PDF
Citation
Print
INTRODUCTION
Hyperdynamic splanchnic circulation is commonly known as a complication of liver cirrhosis. However, strictly, it is a complication of portal hypertension, which develops due to liver cirrhosis. The hyperdynamic splanchnic circulation persists for more than 28 days even after the replacement of a cirrhotic native liver with a new liver graft [1], due to a decrease in the resistance of the vascular bed in the liver graft, pre-existing portal hypertension, and splanchnic vasodilatation [2]. The persistent hyperdynamic splanchnic circulation leads to an increase in portal venous blood flow (PVF) to the liver graft, which is termed “portal hyperperfusion,” and the resulting intrahepatic shear stress stimulates and regulates hepatic regeneration [3,4]. However, in living donor liver transplantation (LDLT) using a partial liver graft, the vascular bed does not always have enough capacity to accommodate the increased PVF. Although the increased PVF plays a crucial role in regeneration of the partial liver graft [1], an excessive increase in the PVF could cause failure of the liver graft to control directly the PVF, eventually threatening the viability of the graft [5] with the development of small-for-size syndrome (SFSS). This review will deal with the two conflicting results caused by the shear stress secondary to portal hyperperfusion, which is inflicted on the sinusoidal bed of a partial graft (hepatic regeneration vs. portal hyperperfusion injury) as well as the underlying physiology and pathophysiology of hepatic hemodynamics, knowledge of which contributes to a better understanding of the effects of portal hyperperfusion in a partial graft.
PHYSIOLOGY OF HEPATIC HEMODYNAMICS
Serving as a major blood volume reservoir in the human body, the liver receives blood at 800–1,200 ml/min, which is equivalent to 100–130 ml/min per 100 g of liver weight [6] and constitutes approximately 25% of cardiac output (1–2 L/min) [7]. The liver contains blood at 25–30 ml/100 g of liver weight – accounting for 10–15% of the total blood volume [8] – although it constitutes only 2.5% or 33 g/kg of total body weight [9,10]. Large capacitance vessels, such as the portal and hepatic veins and hepatic artery, hold 40% of the blood that the liver receives; the sinusoids hold the remaining 60% [6]. The hepatic blood volume can compensate for up to 25% of blood loss [11].
The liver has a unique dual blood supply system consisting of the hepatic artery, which carries approximately 25% of the total hepatic blood flow (30 ml/min per 100 g of liver weight; with well-oxygenated blood [oxygen saturation of 95%] providing 30–50% of the liver oxygen requirement); and the portal vein, which receives all the blood from the spleen, stomach, small and large intestines, gall bladder, and pancreas [12], and provides partially deoxygenated blood (oxygen saturation of up to 85% during the fasting state) responsible for 75–80% of the total blood flow to the liver (90 ml/min per 100 g of liver weight) and 50–70% of the liver oxygen requirement. The two vessels form the portal triad together with the bile duct. In the resting state, the liver consumes approximately 20% of the total oxygen used in the body [13]. With increased oxygen demand, such as isovolemic hemodilution or upregulation of hepatic enzymes, more oxygen is simply extracted from the blood, but the hepatic artery does not dilate [14]. The saturation of hepatic venous blood, which is normally two-thirds saturated with oxygen, would be reduced upon low delivery of oxygen to the liver, leading to an increase in oxygen extraction from the hepatocytes. The portal venous system is protected from hepatic arterial pressure by a reduction in the hepatic arterial pressure towards portal venous pressure (PVP), due to hepatic arteriolar resistance in combination with the intermittent closure of the presinusoidal arterioles in the peribiliary plexus [15].
