THE HEPATIC CIRCULATION 



I 4 I 



partial hepatectomy suggesting the possibility that 

 neurovascular mechanism is invoked. Although the 

 contrast substance appeared to move more rapidly 

 into the hepatic veins with the "restricted" pattern, 

 no evidence of veno-venous shunting could be ad- 

 duced. Circulation time estimated in this way is not a 

 reliable guide to the actual velocity of the blood. Any 

 reduction in the volume of blood held in the vessels 

 would reduce the transit time without necessarily 

 affecting flow. The rise and fall of Thorotrast concen- 

 tration in the entering blood must also be taken into 

 account as well as the extent of dilution by arteriolar 

 inflow. Blood flow was not measured and it is impos- 

 sible to say whether the alteration in portal inflow was 

 associated with a compensatory change in hepatic 

 arteriolar resistance. The phenomenon suggests that 

 the peripheral sinusoids which can be examined 

 directly and in which intermittency has been observed 

 may be peculiarly susceptible to shifts in the pattern 

 of perfusion. It is probable that the same considera- 

 tions are applicable to intermittency in mesenteric 

 and splenic vessels. Determination of the distribution 

 of circulation (or transit) times across the splanchnic 

 bed in the dog (303) indicates that separate popula- 

 tions of path lengths (e.g., hepatic arterial, splenic, 

 and mesenteric channels) overlap markedly, each 

 possessing very short and very long routes. Hence the 

 effect of special distribution patterns within any one 

 circuit (such as the hepatic) would have little detect- 

 able influence upon the composition of draining blood. 

 The fact that BSP transfer remains relatively constant 

 over a wide range of flow suggests that parenchymal 

 cells are uniformly perfused under most circumstances 

 (3 5)- 



I 'iscosity 



The equation of vascular resistance with the num- 

 ber, cross section, and length of the arterioles alone 

 implies that blood flows freely without turbulence as 

 an ideal Newtonian fluid in accord with Poiseuille's 

 law. In reality, of course, blood is a highly complex 

 and heterogenous suspension of red cells in a colloidal 

 solution of proteins. Much evidence indicates that its 

 viscosity is altered by the character of the conduit, by 

 perfusing pressure, and by flow (202). Although there 

 is little reason to believe that critical velocities are 

 frequently exceeded in any portion of the splanchnic 

 vasculature, turbulence may be induced by respira- 

 tory and body movements which check the flow of 

 blood and give rise to transient vortices and eddies. 

 Turbulence may also arise during arterial pulsation 



with a tendency for the blood to move backward, even 

 in the capillaries, during diastole. Laminar flow re- 

 sults in inward movement of red cells and accumula- 

 tion about the axis presumably owing to nonuniform 

 distribution of the shearing force across the lumen of 

 the vessel and to shear rate dependence of plasma 

 viscosity (202, 301). However, this process appears 

 to be inconsistent, so that turbulence of a sort always 

 occurs and produces a "mixed flow." Both the cell- 

 free zone of plasma and the high velocity differential 

 next to the vessel walls permit "slippage" and result 

 in a lower than expected viscosity in vessels of small 

 diameters where the volume of "plasma-lining" is 

 proportionately larger. From this layer is derived the 

 plasma which enters capillaries by the process of 

 plasma-skimming observed in the hepatic sinusoids 

 and mesenteric capillaries by Knisely (185) and 

 others (225, 270, 299). Consequently, the blood 

 perfusing capillaries may vary widely in viscosity as 

 well as hematocrit with resultant irregularities that 

 are not readily resolved in hemodynamic analysis. 

 The development of turbulence under various cir- 

 cumstances and the effects of anomalous viscosity 

 complicate matters still more. 



With turbulence, a more complete admixture of 

 blood results. Thus the blood entering the arteries 

 from the heart has undergone a thorough stirring and 

 may be regarded as having a relatively uniform com- 

 position. Within the large veins, lamination results 

 in an unequal mixing of converging streams of vary- 

 ing composition so that "representative sampling" 

 from the inferior vena cava, for example, may be 

 difficult or impossible. Similarly, "layering" may oc- 

 cur in the portal vein and give rise to nonuniform 

 distribution within the liver, of blood coming from 

 the gastrointestinal, pancreatic, and splenic veins. 

 This possibility, first broached by Glenard in 1890, 

 was given experimental support by studies of Serege, 

 who found that India ink injected into the splenic 

 vein of dogs was carried preferentially to the left lobe 

 of the liver. Later, Bartlett et al. (24) found that 

 absorption of copper sulfate from the stomach and 

 duodenum of dogs resulted in deposition preponder- 

 antly in the left lobe and that absorption from the 

 ileum led to deposition in the right. Gopher & Dick 

 (95) observed "stream lines" in the canine portal vein 

 directly with transillumination following injection of 

 trypan blue into various portal tributaries. Perhaps 

 the most convincing evidence of "bilaterality of 

 portal flow" was reported in 1945 by Hahn et al. (168). 

 These workers injected radioactive phosphorus as 

 orthophosphate into the splenic vein, mesenteric 



