Indicator dilution measurements of lung volumes and alveolar air exchange during breathing

Abstract

A new triple tracer indicator dilution technique has been used to measure alveolar ventilation as well as air and tissue volumes in the lungs of experimental animals and man. The tracers indocyanine green, [(121)I]antipyrine and xenon-133 were rapidly injected into the right atrium, while sampling was carried out from a peripheral artery. Blood flow and tissue volumes were obtained by classical analysis of the indocyanine green and antipyrine concentration-time curves. A double exit-port, constant air flow model was used to analyze the xenon curves, because ventilatory loss led to incomplete recovery of the gas tracer in effluent blood. Uniform ventilation and perfusion were assumed. This analysis permitted calculation of alveolar ventilation (VA(Xe)) and functional residual capacity (FRC(Xe)) during normal breathing. In control studies, VA(Xe) was similar to VA(co2), obtained with the steady-state CO(2) method (r = 0.87), while in critically ill patients the xenon measurement was significantly lower, averaging 54% of VA(co2). In theory, underestimates in VA(Xe) and decrease in the ratio VA(Xe)/VA(co2) relate to nonuniformity in regional ventilation and perfusion. The effect is greatest for the slightly soluble gas, xenon. The significant inverse correlation between VA(Xe)/VA(co2) and the physiologic shunt is consistent with this postulate.FRC(Xe) was similar to the predicted FRC in animals but was 76% of the helium measured FRC in patients. FRC(Xe) was significantly lower than the xenon measured air volumes during breath-holding when nonuniformity of ventilation was not operative. Lung tissue volumes in animals were 83% of gravimetric lung weights, while in patients the volumes were much lower than predicted. Nonhomogeneous lung function, including failure to perfuse the entire capillary bed, with resultant incomplete penetration of tracers into all segments of lung air and tissue, may explain these findings. The resultant errors can be significant in sick patients, and may themselves be used to study nonhomogeneities in the distribution of ventilation and volume.