between wild‐type and TNF‐α knockout mice with pneumothorax. In the second experiment, wild‐type mice were injected intraperitoneally with different doses of dexamethasone (0, 0.25, 0.5 and 1 mg/kg), 5 min before and 24 h after the induction of pneumothorax. Results: After induction of a pneumothorax, TNF‐α knockout mice had significantly fewer total number of cells (P = 0.004), mononuclear cells (P = 0.01), neutrophils (P = 0.017) and eosinophils (P = 0.002) in their PL compared with wild‐type animals. TNF‐α was detected in the PL of most of the control mice but not in TNF‐α knockouts. Dexamethasone induced a significant, dose‐dependent reduction of PL total cells (P < 0.001), eosinophils (P < 0.001), mononuclear cells (P = 0.007) and lymphocytes (P = 0.04) at 48 h, and significantly reduced the number of PL total cells (P = 0.045) and eosinophils (P = 0.005) at 24 h. Furthermore, dexamethasone prevented eosinophil infiltration of lung and pleural tissue. Conclusion: Pneumothorax‐associated pleural eosinophilia in mice is TNF‐α‐dependent and is significantly attenuated by corticosteroid treatment. In addition, both TNF‐α deficiency and dexamethasone treatment were associated with a significant reduction of other types of inflammatory cells in PL.