Mesenteric and splanchnic arteriolar vascular tone, as well as intrahepatic vascular resistance, regulate the valveless portal vein system, maintaining low pressure and resistance. The normal range of PVP is between 5 and 10 mmHg [16]. Due to the large capacity of the sinusoidal structure and intrahepatic vasculature, which easily accommodates additional PVF by releasing nitric oxide (NO) that dilates intrahepatic vessels [17], an increase in PVF does not influence the PVP [18]. The mean hepatic arterial pressure is similar to that of the aorta [19]. The high pressure and resistance hepatic artery system [6] is regulated by classical arterial autoregulation and interacts with the portal vein system via the hepatic arterial buffer response (HABR), which maintains overall hepatic blood flow and an adequate oxygen supply to the liver by compensating for diminished or increased PVF with a reciprocal increase or decrease in hepatic arterial blood flow (HAF) [20]. The compensatory interaction between the hepatic artery and portal vein systems will be discussed in the following section. The both afferent vessels enter the hepatic lobule and merge at the sinusoidal bed, which is drained by the hepatic venous system into the inferior vena cava. Sinusoid pressure or hepatic venous wedge pressure, which may be an indicator of PVP, ranging from 3 to 10 mmHg is between PVP (5–10 mmHg) and inferior vena cava pressure (1–2 mmHg) [19].
REGULATION OF HEPATIC BLOOD FLOW BY THE HEPATIC ARTERIAL BUFFER RESPONSE
In addition to classical arterial autoregulation, in which the HAF is intrinsically regulated via the myogenic response of the hepatic artery to arterial pressure, there is another mechanism for the intrinsic regulation of the HAF, namely HABR. Although an increase in HAF in response to reduced PVF was first reported in 1911, the relationship between the two vascular systems was termed HABR for the first time in 1981 [19]. The hepatic artery dilates or constricts secondarily to a decrease or increase in the PVF, respectively [20]. Temporary occlusion of the portal vein induced a sharp and significant increase in the HAF, by approximately 30% in anesthetized patients with carcinoma of the splanchnic area [20]. The increase in the HAF compensates for 25–60% of the decreased PVF [21]. The hepatic arterial response alleviates the effects of changes in the PVF on hepatic clearance and maintains oxygen supply to the liver [22]. However, HABR is not reciprocal; that is, PVF or portal venous resistance does not change in response to changes in hepatic arterial perfusion [20] because the PVF passively depends on the outflow of the extrahepatic splanchnic organs. Under physiological conditions, HABR operates at all ages [19], even prenatally [23].
The adenosine washout hypothesis well explains the HABR. Adenosine, which is a potent hepatic arterial vasodilator [21], is released at a constant rate into the fluid in the space of Mall, which is a space between the periportal hepatocytes and portal connective tissue surrounding the hepatic arterioles and the portal venules. The concentration of adenosine is dependent on the rate of washout into the portal vein. A reduction in washout of adenosine from the space of Mall secondary to the decreased PVF results in dilation of the hepatic artery, thereby leading to an increase in the HAF [14]. The mediation of adenosine in HABR is supported by several lines of evidence, which show the potentiation of HABR by adenosine uptake inhibitors [21], competitive blockade of HABR by adenosine antagonists [24-27], and a lower effect of adenosine infused into the portal vein compared with that infused into the hepatic artery (thus indicating communication between the two vascular systems) [14]. However, adenosine dose not diffuse from the portal venous vasculature to the hepatic arterial vasculature [28]. The substance that diffuses from the portal venous to the hepatic arterial bed in response to hypoxia [29], caused by a reduction in the PVF [30], is adenosine-5’-triphosphate (ATP), rather than adenosine itself [28]. Only the adenosine produced from ATP catabolism in the hepatic arterial vasculature contributes to hepatic arterial dilation [27], whereas the adenosine created in the portal venous vasculature is rapidly taken up by the endothelium and vascular smooth muscle immediately after its production [31].
