Abstract
As experimentally shown, pulmonary vasculature can be physiologically divided into extra-alveolar vessels and intra-alveolar vessels. The former are composed of pulmonary artery and vein running in the lung interstitium and the latter, capillaries exposed in alveolar space. They show entirely different geometrical responses to a change of lung volume, namely, an increase of cross-sectional area (a decrease of vascular resistance) in extra-alveolar vessels and an decrease (an increase of vascular resistance) in intra-alveolar vessels.
Arteriovenous fistula is a unique clinical condition to test the above-mentioned vascular physiology in man. This is because two vascular channels, a normal one composed of arterial, capillary and venous compartments and an abnormal composed of arterial and venous compartments lacking intra-alveolar vessels, run parallel in this pathological condition.
In this report, studies from this view point were carried out in 2 cases of pulmonary arteriovenous fistula. PaO2 under pure oxygen breathing changed, evidently due to a voluntary change of lung volume, the minimum in the TLC level and maximum in the RV level. Rather drastic, but identical PaO2 changes were observed during a passive change of lung volume under general anesthesia due to increases of airway pressure from atmospheric to 40 cmH2O. QS/QT was inversely proportional to these changes of PaO2. These results fit clearly with the behavior of geometrical changes of intra-alveolar and extra-alveolar vascular comparments during lung volume change experimentally proposed in animals.
Evaluations of the disease were simaltaneously carried out using pulmonary angiography, pulmonary artery occlusion test during right heart catheterization, perfusion lung scan with 99mTc MAA and echocardiogram with contrast media in these patients, but it is clear that dynamic analyses of PaO2 during these procedures have equivalent significance with these meticulous tests in diagnosis of pulmonary arteriovenous fistula.
