Correlation of oxygenation with vascular permeability-surface area but not with lung water in humans with acute respiratory failure and pulmonary edema

Abstract

We used a single-pass multiple tracer technique to measure cardiac output, extravascular lung water (EVLW) and lung vascular [14C]urea permeability-surface area (PS(u)) in 14 patients with acute respiratory failure and pulmonary edema. All patients had increased EVLW, but EVLW in the 10 surviving patients (0.26 ± 0.06 SE ml/ml total lung capacity [TLC] was not significantly different from that in the five patients who died (0.22 ± 0.05). EVLW did not correlate with inravascular pressures or with alveolar-arterial oxygen pressure difference (A-aDO2). PS(u) was lower in surviving patients (0.50 ± 0.16 SE ml/s x liter TLC) than in patients who died (3.44 ± 0.36; P < 0.05) and also lower than in previously reported data in patients with normal PS(u). PS(u) correlated significantly with A-aADO2. Serial studies showed that PS(u) returned from a low value toward normal in a patient who survived but remained high in a patient who died. We conclude that the amount of edema in the lungs measured by indicator methods was not the principal determinant of either the magnitude of oxygenation defect or survival in the patients studied. We interpret the low PS(u) in surviving patients as decreased surface area and infer that the ability of the lung circulation to reduce perfusion of damaged and edematous areas was important in preserving oxygenation. A high PS(u), presumably reflecting perfusion of areas with increased permeability, was a sign of especially poor prognosis. Multiple tracer techniques for measuring lung vascular PS(u) may help to define the pathogenesis and to evaluate therapies of acute lung injury in humans. Such measurements may be a more useful clinical tool than measurements of lung water in patients with acute respiratory failure and pulmonary edema.