PATHOPHYSIOLOGY OF PORTAL HYPERTENSION AND SUBSEQUENT HYPERDYNAMIC SPLANCHNIC CIRCULATION IN LIVER CIRRHOSIS
Portal hypertension is defined as a sustained increase, by more than 5 mmHg, in the hepatic venous pressure gradient (HVPG) [32]. The HVPG represents the difference between hepatic venous wedge pressure reflecting sinusoidal pressure, which provides an accurate estimate of PVP due to loss of normal intersinusoidal communications in cirrhotic patients, and free hepatic venous pressure [33]. However, an HVPG between 5 and 10 mmHg presents with no clinical manifestations and is considered to be subclinical portal hypertension. A diagnosis of clinically significant portal hypertension is made when the HVPG reaches more than 10 mmHg, which predicts the development of complications of liver cirrhosis, including ascites, portosystemic collaterals, varices, circulatory dysfunction, and even death [34]. The threshold of the HVPG for variceal hemorrhage is 12 mmHg [35]. In addition, liver cirrhosis patients with HVPG greater than 20 mmHg following variceal bleeding have a five-fold increase in the risk of death [36].
In portal hypertension, intrahepatic vascular resistance increases not only upon mechanical obstruction caused by architectural changes in the liver (fibrosis and capillarization of the sinusoids, development of microthrombi in the intrahepatic vasculature, and regenerative nodule formation) [37], but also upon the dynamic obstruction that results from impaired sinusoidal relaxation due to overproduction of inflammatory mediators. The overproduction in turn results in oxidative stress, which causes sinusoidal endothelial cell dysfunction thus leading to overproduction of, and enhanced sensitivity to, vasoconstrictors such as endothelin-1, angiotensin II, leukotrienes, and norepinephrine, coupled with underproduction of vasodilators such as NO, carbon monoxide, and prostaglandin E2 [18]. Activated hepatic stellate cells differentiating into contractile fibrogenic myofibroblasts also increase intrahepatic vascular resistance by producing an excessive amount of extracellular matrix, inflammatory cytokines, and endothelin-1. Vasoconstrictive substances, such as endothelin-1, reduce the sinusoidal space and increase intrahepatic vascular resistance by inducing hepatic stellate cell contraction [37].
In contrast to vasoconstriction, which occurs in the intrahepatic vasculature, the splanchnic vasculature experiences progressive vasodilatation during the progression of liver cirrhosis. However, in the early stage of cirrhosis, when portal hypertension develops due to an increase in intrahepatic vascular resistance in the absence of extensive collateral circulation, the splanchnic circulation is normodynamic or hypodynamic, rather than hyperdynamic [38]. Consequently, a reduction in the PVF with a reactive increase in the PVP ensues [39]. This physiological alteration constitutes the main postulation of the “backward flow” theory. As liver cirrhosis progresses to an advanced stage, the consequent physiological alteration would be explained by the “forward flow” theory, in which the splanchnic circulation becomes hyperdynamic due to portal hypertension-induced splanchnic vasodilatation with subsequent development of extensive collaterals [40]. Under this physiological milieu, the decreased PVF during the early stage of liver cirrhosis increases again [41].
NO, which causes vasodilatation by the activation of soluble guanylyl cyclase, in turn leading to the generation of cyclic guanosine monophosphate in the vascular smooth muscle [42], is the most important vasodilator associated with portal hypertension and subsequent hyperdynamic splanchnic circulation. There are three isoforms of NO synthase (NOS) that synthesize vascular NO: endothelial NOS (eNOS) [43], inducible NOS (iNOS) [44], and neuronal NOS (nNOS) [45]. Among these isoforms, eNOS, which produces NO in the endothelial cell in response to mechanical forces caused by sheer stress, endothelial growth factor, or inflammatory cytokines [46], plays a major role in the overproduction of vascular NO in the splanchnic arterial circulation [47]. In addition, an increase in intestinal absorption of lipopolysaccharides, due to changes in intestinal permeability, activates iNOS by releasing inflammatory cytokines, such as tumor necrosis factor-α, from macrophages [48], which contributes to hyperdynamic splanchnic circulation. nNOS, which is synthesized in neuronal and vascular smooth muscle cells, was also found to be upregulated in the mesenteric artery [49] and in the aorta [50], resulting in the development and maintenance of hyperdynamic splanchnic circulation in liver cirrhosis [50]. In contrast, intrahepatic NO decreases because of reduced activity of eNOS in portal hypertension [37].
