Microsphere passage through intestinal circulation: via shunts or capillaries?

Significant quantities of 9-micron microspheres (20-30%) are not trapped in the intestine following intracardiac or intra-arterial injection, but reach venous blood. Some investigators propose that the passage of 9-micron spheres measures blood flow through noncapillary connections. Because frequency distributions of intestinal capillary diameters and 9-micron spheres overlap, microspheres could simply pass through capillaries. Therefore, we developed simple probabilistic models to predict both the size distribution and the percentage of injected spheres [9 +/- 1 (SD) micron] that should appear in venous blood. Chief assumptions in models are that microsphere delivery and sphere diameter are independent and that microspheres pass through capillaries of equal or larger size. The passage predicted by the models was consistent with values in canine intestinal circulation, demonstrating that passage through capillaries [7.38 +/- 1.4 (SD) micron] adequately accounts for spheres in venous blood. Because the diameters of nominal 9-micron spheres were distributed too narrowly to show a marked sieving effect on passage through the intestinal circulation, we also injected microspheres varying from 5 to 20 micron in diameter. This mixture demonstrated a marked sieving effect. The predicted frequency distribution for microsphere diameters in venous blood agreed with the observed distribution. Our models demonstrate that the passage of 9-micron spheres through capillaries, rather than through "shunts," adequately accounts for the appearance of spheres in venous blood and suggests that the frequency distribution of venous microspheres can provide an in vivo method for estimating the frequency distribution of intestinal capillary diameters.