Liver and systemic hemodynamics in cirrhotic children

Abstract

Portal hypertension and cirrhosis are associated with severe hemodynamic changes in hepatic and systemic circulation in the adult population. During cirrhosis progression, circulation becomes hyperdynamic, with cardiac, pulmonary and renal consequences. Cirrhotic adults also present with cirrhotic cardiomyopathy, with systolic and diastolic dysfunction and electrophysiological abnormalities. This article provides an update on normal liver hemodynamics, a brief reminder of the liver and systemic hemodynamics in cirrhotic adults, and a description of liver and systemic hemodynamics in cirrhotic children. This review attempts to clarify whether liver and systemic hemodynamics are altered in cirrhotic children like they are in adults. The characterization of these hemodynamic disturbances could contribute to a better understanding of hepatic and systemic physiopathology in pediatric cirrhosis.

Key Words: Cirrhosis; Children; Portal hypertension; Liver hemodynamics; Systemic hemodynamics; Cirrhotic cardiomyopathy

Core Tip: Portal hypertension and cirrhosis are associated with severe hemodynamic changes in hepatic and systemic circulation in adults. With the progression of cirrhosis, the circulation becomes hyperdynamic, with cardiac, pulmonary and renal consequences. Cirrhotic adults also present with cirrhotic cardiomyopathy. This article provides a brief review of liver and systemic hemodynamics in cirrhotic adults and a detailed description of liver and systemic hemodynamics in cirrhotic children. The characterization of these hemodynamic disturbances could contribute to a better understanding of hepatic and systemic physiopathology in pediatric cirrhosis, elements that should be included in the management of cirrhotic children.

INTRODUCTION

In adults, portal hypertension and cirrhosis are associated with hemodynamic changes in both hepatic and systemic circulation[1]. In addition to liver-related complications, patients with cirrhosis and portal hypertension develop extrahepatic functional disturbances of multiple organ systems. Progressive liver fibrosis and subsequent metabolic impairments lead to systemic and splanchnic arteriolar vasodilatation. This affects the hemodynamic and functional homeostasis of many organs and largely determines disease evolution. As cirrhosis progresses, circulation becomes hyperdynamic, with cardiac, pulmonary and renal consequences[2]. Cirrhotic cardiomyopathy implies systolic and diastolic dysfunction and electrophysiological abnormalities, which may further aggravate complications, such as hepatorenal syndrome[2,3]. However, knowledge regarding corresponding hemodynamic alterations in children with cirrhosis is limited[4,5]. This article provides an update on normal liver hemodynamics, a brief description of the liver and systemic hemodynamics observed in cirrhotic adults, and a detailed description of liver and systemic hemodynamics in cirrhotic children. We reviewed the literature for specific hemodynamic changes observed in pediatric cirrhosis. A MEDLINE literature search was conducted using the following search terms: ‘cirrhosis’, ‘hemodynamic’, ‘systemic’, ‘liver’, ‘cardiomyopathy’, ‘circulation’, and ‘children’.

LIVER HEMODYNAMICS IN NORMAL PHYSIOLOGY

Historical perspectives

William Harvey’s description of blood circulation in 1628 allowed for the first time the development of accurate knowledge about liver circulation. In the mid-17^th century, Glisson, after whom the hepatic capsule is named, advanced our knowledge of the gross anatomy of the hepatic vascular bed and demonstrated that portal blood flowed through the liver. In 1664, Wepfer was reportedly the first to observe the glandular appearance of hepatic acini beneath Glisson’s capsule. Twenty-one years later in 1685, Malpighi confirmed the presence of similar microvascular units, which he redefined as hexagonal lobules. To this day, there remains a lack of consensus on whether the liver microvascular unit should be referred to as a lobule (centering on a hepatic vein) or an acinus, centering on a portal triad, consisting of a terminal branch of the hepatic artery (HA), portal vein (PV), and bile duct (BD)[6].

In the late 19^th century, Claude Bernard and Ernest Starling established the liver as a central organ with critical endocrine, metabolic, and vascular functions.

