[Impairment of gas exchange in acute lung injury]

To demonstrate the characteristics of impaired gas exchange in acute respiratory distress syndrome (ARDS), an experimental model of lung injury was produced in 24 mongrel dogs by intravenously injecting 0.07 ml/kg of oleic acid. While allowing the animal to breathe a mixture of 0.1% CO in air, normal saline, which contained appropriate amounts of six inert gases such as SF6, ethane, cyclopropane, halothane, ether and acetone, was infused at a constant rate through a peripheral vein. After a steady state was established, the expired gas was collected and the samples of both arterial and mixed venous blood were taken simultaneously. The concentrations of the nine indicator gases (O2, CO2, CO and six inert gases) in the samples were measured in terms of a gas chromatograph, permitting analysis of the distribution of ventilation to perfusion (VA/Q) as well as the diffusing capacity to perfusion (G/Q) in injured lungs. To determine the role of hypoxic pulmonary vasoconstriction (HPV) in maintaining gas exchange in ARDS, hemodynamic and gas-exchange parameters were investigated at inspired O2 concentrations (FIO2) of either 21 or 60%. The impairment of gas exchange was examined by measuring the fractional retention (R) of the inert gases in arterial blood. Furthermore, to assess the possible contribution of vasoactive prostanoids in regulating vascular reactivity in ARDS, observations at FIO2 of 60% were repeated after administered indomethacin at a dose of 5 mg/kg. Analytical results revealed that shunt flow in experimental dogs with lung damage caused by oleic acid averaged 17%. Furthermore, widening of VA/Q distribution was found accompanied with significant contribution of extremely low VA/Q areas. In addition, most of the lung was operating in G/Q units with values ranging from 10(-3) to 10(-2) but 9.8% of total Q(QT) was received by the area with G/Q less than 10(-3), which might limit O2 exchange between the alveolar gas phase and capillary blood. Although pulmonary vascular resistance (PVR) in injured lungs observed at FIO2 of 60% was significantly smaller than the value obtained at FIO2 of 21%, QT as well as extravascular lung water did not differ in the two conditions. The R value for the indicator gas was consistently lower at FIO2 of irrespective of gas species. Administration of indomethacin caused a considerable diminution of the R value for inert gas but a rise in PVR without an appreciable change of either QT or extravascular lung water. This was followed by a significant rise in arterial PO2 from 84 to 99 Torr.(ABSTRACT TRUNCATED AT 400 WORDS)