Twenty-five percent of the cardiac output flows through the liver with at least two thirds of this derived from the portal vein. This huge portal venous flow is driven through the liver across a minute pressure gradient. The pressure gradient between the portal inflow to the liver and the hepatic venous outflow from the liver is usually no more than 5 mmHg. Precise measurements must take into account the level of the tip of the recording catheter as the small gravitational differences can be significant. The resistance to blood flow through the portal vein is so low because of the unique hepatic vasculature, with conducting blood vessels terminating in each of the microvascular units of the acinus and flowing past only approximately 20 hepatocytes before exiting into the wide hepatic venules. The resistance is so low that at least 50% of the entire blood content of the liver can be expelled without adding significant vascular resistance.
For years, it was believed that the major site of resistance to portal blood flow was at the portal venous inlet vessels. By this assumption, sinusoidal pressure was better represented by hepatic venous outflow (or inferior vena caval) pressure than by portal pressure. However, in 1981, Greenway discussed the new data that changed his mind to the view that the primary site of portal vascular resistance at rest was across the hepatic veins and that portal venous pressure was a more accurate index of mean sinusoidal pressure. He demonstrated zones of resistance identifiable in the small hepatic veins through the use of catheters with the tips sealed and measuring blood pressure through side holes. We later used the same techniques and further validated Greenway’s conclusion. In addition, we demonstrated the extreme distensibility of these resistance sites and described the mathematical relationship between the distending pressure of the venous blood and the vascular resistance at the distensible resistance sites in both the hepatic vein and portal venule inlets.
A small change in vascular distending pressure resulted in a large change in the resistance that was defined by the equation R = 1/Pd3, where R is the resistance, calculated based on the pressure on either side of the distending tissue, divided by blood flow across the resistance site. Distending pressure (Pd) is calculated as the mean of the pressure above and below the resistance site. By this equation, a linear relationship is plotted between resistance and 1/Pd3 with the intercept passing through zero. The ope of the line is the index of contractility and is calculated as IC = R × Pd3. My 13-year-old daughter, Kelly, solved the equation (Figure 6.1).
The concept of IC allows for determination of whether changes in portal vascular resistance (or any venous resistance for that matter) are active, or passive secondary to altered distending pressures. The passive nature of the distensibility explains what we have referred to as portal pressure autoregulation. Autoregulation generally refers to regulation of blood flow, whereas, in this case, the reference is to regulating portal pressure. The liver cannot control portal blood flow, which is simply the total venous efflux from all of the splanchnic organs. Hepatic venous resistance is so low that even maximal stimulation of the hepatic sympathetic nerves, although resulting in a doubling or tripling of portal venous pressure, does not alter portal outflow from the splanchnic organs and therefore does not alter portal inflow to the liver. The liver must accommodate the entire portal flow and must accommodate to the quite large fluctuations that occur in portal blood flow, while maintaining its other homeostatic functions. The passively distensible portal venous and hepatic venous resistance sites explain how it is possible for portal blood flow to double while producing no more than a 2-mmHg change in portal pressure.
The resistance to blood flow through the liver is extremely low with pressure gradients between the portal venous inflow and hepatic venous outflow of the liver being in the range of 5 mmHg or less. Considering that the pressure gradients across all other organs are in the range of 115 mmHg, the vascular resistance within the hepatic portal system might erroneously appear to be trivial and of no consequence. In fact, the pressures are regulated at precise regions of the venous circuit at presinusoidal and postsinusoidal sites, and the proportion of pressure drop across the two sites and the total resistance determines intrahepatic and portal venous pressure and volumes. In chronic diseases of the liver, the most common cause of death is directly related to the vascular consequences of increased resistance to portal blood flow. If the pressure gradient becomes elevated by as little as 15 mmHg, hemodynamic and homeostatic instability can lead to the demise of the patient. Portal venous pressure reaches approximately 25 mmHg in the most severely cirrhotic conditions .