Overview of normal liver hemodynamics

The liver receives 25% of the cardiac output (CO), although it constitutes only 2.5% of the body’s weight. Hepatic parenchymal cells are the most richly perfused of any of the organs. The liver is characterized by two inflow systems, a low resistance and high flow system through the PV and a high resistance and low flow system through the HA. Of the total hepatic blood flow (100-130 mL/minute per 100 g of liver, 30 mL/minute per kilogram of body weight), 20%-30% is supplied by the HA (± 20 mL/minute per 100 g of liver). About 70%-80% of the hepatic blood supply is portal venous blood (± 80 mL/minute per 100 g of liver)[6,7]. Therefore, the portal-to-arterial flow ratio is approximately 4. A summary of key figures is presented in Table 1. Within the hepatic sinusoids, high pressure, oxygen-rich arterial blood thoroughly mixes with low-pressure, less oxygenated but nutrition-rich portal venous blood. This mixed blood is then drained in a common outflow venule, which ultimately drains into the hepatic veins and through the inferior vena cava[8,9]. The blood drained by the hepatic veins represents 66% of venous return through the inferior vena cava.

Table 1 Overview of reported key values.

Normal liver hemodynamics (mL/min per 100 g of liver)
Total hepatic blood flow	100-130
Supplied by the HA	± 20
Supplied by the PV	± 80
Porto-to-arterial flow ratio	4/1
Normal portal pressure (mmHg)	7-10
Cirrhotic adults (mmHg)
Portal hypertension, HVPG	> 5
Significant portal hypertension (varices, gastropathy, ascite, etc.)	> 10
Portopulmonary hypertension, Pulmonary arterial pressure	> 25
Cirrhotic children
Cirrhosis, PV velocity of (cm/second)	Less than 20
Arterial resistance index (nL 0.6-0.7)	> 0.7
Normal portal vein diameter (mm)
At birth	3-5
At 1 year	4-8
At 5 years	6-8
At 10 years	6-9
At 15 years	7-11
PV hypoplasia internal PV diameter:	≤ 4 mm at Doppler ultrasound

HA: Hepatic artery; PV: Portal vein; HVPG: Hepatic venous pressure gradient.

In normal physiological conditions, the volume of blood contained within the liver is high, and the liver plays a role as a major blood reservoir that can be mobilized if required. It has a crucial role in the response to blood loss or expanded fluid volume. The main point of resistance to the venous flow out of the splanchnic vasculature is located within the hepatic veins or within the liver itself[10-12]. An increase in resistance within the distal portion of the splanchnic venous system would impede the outflow of blood from splanchnic organs, sequestering blood within the liver and more proximal parts of the splanchnic veins. This has been confirmed in experimental settings where profound arterial hypotension during septic shock was associated with a drastic increase in venous resistance within the distal part of the splanchnic vasculature[13,14]. A decrease in resistance to venous flow within the liver and/or hepatic veins would facilitate blood flow and volume shift from splanchnic vasculature to the inferior vena cava and right atrium, thereby increasing venous return. In the hemodynamic circuit, the liver is located downstream of the right side of the heart and upstream of the splanchnic compartments. Any variations of these may influence hepatic flows.

Portal pressure is due to intrahepatic resistance and portal blood flow and is defined as a function of flow and resistance across the hepatic vasculature (pressure = flow × resistance). Portal pressure is commonly measured by the hepatic venous pressure gradient (HVPG), defined as the difference between wedged (occluded) and free hepatic venous pressures, with normal values ranging between 1 and 5 mmHg. The liver is a passive recipient of fluctuating amounts of blood flow, which can encompass a wide range of flow with minimal effects on portal system pressure[15-18]. Intrahepatic and PV pressures are primarily regulated by hepatic venous sphincters and, in the basal resting state, portal pressure is insignificantly different from sinusoidal pressure (7-10 mmHg). However, veins have limited elasticity once they are fully distended and, at this point, PV pressure rapidly increases with increased volume[19]. The hepatic circulation has multiple interacting mechanisms that strive to maintain constant hepatic blood flow under both acute and chronic physiological conditions. Portal blood flow is determined by the net outflows from splanchnic organs including the stomach, spleen, pancreas, intestines, and omentum. In response to fluctuations in portal blood flow, the hepatic arterial flow adjusts in the opposite direction, thus helping to stabilize total hepatic blood flow. This mechanism does not completely compensate but rather buffers the effect of portal blood flow changes on total hepatic flow, a phenomenon known as the hepatic arterial buffer response (HABR)[6,7]. The HABR is thought to be mediated through adenosine. Adenosine is continuously released into the space of Mall, a small, anatomically confined fluid compartment through which the portal triad passes - comprising terminal branches of the HA, PV, and bile ductule. As a potent vasodilator, the local adenosine concentration is regulated by the rate of washout from the space of Mall into the blood vessels. When portal blood flow decreases, adenosine clearance from the space of Mall is reduced, resulting in its accumulation. The elevated adenosine concentration subsequently induces HA vasodilation, thereby contributing to the compensatory increase in arterial inflow[20].

