Liquid ventilation provides uniform distribution of perfluorocarbon in the setting of respiratory failure.

BACKGROUND

We evaluated the effect of perfluorocarbon liquid ventilation (LV) on gas exchange and pulmonary function in the setting of respiratory failure and the distribution of the ventilating medium during LV when compared to gas ventilation (GV).

METHODS

Ten sheep, 17.3 +/- 4.2 kg in weight, underwent oleic acid induction of lung injury followed by either GV (n = 5) or perfluorocarbon LV (n = 5). After 1 hour animals were killed, and chest computed tomographic (CT) imaging was performed. Average CT attenuation number was assessed as an indicator of the distribution of gas or perfluorocarbon in the dependent (posterior) and nondependent (anterior) zones of the lung (air = -1000; soft tissue = 0; perfluorocarbon = +2300 Hounsfield units [H]).

RESULTS

Significant increases in PaO2 (LV = 298 +/- 76 mm Hg, GV = 43 +/- 18 mm Hg, p < 0.001), SvO2 (LV = 74% +/- 6%, GV = 32% +/- 18%, p < 0.01), and lung compliance (LV = 1.65 +/- 0.50 ml/cm H2O/kg, GV = 0.58 +/- 0.06 ml/cm H2O/kg, p < 0.01) were observed. Significant decreases in physiologic shunt (LV = 24% +/- 6%, GV = 62% +/- 14%, p < 0.01) were noted. CT attenuation data showed the presence of minimal gas ventilation in the dependent regions during GV although the nondependent regions remained well aerated (CT attention number during GV: ND = -654 +/- 160 H; D = -92 +/- 160 H, p < 0.0001). During LV, there was a fairly homogenous distribution of perfluorocarbon in the lungs (CT attenuation number during LV: D = 1071 +/- 330 Hounsfield units; ND = 1112 +/- 287 Hounsfield units; p = 0.240). Lung biopsy analysis in the LV animals was consistent with a reduction in intraalveolar hemorrhage, intraalveolar edema, and the inflammatory infiltrate.

CONCLUSIONS

On the basis of the data, we conclude that in this lung injury model, (1) the distribution of the ventilating medium is uniform during LV when compared to GV, (2) LV improves gas exchange and pulmonary function, and (3) histologic evidence of lung injury is reduced after LV when compared to GV.