3.1. FLOW-LIMITED DISTRIBUTION OF BLOOD-BORNE SUBSTANCES

The “multiple indicator dilution technique” consists of determining the hepatic venous efflux after rapid intraportal injection of a reference substance, whose behavior is known, and one or more other substances, whose behavior is to be characterized. One of the pioneers who used the technique for studies in the liver was Carl Goresky. Carl contributed a chapter in the 1981 symposium that provides a theoretical basis for the method and its interpretation [97]. For this method, red blood cells were usually used as the reference curve. A labeled red cell travels faster than a bolus of labeled albumin. The difference in rate of efflux from the liver represents different accessibility to the hepatic spaces. Access to intracellular space is restricted according to molecular weight. The restriction by molecular size demonstrated by the indicator dilution outflow curves from the liver is compatible with the completely different approach of demonstrating physical size limitation to hepatic lymph outflow [12].

On the basis of a series of studies with multiple-indicator dilution techniques, Goresky [98] confirmed that no continuous anatomic barrier exists between plasma and the space of Disse. The endothelial lining serves to contain red cells, but virtually immediate lateral diffusion equilibrium is expected for most molecules within the space of Disse. As blood cells squeeze through the sinusoids, they massage the endothelial cells and further mix plasma and Disse fluid.

Sinusoids in zone 1 of the acinus form richly anastomotic channels, whereas sinusoids in zone 3 are more radially arranged. There are fenestrations throughout the length of the sinusoids composed of single large 3-µm holes and small clusters of 1-μm holes surrounded by microfilaments [128]. There are no lymphatics associated with the sinusoids. Lymph vessels appear to arise in the space of Mall around the portal tract, forming extensive lymphatic plexuses that anastomose with other lymphatic plexuses on the surface of the liver underneath Glisson’s capsule. These anastomoses thus provide a lymphatic pathway from deep in the liver to the surface. Other lymphatic plexuses are found around the larger hepatic veins [30,57]. When formation of lymph exceeds the capacity of the major lymph trunks, exudation of lymph on the surface of the liver occurs [30,31,110,141].

Morphological studies suggest that the sinusoidal lumen is 10.6% of liver volume, the space of Disse is 5%, and the bile canaliculi comprise 0.4% of volume [24]. The picture that emerges from this anatomy is compatible with the physiological data on extracellular spaces in the liver. Sodium (8.9%), sucrose (8.8%), and inulin (7.7%) spaces were almost equal. Albumin extravascular space was 5.7% of liver weight and 64% of the Na⁺ space (98). The albumin space was 59% of the interstitial space, and this increased to 66% when hepatic venous pressure was raised. The lymph-to-plasma concentration ratio varied only slightly with molecular size and was 1.0 for lactoglobulin, 0.88 for albumin, and 0.69 for γ-globulin. Protein content of lymph was 80–95% that of plasma, and hepatic lymph proteins originated from blood, not from new synthesis [30,71,101,122,180].

These data indicate that 60% of the interstitial space is accessible to albumin, and this presumably includes the spaces of Disse, the spaces of Mall, and the lymphatics. Lymph appears to be formed by filtration of high-protein fluid from the spaces of Disse across the limiting plate into the lymphatics in the spaces of Mall, and across the limiting plate into the tissue spaces and lymphatics around the hepatic veins. The limiting plate appears to represent a very minor barrier to passage of proteins, depending on their molecular weights [71,101,180].

3.2. ASCITES FORMATION

Ascites is a collection of fluid in the peritoneal cavity that represents an imbalance between the rate of fluid filtration into the peritoneal space and the rate of reabsorption from that space. The liver does not become edematous. Fluids filtered out of the sinusoidal space rapidly exit with apparently minor restriction. There is no evidence that fluid reabsorption from the peritoneal space through the liver occurs, nor do the characteristics of the vasculature suggest this would be possible.

Zink and Greenway [386] studied the reabsorption of fluid from the peritoneal cavity by encasing the abdomen in a rigid plaster cast to form an abdominal plethysmograph. The rate of fluid absorption from the peritoneal cavity was directly proportional to the intraperitoneal pressure regardless of whether the intraperitoneal fluid was free from protein or contained a protein concentration equivalent to that of plasma. The relationship between absorption of fluid from the peritoneal space and intraperitoneal pressure was approximately linear with a rate of 0.01 ml min^–1 mmHg^–1 per kilogram of body weight or 0.04 ml min^–1 mmHg^–1 per 100 g of liver. This contrasts with the filtration rate out of the liver of 0.08 ml min^–1 mmHg^–1 per 100 g of liver [103]. Thus, removal from the peritoneal cavity per unit pressure appears to be substantially slower than formation, and this may explain the occurrence of ascites in some pathological situations. Further studies with ¹³¹I-labeled albumin showed that protein was absorbed from the peritoneal cavity in equal proportion to the absorption of fluid, and the fractional rates of protein absorption were never significantly different from the fractional rates of fluid absorption. Both fractional rates were independent of the protein concentration in the peritoneal cavity, and this indicated that the removal process involved lymphatic absorption rather than transcapillary absorption [261].

It is very difficult to reverse the hydrostatic pressure gradient across the small hepatic vessels because raising intraperitoneal pressure compresses the large veins. Thus, sinusoidal pressure increases by the same amount as extravascular pressure [385]. It appears that protection of the liver against edema is achieved by lymphatic drainage and transudation across the capsule rather than by mechanisms that limit filtration or that allow reabsorption of the fluid into the hepatic vessels [110,180].