The microvascular unit of the liver seems to be the acinus. The acinus represents a cluster of parenchymal cells approximately 2 mm in diameter, organized around the terminal branches of the hepatic arteriole and portal venule. In the event of localized portal venous stasis, the HA supplying that acinus dilates to increase the arterial input, thereby flushing the sinusoids and restoring sinusoidal blood flow. The uptake of substances by parenchymal cells, as well as the exchange between these cells and plasma, is influenced by several features of the hepatic microvascular circulation. The characteristics of uptake and exchange by the liver play a critical role in processes such as lipoprotein metabolism, endocrine homeostasis, and clearance of therapeutic agents[6].

In the following paragraphs, we will first focus on liver hemodynamics in adults with cirrhosis, followed by an overview of their systemic hemodynamics. We will then compare it with hepatic and systemic hemodynamics in cirrhotic children, an area where the literature is less extensive. The final section will address hemodynamics in children with cirrhosis secondary to biliary atresia (BA), the leading indication for liver transplantation (LT) in children.

LIVER AND SYSTEMIC HEMODYNAMICS IN CIRRHOTIC ADULTS

Liver hemodynamics in cirrhotic adults

Liver cirrhosis is the final pathological and clinical expression of a wide variety of chronic liver diseases. In the cirrhotic liver, there is an increase in vascular resistance. Approximately 70% of this resistance is mechanical due to architectural changes of fibrosis and potential vascular occlusion. The remaining 30% of the increased vascular resistance in cirrhotic livers is a dynamic component due to endothelial dysfunction and increased vascular tone[21]. This increased vascular tone is characterized by an imbalance between vasoconstrictor and vasodilator agents in the hepatic microcirculation. In addition, there is a hyper-responsiveness to vasoconstrictors and decreased vasodilator production, with an insufficient response of the hepatic vascular bed to vasodilators[22].

Pathophysiologically, the clinical manifestation of liver cirrhosis arises from two major events: hepatocellular insufficiency and portal hypertension[23]. In cirrhosis, loss of normal hepatic architecture leads to increased sinusoidal or post-sinusoidal portal resistance to blood flow, resulting in portal hypertension (defined as a HVPG greater than 5 mmHg)[24]. The major pathophysiologic consequences of portal hypertension include portal-systemic collaterals formation, ascites and splenomegaly. It is well known that varices, variceal bleeding, portal hypertensive gastropathy, and ascites do not occur until the HVPG increases above 10 mmHg, a pressure gradient threshold that clinically defines significant portal hypertension.

In cirrhosis, portal perfusion progressively decreases because of increasing sinusoidal resistance and development of spontaneous portosystemic collaterals. Thus, total liver blood flow becomes increasingly dependent on HA blood flow. Results on HA flow and resistance in cirrhosis are inconsistent and whether HABR is preserved in adult patients with cirrhosis and is dependent on disease stage is unclear[25-35]. Most studies have used non-invasive transcutaneous Doppler ultrasound or intraoperative electromagnetic probes. Several of these studies including cirrhotic adults described low PV flow and increased HA flow in their native livers[25,31,33,34]. Indeed, Kleber et al[31] reported an increased hepatic arterial flow in Child-Pugh class C adult patients, suggesting that HABR might be independent of the liver disease stage. Conversely, other studies have demonstrated an increase in hepatic arterial resistance in adults with cirrhosis, which correlates with the severity of portal hypertension, the level of portal venous resistance, and Child-Pugh classification[26,29,30,35]. Consistent with these findings, HA blood flow appears to be reduced in patients with cirrhosis during laparotomy, as measured using electromagnetic flow probes[25]. The hypothesized increase in HA resistance in cirrhosis has been attributed to fibrotic scarring surrounding the arterial vessels. According to this hypothesis, resistance should increase as cirrhosis progresses.