In response to splanchnic and peripheral vasodilatation, retention of sodium and water occurs, resulting in the expansion of the plasma volume, a large part of which accounts for the portosystemic collateral circulation [51]. The resulting increase in venous return to the heart leads to an increase in cardiac output [51]. Despite the essential role of vasodilatation, the systemic hyperdynamic circulation cannot be maintained in the absence of the portosystemic shunt, which contributes to the expansion of plasma volume [52]. A cardiac index higher than normal (> 4 L/min/m2) often does not sufficiently compensate for the decreased arterial pressure caused by progressive vasodilatation. The splanchnic and systemic hemodynamic derangement in cirrhotic patients is reversible by liver transplantation [53].
HEPATIC REGENERATION INDUCED BY PORTAL HYPERPERFUSION
Upon replacement of the native cirrhotic liver with a new liver graft from a living donor, the partial liver graft should receive a large amount of PVF from the hyperdynamic splanchnic circulation such that liver cells, including Kupffer cells and sinusoidal endothelial cells, are exposed to excessive hemodynamic forces caused by PVF against the vessel walls. An increase in sinusoidal blood flow to the partial liver graft inflicts shear stress on the endothelial surface, which stimulates hepatic regeneration [54]. Shear stress is defined as a viscous drag at the surface of the sinusoidal endothelial cells, which is created by adjacent blood flow [55]; its amount is proportional to the volume of the blood flow and the inverse of the cube of the vessel radius [55]. However, excessive shear stress may induce liver failure [54].
Because the sinusoidal domain of the hepatocyte faces the perisinusoidal space of Disse, which is the space between the hepatocytes and endothelial sinusoidal-lining cells (with fenestrae into the sinusoidal lumen forming the sieve plate structure), the hepatocyte may be exposed directly to the PVP through the fenestrae. Thus, not only the sinusoidal endothelial cells, but also the hepatocytes, respond to PVP-induced shear stress [56]. The shear stress arising from alteration of the hemodynamic pattern of hepatic blood flow initiates signal cascades that trigger hepatic regeneration. The existence of shear stress-responsive cell surface modulators on the hepatocytes and sinusoidal endothelial cells, such as the calcium ionic channel, sodium ionic channel [57], and gap junctions [58], has been suggested. Electron microscopic observations showed widening of sinusoidal endothelial fenestrations and spaces of Disse, contributing to an increase in sinusoidal endothelial permeability to circulating hepatotrophic substances and enlargement of the intercellular spaces. This in turn leads to extracellular matrix degradation, with subsequent induction of responsiveness of hepatocytes in the periportal area to circulating and liver-borne growth factors [59]. In addition, ATP, which enhances cell cycle progression and hepatocyte proliferation [60], is released from isolated rat hepatocytes and hepatectomized livers of rats and human living donors by mechanical stress [61,62]. The rapid loss of adenine nucleotides after partial hepatectomy also generates early stress signals to the remnant liver, contributing to the onset of hepatic regeneration [63].
Interestingly, portal hypertension does not induce hepatic regeneration in liver cirrhosis. As liver cirrhosis progresses, the sinusoidal endothelial cells become capillarized [64], thereby leading to disappearance of the sieve plate structure. The absence of the sieve plate structure hinders any direct influence of portal hypertension-induced shear stress on the hepatocytes, thereby preventing the initiation of hepatic regeneration. However, hepatocytes isolated from cirrhotic livers have the potential to proliferate [56].
Within the normal liver, sinusoidal endothelial cells express eNOS that produces NO, the amount of which is dependent on flow [65]. An increase in endothelial shear stress modulates the release of endothelial NO and prostaglandins [66,67] and increased PVP immediately after partial hepatectomy activates eNOS and upregulates iNOS [3]. In contrast to adenosine, which is dependent on hepatic blood flow, NO acts in response to vasoconstriction-induced shear stress [11]. Accordingly, NO produced in response to an increase in shear stress affects vascular accommodation following a partial hepatectomy [68] and provides the residual liver with permeability to growth factors, contributing to triggering of the hepatic regeneration cascade [3,66,67]. In addition, NO downregulates the level of S-adenosylmethionine – the synthesis of which is essential for methionine metabolism – causing hepatocytes to respond to hepatocyte growth factor [69].
