Xenon handling in the liver: red cell capacity effect.

Xenon, despite its lack of chemical reactivity, associates preferentially with red cells in blood. To characterize the effect of this and the nature of xenon-tissue interaction in the liver, multiple indicator dilution studies were performed in the anesthetized normal dog through portal vein injection and hepatic vein collection of anaerobic blood samples. Two experimental runs were carried out in each animal, one at the prevailing hematocrit and the other at reduced hematocrit after bleeding and replacement with dextran. For comparison, the injection mixtures contained labeled red blood cells (a vascular reference), sucrose (an interstitial space reference), and labeled water (which freely enters liver cells), as well as labeled xenon. At the higher hematocrit, the labeled xenon curves generally rose earlier, peaked higher, and decayed more quickly than the labeled water curve; at the lower hematocrit, the xenon curve was delayed and diminished in magnitude in relation to the labeled water curves. Analysis of the curve shapes indicated that xenon, like labeled sucrose and water, underwent delayed wave flow-limited distribution. With knowledge of the red cell plasma partition coefficient (2.89 ml/ml), it was possible to both account for the change in form of the xenon curves with hematocrit and to use the data to estimate the liver cell tissue plasma xenon partition coefficient. Values averaged 1.93 ml/ml liver space, or 1.79 ml/g, and did not change significantly from first to second runs. Theoretical analysis indicated that flow cannot be estimated from xenon downslopes.