3.3. EFFECTS OF DRUGS ON FLUID EXCHANGE

When hepatic venous pressure was mildly elevated to produce a steady-state filtration across the liver, infusions of epinephrine, isoproterenol, and histamine had no effect on the steady-state filtration, and it was concluded that these drugs did not modify either surface area or permeability within the liver [111]. There are clear species differences in this regard as histamine is capable of stimulating vasoconstriction in the hepatic veins of dogs, thereby increasing intrahepatic pressure and fluid filtration [19] (also see Figure 6.5 in Chapter 6).

3.4. EFFECTS OF HEPATIC NERVE STIMULATION

Some of the earlier responses of fluid exchange in response to vasoactive stimuli were misinterpreted based on the assumption that central venous pressure was an estimate for intrahepatic sinusoidal pressure. Vasoconstriction resulting in elevated portal venous pressure was thought to primarily occur at the presinusoidal portal vascular resistance vessels. Vasoconstrictors have since been shown to have the primary effect in small hepatic venules so that portal venous pressure is more reflective of intrahepatic sinusoidal pressure (Chapter 6).

Regardless of the proportion of the vasoconstriction that occurs at presinusoidal or post-sinusoidal sites, stimulation of the hepatic sympathetic nerve branch will result in some degree of increase in sinusoidal pressure. Therefore, one should expect to see an increase in fluid filtration from the liver. This is not seen. In fact, if a background filtration is established by elevating hepatic venous outflow pressure to produce a portal pressure of 8.7 mmHg, hepatic nerve stimulation at 2, 4, and 8 Hz results in a frequency-dependent decrease in filtration rate. If hepatic venous pressure is then elevated to levels of 12 and 16 mmHg, the ability of nerve stimulation to decrease the large filtration rate is overwhelmed but the trends still remain, indicating that the nerves result in reduced fluid filtration even in the face of elevated sinusoidal pressure [105]. These studies suggest that the hepatic sympathetic nerve stimulation reduced fenestration size and impaired fluid filtration out of the plasma compartment.

Even if the entire vasoconstrictor response to the nerves was at the presinsuoidal site, sinusoidal pressure would not decrease to reduce fluid filtration. Hepatic blood flow is neither significantly redistributed nor heterogeneous when hepatic nerves are stimulated (Chapter 11). Therefore, heterogeneity of flow cannot account for the ability of the hepatic nerves to decrease fluid filtration. This observation requires further evaluation as it may have significant consequences for liver metabolism in situations of activated sympathetic nerves. Furthermore, if the endothelial fenestrations can be shown to be regulated by constrictor influences, then dilator stimuli should be identified for evaluation of utility in diseased livers where sinusoidal fenestrations become decreased in number and size. Access of lipoproteins transporting triglycerides from the liver (very low-density lipoprotein) and cholesterol to the liver (low-density lipoprotein) may be manipulated by regulating fenestrae size.

3.5. BLOOD FLOW AND HEPATIC CLEARANCE OF DRUGS AND HORMONES

As a general rule (with many important exceptions), clearances of substances through any organ have certain kinetic limitations. Generally, if a substance has a very high extraction ratio passing through the liver, a reduction in hepatic blood flow will lead to a similar reduction in hepatic clearance of that substance. That can have major homeostatic consequences. If substances have a very low hepatic clearance rate, it is assumed that the rate-limiting step is not the delivery of the compound to the extracting hepatocytes but rather a rate-limiting control within the processing cells. This simplistic pharmacokinetic model cannot be relied upon. For example, we showed that hepatic extraction of lidocaine in the cat was only 28%. At that moderate level of hepatic extraction, classic pharmacokinetic theory at that time would have predicted a very significant impact on hepatic clearance as a result of a reduction in hepatic blood flow. However, as hepatic blood flow decreased, the hepatic extraction ratio for lidocaine did not change significantly and lidocaine clearance decreased in parallel with blood flow (Figure 3.2).

FIGURE 3.2

The effect of stepwise reduction in hepatic blood flow on lidocaine clearance rate. Control extraction ratio of 30% was not significantly altered by reduced flow. Lidocaine clearance is linearly related to blood flow in the cat. Reproduced with permission (more...)

In summary, the hepatic lymphatic system is poorly understood in terms of function, regulation, and role. The pathway by which fluid, filtered from the plasma space, enters both the lymphatics that drain into the thoracic duct and into lymphatics, through which fluid exudates across Glisson’s capsule, is unclear. The role of the lymphatic system in other tissues does not appear to be of significance in the liver. The Starling forces of hydrostatic pressure and colloid osmotic pressure acting across the endothelial cells to regulate fluid exchange between the extracellular fluid compartments of plasma and interstitial fluid seems irrelevant to the liver because the equivalent of the extracellular interstitial fluid, the space of Disse, appears to have a protein content and hydrostatic pressure level indistinguishable from that of the plasma compartment. Although these observations may suggest a trivial role for the hepatic lymphatics, the huge contribution of hepatic lymph flow to total lymph flow and the large volume of hepatic lymph that is formed per day suggest that our knowledge rather than the function of the lymphatics is insignificant. I cannot update Child’s conclusion in 1954 that “At the moment, then, it can safely be stated that the last word has not been written on the hepatic lymphatics.”