Production of inflammatory cytokines in ventilator-induced lung injury: a reappraisal.

We investigated the production of proinflammatory cytokines by the lung during high mechanical stretch in vivo. To do this, we subjected rats to high-volume (42 ml/kg tidal volume [VT]) ventilation for 2 h. The animals developed severe pulmonary edema and alveolar flooding, with a high protein concentration in bronchoalveolar lavage fluid (BALF). The animals' BALF contained no tumor necrosis factor (TNF)-alpha, negligible amounts of interleukin (IL)-1beta, and less than 300 pg/ml of the chemokine macrophage inflammatory protein (MIP)-2, an amount similar to that found in rats ventilated with 7 ml/kg VT. Systemic cytokine levels were below the detection threshold. Because isolated lungs have been shown to produce high levels of proinflammatory cytokines when ventilated with a similarly high VT for the same duration (Tremblay, et al. J Clin Invest 1997;99:944-952), we reconsidered this specific issue. We ventilated isolated, unperfused rat lungs for 2 h with 7 ml/kg or 42 ml/kg VT, or maintained them in a statically inflated state. Negligible amounts of TNF-alpha were found in the BALF whatever the ventilatory condition applied. The BALF IL-1beta concentration was slightly elevated and higher in lungs ventilated with 42 ml/kg VT than in those ventilated with 7 ml/kg VT or in statically inflated lungs (p < 0.05). The BALF MIP-2 concentration was moderately elevated in all isolated lungs (200 to 300 pg/ml), and was slightly higher (p < 0.05) in lungs ventilated with 42 ml/kg VT. After lipopolysaccharide (LPS) challenge, high levels of TNF-alpha, IL-1beta, and MIP-2 were found in the animals' plasma before the lungs were removed. Negligible amounts of TNF-alpha and IL-1beta were retrieved from the BALF of statically inflated lungs. The concentrations of TNF-alpha and IL-1beta were higher in the BALF of ventilated lungs (p < 0.001). The TNF-alpha level did not differ with the magnitude of VT, whereas the level of IL-1beta was significantly higher in BALF of lungs ventilated with 42 ml/kg VT (p < 0.01). The MIP-2 concentrations were similar for the two ventilatory conditions. These results suggest that ventilation that severely injures lungs does not lead to the release of significant amounts of TNF-alpha or IL-1beta by the lung in the absence of LPS challenge but may increase lung MIP-2 production.