Pulmonary capillary transport function from flow-limited indicators.

The objective of this study was to examine the use of rapidly diffusing (flow-limited) indicators for estimating the pulmonary capillary blood volume (i.e., fraction of the lung blood volume wherein the diffusible indicators equilibrate with the tissue) and the capillary transit time distribution. Supporting theory and an application to experimental data are presented. The theory leads to the following equations, which relate the mean transit time (t), the variance (sigma 2), and the third central moment (m3) of the capillary transport function, hc(t), to the moments of the venous concentration-time curves for a vascular reference indicator, CR(t), and a flow-limited diffusible indicator, CD(t), after a bolus injection of the indicators upstream from an organ: sigma 2D - sigma 2R = ([1 + (te/tc)]2-1)sigma 2c and m3D-m3R = ([1 + (te/tc)]3-1)m3c, where te = tD - tR and tc is capillary t. The moments of hc(t) can be estimated if the injected bolus includes, along with the vascular reference indicator, at least two flow-limited diffusible indicators, each with a different te. A least-squares optimization procedure can then be used to specify the moments of hc(t). This approach was applied to isolated dog lung lobes with [14C]-diazepam as the diffusible indicator. The tissue-to-perfusate partition coefficient for [14C]diazepam could be adjusted to any desired value by altering the perfusate albumin concentration. Thus, by making a number of injections, each at a different perfusate albumin concentration, data were obtained in a manner equivalent to making one injection with a number of flow-limited diffusible indicators, each with a different te. On average, the estimated capillary volume and mean transit time were approximately 48% of the total lobar volume and mean transit time, and the relative dispersion of the hc(t) was approximately 75%.