If the graft-to-recipient weight ratio (GRWR) is > 0.8, which is known to be the minimum size of a liver graft to fulfill the metabolic demand of a recipient [70], in the absence of any occlusion of hepatic venous outflow, SFSS rarely develops. For liver grafts uncomplicated by SFSS, portal venous hemodynamics have a major impact on hepatic regeneration. The portal venous velocity and PVF increased considerably after the reperfusion of a liver graft and then returned to the baseline value, which was measured before hepatic parenchymal transection in donors, at 3 months after LDLT [71]. The size of the liver graft is restored to the standard liver volume of recipients in 2 weeks after LDLT [1]. Many previous studies have found correlations between PVP or PVF and the degree of graft regeneration, which were measured at different postoperative time points [1,72-74] (Table 1). However, there is still no consensus about which portal venous hemodynamic parameter is of paramount importance in predicting the degree of graft regeneration.
Table 1
Correlations between Portal Venous Hemodynamic Parameters and the Degree of Hepatic Regeneration after Living Donor Liver Transplantation
| Reference | Number of patients | Etiology of liver disease | Portal venous hemodynamic parameters | Parameters representing the degree of hepatic regeneration | Statistical method | Correlation coefficient (r) or coefficient of determination (r2) | P |
|---|---|---|---|---|---|---|---|
| Eguchi et al. (2003) [1] | 15 | Fulminant hepatic failure (n = 4), liver cirrhosis (n = 9) | Mean portal venous velocities measured on the 7th postoperative day (cm/s) | The proportion of the liver graft to standard liver volume at postoperative month 1 (liver volume ratio) (%) | Simple linear regression | r = 0.666* | 0.0181 |
| Mean portal venous velocities measured on the 1st and 28th postoperative days (cm/s) | Liver volume ratio at postoperative month 1 (%) | UA | < 0.05 | ||||
| Mean portal venous velocities measured on the 1st and 7th postoperative days (cm/s) | Liver volume ratio at postoperative month 3 (%) | UA | < 0.05 | ||||
| Mean portal venous velocities measured on the 7th postoperative day (cm/s) | Liver volume ratio at postoperative week 1 or 2 (%) | UA | < 0.05 | ||||
| García-Valdecasas et al. (2003) [72] | 22 | Liver cirrhosis | Difference in portal venous blood flow per 100 g of liver weight between recipient and donor (ml/min/100 g) | Percent change in liver graft weight between time of graft harvest and postoperative month 2 (%) | Spearman rank correlation | r2 = 0.2343† | < 0.05 |
| Park et al. (2008) [73] | 31 | Fulminant hepatic failure (n = 3), liver cirrhosis (n = 28) | Portal venous velocity per 100 g of liver graft at postoperative day 1 (cm/s) | Ratio of liver graft volume at postoperative day 7 to graft weight measured immediately after graft retrieval (%) | Simple linear regression | r2 = 0.273 | 0.009 |
| Portal venous blood flow per 100 g of liver graft at postoperative day 1 (ml/min/100 g) | r2 = 0.275 | 0.009 | |||||
| Portal venous blood flow per 100 g of liver graft at postoperative day 5 or 6 (ml/min/100 g) | r2 = 0.237 | 0.012 | |||||
| Jiang et al. (2009) [74] | 18 | Liver cirrhosis | Mean maximal velocity of the portal vein multiplied by a coefficient of 0.57 (PBV) at postoperative day 1 (cm/s) | Percent change in the liver graft volume between the preoperative period and postoperative day 30 (%) | Simple linear regression | r = 0.67492* | 0.0001 |
| Multiple regression | UA | 0.0011 | |||||
| Portal venous blood flow at postoperative day 1 (ml/min) | Simple linear regression | UA | < 0.05 | ||||
| Multiple regression | UA | 0.0003 | |||||