The adjustment of pressure in the normal state is achieved by changes of vascular resistance within the presinusoidal portal venules and the postsinusoidal hepatic veins. These sites of resistance have three important characteristics: they offer very low resistance; the resistance is able to be more than doubled by active vascular constriction; the resistance sites are passively distensible. To understand the venous responses to vasoactive stimuli, it must first be appreciated that the active responses interact with and are severely modulated by the passive distensibility of the resistance site. The ability to describe this interaction is dependent on being able to measure resistance at the key regulatory locations.
6.1. ESSENTIAL ASSUMPTIONS
The description of active and passive regulation of presinusoidal and postsinusoidal resistance sites is dependent on the assumption that the pressures measured are valid and not subject to significant artifact. The pressures of relevance are the portal venous pressure before entry into the liver, the central venous pressure (CVP) at the exit of the hepatic veins, and the intrahepatic pressure representing sinusoidal blood pressure. With these three pressures, and portal and hepatic blood flows, the presinusoidal and postsinusoidal resistance can be quantified. The portal and vena caval pressures are technically easily measured and validated because of ready access to these vessels under experimental conditions. The intrahepatic pressure is the contentious pressure.
As a pressure catheter is advanced via the vena cava into the hepatic veins, pressure is similar to CVP until the catheter tip passes through a narrow length of hepatic vein (Figure 6.2). This region has the characteristics of a smooth muscle sphincter, and morphological studies support the existence of sphincter-like regions in large hepatic veins . These sphincters are localized to third-order branches (ramuli, according to the nomenclature of Elias and Petty ) of the hepatic veins in cats  and in the terminal 2 cm of the lobar hepatic vein proper in dogs  (Figure 6.3). Physiological sphincters have not been proven in other species, and morphological data directly correlating functional and structural evidence are missing. The sudden rise of measured pressure from CVP to portal pressure as a catheter is advanced has usually been assumed to represent wedging of the catheter, and the pressure readings were assumed to be via a static column. Several factors indicate that this measurement is not a “wedged” pressure. First, the catheter need not be wedged and can often be advanced at least an additional 1–2 cm beyond the sphincter even in cats . Second, the catheters we have used have the tips sealed and pressure is recorded via side holes cut 3 and 7 mm back from the tip. We assume that the ability to record a valid pressure beyond the sphincter, but distal to the sealed catheter tip, is dependent on the existence of collaterals between the veins proximal to the sphincters. The sphincters are contracted by neural and pharmacological stimuli, and the site of the sphincter is not altered by catheter size or state of contraction [215,223,236] as would be expected if the “sphincter” really represented an artifact of wedging in the vein (Figures 6.2 and 6.4). Finally, in dogs, an unusual species in that the hepatic resistance site is located in the terminal portion of the hepatic vein, it is possible to position a catheter in the hepatic vein proximal to the sphincter site via an incision in the surface of the liver. By locating a small hepatic venous tributary and passing a catheter downstream into the hepatic vein (the catheter thus does not pass through the putative sphincter zone), pressure similar to portal pressure can be demonstrated (Figure 6.5). With the catheter thus placed, histamine-induced sphincter contraction led to equal and parallel elevations in portal venous pressure and in hepatic venous pressure proximal to the hepatic veins in a position where it is impossible to be measuring a “wedged” pressure . Other arguments have been presented [114,213,218] to support the view that measurement of lobar venous pressure (LVP) using a catheter that is passed proximal to the hepatic venous resistance site is representative of a true pressure measured at that site. Use of a balloon occluder in the large hepatic veins does, however, block a sufficiently large venous outflow that pressure escape via interconnected venous and sinusoidal pathways is not possible . In that case, the balloon occluder measures pressure resembling portal pressure .
Thus, LVP is assumed to be representative of sinusoidal pressure. The pressure gradient between LVP and CVP is the pressure gradient primarily regulated by a sphincter-like zone in the hepatic veins. In the basal state, this pressure gradient usually represents virtually the entire pressure gradient across the liver. Therefore, in the basal state, the vascular resistance of the portal venules and sinusoids is normally trivial with sinusoidal pressure and portal pressure being insignificantly different. The observation that equal reductions in arterial and portal flow lead to the same changes in hepatic volume  supports the contention that virtually all of the resistance to venous flow is in the hepatic veins. The pressure gradient between the portal vein (PVP) and the pressure (LVP) measured just proximal to the hepatic venous sphincters is assumed to represent primarily resistance within the portal venules. Note, however, that the pressure gradient represents pressure lost across the large and small portal tributaries, and the sinusoidal bed, as well as the small hepatic venules proximal to the LVP catheter. Active vasoconstriction results in a very significant increase in the PVP to LVP gradient.