| PBV (cm/s) and portal venous blood flow (ml/min) at postoperative day 3 | Simple linear regression | UA | < 0.05 |
PORTAL HYPERPERFUSION INJURY AND SMALL-FOR-SIZE SYNDROME
SFSS has emerged as a new challenge with the increasing practice of LDLT [75]. It is manifested as graft function impairment, characterized by portal hypertension, refractory ascites, encephalopathy, prolonged cholestasis with hyperbilirubinemia, and coagulopathy during the 1st postoperative week in the absence of technical problems such as hepatic venous outflow obstruction [76,77]. In severe cases, the syndrome progresses to acidosis, hypoglycemia, renal failure, and septic shock unless prompt retransplantation is performed [78]. The histopathological examination of grafts complicated by SFSS in a porcine model showed evidence of hepatic artery vasospasm and resulting cholestasis, centrilobular necrosis, and biliary ischemia [79]. In the early era of LDLT, SFSS was known to develop in liver grafts with a GRWR < 0.8%, or a graft weight to standard liver volume < 35% [80], which might indicate a relative shortage of hepatic parenchymal volume for life maintenance. However, several definitions of SFSS have been proposed without a single accepted definition [76,81-83] (Table 2) and the size of the liver graft is no longer an absolute determinant of the development of SFSS. Hepatic hemodynamic derangement, particularly portal hyperperfusion producing high intravascular shear stress, might also have an influence on its development [84]. Recently, it was suggested that high portal blood inflow to a partial graft, transplanted in a recipient with a persistent hyperdynamic splanchnic circulation, caused hepatic dysfunction and impairment of hepatic regeneration [85-87]. The generated shear stress also leads to an imbalance in endothelin-1 and NO, which contributes to graft injury [88]. Accordingly, improvements in surgical techniques for modulating PVF led to a reduction in the lower limit of GRWR to 0.6 [89].
Table 2
Various Definitions of Small-for-size Syndrome
| Reference | Definition |
|---|---|
| Dahm et al. [76] |
Small-for-size dysfunction Dysfunction* of a ‘small’ partial liver graft (graft-versus-recipient weight ratio < 0.8%) during the 1st postoperative week after the exclusion of other causes† |
|
Small-for-size non-function Failure‡ of a ‘small’ partial liver graft (graft-versus-recipient weight ratio < 0.8%) during the 1st postoperative week after the exclusion of other causes |
|
| Soejima et al. [81] |
Small-for-size syndrome Prolonged cholestasis (total serum bilirubin > 10 mg/dl on the 14th postoperative day without any other definitive causes of cholestasis) and intractable ascites (daily production of ascites of > 1 L on the 14th postoperative day or > 500 ml on the 28th postoperative day |
| Hill et al. [82] |
Small-for-size syndrome The presence of significant cholestasis with serum bilirubin > 10 mg/dl (and continuing to increase) after the 7th postoperative day, coagulopathy with an international normalized ratio > 1.5, and ascites with drain out > 2 L/day in the absence of an obvious technical problem such as biliary leak, vascular thrombosis, or stenosis. |
| Ikegami et al. [83] |
Delayed functional hyperbilirubinemia (yields the highest area under the receiver operating characteristic curve (0.977), representing a sensitivity of 100% and specificity of 95.4% for detecting early graft loss compared with the three definitions above.) Total serum bilirubin > 20 mg/dl for > 7 consecutive days occurring after the 7th postoperative day, excluding technical, immunological, and hepatitis factors. |
* The presence of two of the following on 3 consecutive days: serum bilirubin > 100 μmol/L, international normalized ratio > 2, and encephalopathy grade 3 or 4.
Portal hyperperfusion activates HABR, which causes hepatic arterial hypoperfusion in the partial liver graft. In patients undergoing LDLT, an increase in PVF to the grafts is by more than two-fold [19]. In the absence of HABR, the HAF would increase in line with the increase in the PVF. However, a considerable decrease in the HAF was observed in patients receiving right-lobe grafts [90], indicating that HABR operates in response to an increase in the PVF to maintain the total hepatic blood flow within a physiological range [91]. The mean PVF of recipients transplanted with small-sized liver grafts (GRWR < 0.8) was at least three times higher than that of donors [92]. In contrast, the hepatic arterial contribution to total hepatic blood flow was decreased from 30% in donors to 6% in recipients, due to a substantial reduction in the HAF [92]. SFSS developed in 3 of 11 (27%) patients who did not undergo graft inflow modulation and a PVF of 250 ml/min per 100 g of liver weight predicted the development of SFSS [92]. In a porcine model transplanted with a small-for-size graft, the portal-to-hepatic flow ratio, which was poorly tolerated by the liver graft, remained increased until the 5th postoperative day [93].
Graft inflow modulation
Because a recipient transplanted with a liver graft, the GRWR of which was 0.61%, survived following a mesocaval shunt that reduced PVP [94], several techniques to divert the PVF to the systemic circulation – or to decrease the portal inflow by modifying the splenic blood flow – have been developed, such as hemiportocaval shunt [95,96], mesorenal shunt [97,98], delayed ligation of spontaneous portosystemic shunts [99], splenectomy [100], splenic artery ligation and embolization [101,102].
A hemiportocaval shunt between the right portal vein and inferior vena cava reduced the HVPG from 18 to 5 mmHg in 16 patients with a median GRWR of 0.67, among whom only one patient required retransplantation due to the development of SFSS [96]. Given the complications associated with the procedure (e.g., encephalopathy due to systemic shunting of the PVF with subsequent hyperammonemia [96], and graft atrophy due to insufficient portal inflow caused by the systemic shunting), the shunt was closed postoperatively to restore the PVF back to the liver by deploying an aortic covered endograft [103] or by tightening of the endo-loop left around the shunt at the time of transplantation [104]. As an alternative, a mesorenal shunt between the inferior mesenteric vein to the left renal vein is easy to perform and prevents excessive portosystemic shunting while also decreasing the PVP [97,98].
Occlusion of the splenic circulation (splenectomy, splenic artery ligation or embolization), which contributes considerably to the portal inflow, not only decreases the PVF [92] and PVP [85,87], but also increases the HAF [92] with a subsequent increase in oxygen supply [105] via HABR [106]. Although splenectomy, which abolishes both arterial and venous blood flow to the portal vein, is more effective in decreasing the PVF than splenic artery ligation or embolization, it increases the length of the operation [107] and the risks of bleeding, infection [107], and portal vein thrombosis [108] in cirrhotic patients exhibiting a hyperdynamic state with collateral circulation around the splenic artery (such as a gastric coronary vein and spleno-renal shunt) [109]. However, despite the absence of complications associated with splenectomy, a splenic abscess resulting from a splenic infarction may occur following splenic artery ligation in patients with an enlarged spleen due to portal hypertension [110]. The effect of splenic artery embolization is comparable to that of splenic artery ligation [101].
The criteria of the hepatic hemodynamic parameters for graft inflow modulation are still debated. An elevated mean PVP of more than 20 mmHg early in the first week after LDLT was found to be associated with an increased incidence of bacteremia in the first 3 months and worse graft survival at the 6th postoperative month [87]. Patients with a final PVP less than 15 mmHg had better 2-year survival and recovery from hyperbilirubinemia and coagulopathy after LDLT than those with a PVP more than 15 mmHg [100]. Thereafter, a PVP less than 15 mmHg was suggested as a surgical strategy for small-for-size grafts in the subsequent study in which intentional PVP modulation was performed to achieve a target PVP of less than 20 mmHg [111].