I will refer to this pressure gradient as being due to presinusoidal or portal venous resistance. Using these assumptions to analyze the vascular responses to norepinephrine and nerve stimulation reveals some interesting characteristics. For example, in cats, activation of sympathetic nerves results in an initial vasoconstriction of the hepatic arteriole and portal venule sites, which gradually undergo vascular escape, in contrast to the responses of the hepatic venous resistance sites that actively contract and maintain contracted throughout the period of nerve stimulation. Vascular escape of the presinusoidal portal resistance vessel sites does not, however, occur in response to norepinephrine infusion in contrast to the escape seen in the hepatic artery . The observation of almost complete vascular escape of the active nerve-induced constriction in the portal vein indicates that a very small pressure gradient between the portal vein and hepatic vein existed after 5 min of stimulation. The reduction in hepatic blood volume, which is well maintained, obviously did not significantly elevate vascular resistance. This observation suggests that the sinusoidal cross-sectional area is so vast as to offer trivial resistance and that very large changes in this cross-sectional area can be achieved without adding significantly to overall resistance.
Changes in CVP are transmitted past the venous sphincter to the LVP according to the distensibility of the resistance site. Increasing the IC using norepinephrine results in a reduced transmission (Figure 6.6).
In summary, the principle assumptions here are that the LVP measurement is a valid pressure measurement representative of pressures proximal to the hepatic venous sphincter-like zone and that the PVP–LVP gradient represents largely presinusoidal or portal venous resistance and the LVP–CVP gradient represents primarily resistance across the hepatic venous sphincter-like zones.
6.2. PASSIVE DISTENSIBILITY
The large passive distensibility of the venous resistance sites has been quantified and studied under a variety of conditions [214,218,219]. Hepatic venous resistance is reduced by 75% in response to an elevation of vena caval pressure of roughly 3 mmHg . The small rise in distending pressure produces very large decreases in vascular resistance that has been quantified as having a constant relationship under basal condition and under conditions of actively increased basal tone. The distending pressure (Pd) is able to be estimated from the average value of the pressures measured on either side of the resistance site . Resistance (R) is inversely linearly related to the distending pressure cubed (Pd3) and the slope of this linear relationship is referred to as the “index of contractility” where IC = R × Pd3. Changes in this IC are affected acutely only by active vascular responses and are independent of passive alterations in the distending blood pressure. Thus, active vascular responses can be measured using the IC, whereas the use of calculated vascular resistance provides a measure that is the result of interaction between active and passive influences (see Chapter 11, Hepatic Nerves).
The utility of the IC can be seen when attempting to interpret the vascular responses to pharmaceutical compounds given to modify portal vascular resistance, for example, for treatment of portal hypertension. A compound such as propranolol that reduces portal pressure secondary to reductions of cardiac output and portal inflow actually result in an increase in calculated vascular resistance, thus giving the false impression that propranolol results in active vasoconstriction of the portal vessels. The calculated increase in resistance is, however, most likely explained entirely by the reduction in portal flow and subsequent portal pressure and passive recoil of the resistance sites. Using the IC, an increase in resistance with no change in IC offers a clear differentiation between active and passive responses. Similarly, pharmaceutical evaluation of compounds capable of dilating resistance sites in portacaval shunts, as a means of reducing portal pressure, is best assessed using both calculated resistance and IC.
Morgan & Claypool Life Sciences, San Rafael (CA)
Lautt WW. Hepatic Circulation: Physiology and Pathophysiology. San Rafael (CA): Morgan & Claypool Life Sciences; 2009. Chapter 6, Resistance in the Venous System.