The PVF is another hepatic hemodynamic parameter used to determine whether to perform graft inflow modulation. In full-size liver transplantation, four times the PVF of healthy donors (360 ml/min per 100 g of liver weight) was shown to be a risk factor for graft failure; a flow rate below 180 ml/min per 100 g of liver weight was also associated with lower survival rates [112]. These results were confirmed in an experimental study that used small-for-size grafts [113]. In addition, the results of other studies showed a PVF less than 260 ml/min 100 g of liver weight for preventing allograft dysfunction [114]. Similarly, a PVF of 250 ml/min per 100 g of liver weight predicted the development of SFSS [92]. By incorporating several criteria, including PVF, PVP, or HVPG, many flowcharts for graft inflow modulation have been proposed [84,112,115]. However, a detailed discussion of these flowcharts is beyond the scope of this review.
Pharmacological intervention
Adenovirally overexpressed redox factor-1 in a partial liver graft [116], an endothelin A receptor antagonist that maintains a balance between endothelin-1 and NO [88], low-dose somatostatin (which attenuates acute-phase shear stress due to transient portal hypertension) [117], a preservation solution containing activated protein C (which has cytoprotective properties due to its anti-inflammatory and anti-apoptotic effects) [118], venous systemic oxygen persufflation with NO gas [119], and a newly developed low-viscosity preservation solution (POLYSOL) [120] reduced graft injury after reduced-size rat liver transplantation. Subcutaneous injection of granulocyte colony-stimulating factor [121] and hyperbaric oxygen treatment [122] also reduced liver damage in rats undergoing massive partial hepatectomy. Intraportal infusion of nafamostat mesilate (protease inhibitor), prostaglandin E1, and thromboxane A2 synthetase inhibitor for 7 days prevented SFSS in patients undergoing LDLT [123].
Anesthetic considerations for portal hyperperfusion
The role of anesthesiologists in liver transplantation is to maintain multiple systemic hemodynamic parameters (e.g., radial, femoral, and pulmonary arterial pressure, central venous pressure [measured from the internal jugular or subclavian vein, and femoral vein], cardiac index, stroke volume index, stroke volume variation, and systemic vascular resistance index) within normal ranges for patient protection against the physiological derangements caused by the surgical procedure. Among these parameters, central venous pressure measured from the internal or subclavian vein (into which a catheter is inserted with its tip located within the lower third of the superior vena cava close to the junction of the superior vena cava and right atrium) [124] is comparable to hepatic vein pressure [125] which might influence the portal hemodynamic parameters. Accordingly, PVP has been reported to be modulated by central venous pressure [112,126,127]. An increase in central venous pressure by 58% was found to be transmitted to the PVP in patients receiving liver transplantation [112] and 60% [126] and 90% [127] of inferior vena cava pressure contributed to PVP in experimental studies. Recently, maintenance of central venous pressure between 5 and 10 mmHg during the neohepatic phase was recommended to prevent portal hyperperfusion of a liver graft. This recommendation was derived from a predictive model built with percent change in peak portal vein flow velocity (which was measured by spectral Doppler ultrasonography after reconstruction of the hepatic artery and bile duct as well as on the 1st postoperative day), and the central venous pressure measured from the right internal jugular vein (which was averaged for 5 min after the measurement using spectral Doppler ultrasonography during the neohepatic phase) [128].
CONCLUSIONS
In accordance with the clinical conditions that liver grafts encounters, the graft either regenerates or fails to accommodate portal hyperperfusion. Although GRWR < 0.8, standard liver volume < 35%, or excessive portal blood inflow to a partial graft in the presence of hyperdynamic splanchnic circulation were found to contribute to hyperperfusion injury, they are not yet used as absolute determinants of hyperperfusion injury and still have to be confirmed in further investigations. Furthermore, clinical strategies and modalities to maintain the viability of grafts showing reduced direct control of enhanced PVF have not been standardized. It is hoped that anesthesiologists will contribute to the prevention of hyperperfusion injury of liver grafts by maintaining optimal systemic hemodynamic stability, and by conducting systemic pharmacological interventions in collaboration with transplantation surgeons based on a comprehensive understanding of the clinical implications of portal hyperperfusion caused by the underlying hyperdynamic splanchnic circulation resulting from portal hypertension.
XML Download