Systemic hemodynamics in cirrhotic adults

Adult patients with end-stage liver disease have hyperdynamic circulation characterized by decreased systemic vascular resistance and arterial pressure, and increased heart rate and CO[36]. The clinical manifestations of hyperdynamic circulatory syndrome include warm skin, spider angioma, palmer erythema, and bounding pulse. These cardiovascular changes were described by Kowalski and Abelmann[37] over 60 years ago. Hyperdynamic circulation is most likely initiated by splanchnic and peripheral vasodilatation, leading to reduction in the effective arterial blood volume. This leads to a diminished renal blood flow in cirrhotic patients, which in turn stimulates the renin-angiotensin-aldosterone system, sympathetic nervous system, and antidiuretic hormone, resulting in renal artery vasoconstriction, sodium retention, and volume expansion[38]. Worsening liver disease results in progressive vasodilatation, making the hyperdynamic circulation and renal artery vasoconstriction more pronounced[39,40]. These circulatory changes lead to the development of multiple life-threatening complications including hepatorenal syndrome, ascites, gastroesophageal varices, and hepatopulmonary syndrome[41,42].

Portopulmonary hypertension syndrome (mean pulmonary arterial pressure > 25 mmHg at rest) is also associated with end-stage liver disease. Portopulmonary hypertension is less prevalent (3.5%-8.5% in LT candidates) than hepatopulmonary syndrome[43,44]. It is associated with a high five-year mortality rate, estimated between 50% and 90%. Unlike hepatopulmonary syndrome, LT may be contraindicated in cases of severe portopulmonary hypertension.

In addition to hyperdynamic circulation, impaired ventricular contractility in response to stimuli has also been described in cirrhotic patients[45,46]. These cardiovascular changes are now termed ‘cirrhotic cardiomyopathy’[47-49]. The characteristic features of cirrhotic cardiomyopathy include: Attenuated systolic and diastolic response to stress stimuli despite an increased basal CO (because of increased cardiac contractility and peripheral vasodilatation), myocardial hypertrophy, structural changes in cardiac chambers, and electrophysiological abnormalities, including prolonged QT interval[50].

WHAT DO WE KNOW ABOUT LIVER AND SYSTEMIC HEMODYNAMICS IN CIRRHOTIC CHILDREN?

Liver hemodynamics in cirrhotic children

Available data concerning liver hemodynamics in the cirrhotic pediatric population have been collected by Doppler sonography. Although Doppler sonography has become widely available, reports of hepatic hemodynamic measurements in pediatric patients with liver disease are scarce.

Portal hypertension secondary to cirrhosis is readily diagnosed when there is reversal of flow within the PV, directed away from the liver (hepatofugal), although this is a very late and severe finding that can be caused by other hepatic vascular abnormalities[51]. More typically, PV velocities decrease in cirrhosis, and the PV waveform becomes more pulsatile[52-54]. PV pulsatility is reported to be 94% sensitive for portal hypertension. It has also been reported that there is a significant difference in PV velocity between Child-Pugh class A and Child-Pugh B and C classes of liver cirrhosis in children[55]. In the same way, El-Shabrawi et al[56] showed a reduction in PV flow velocity according to Child-Pugh classification in another pediatric population. In fact, PV velocity of less than 20 cm/second is 67%-83% sensitive for cirrhosis[52-54].

With decreased PV flow, there is a relative increase in HA flow. The ratio of HA to PV flow at Doppler ultrasound has been suggested to be a good indicator of portal hypertension, with a ratio of more than 3 having a sensitivity between 75% and 78%[53,54]. While the contribution of hepatic arterial flow to liver perfusion may be increased, continued hepatic parenchymal disease may lead to increased vascular resistance and therefore decreased diastolic blood flow (with increased waveform pulsatility). The hepatic arterial resistance index (ARI; calculated by ARI = (maximal systolic velocity - minimal diastolic velocity)/maximal systolic velocity; normal values 0.6-0.7) is increased in cirrhotic children compared to healthy control subjects[57] (Figure 1).

Figure 1 Measure of hepatic arterial resistance index by Doppler ultrasound. A: Arterial resistance index (ARI) < 1; B: ARI ≥ 1. VSM: Peak Systolic Velocity; VTD: End Diastolic Velocity; VDM: Mean Diastolic Velocity; IR: Resistance index.

Hepatic venous waveforms are also affected by portal hypertension and intrinsic liver disease. Both decreased hepatoportal blood flow and parenchymal compliance dampen hepatic venous waveforms, (which normally have a triphasic pattern[53,56]. Gorka et al[53] identified loss of reverse flow in the hepatic veins as the most reliable indicator of liver cirrhosis in children, with 70% sensitivity for early and 100% sensitivity for established cirrhosis. However, other studies have found that a monophasic flow pattern was the most common in children less than 1 year of age, and therefore rejected the absence of triphasic flow pattern as a sign of liver pathology in children, especially when they are younger than 1 year of age[58].

Our group has also described important alterations of native liver hemodynamic parameters in cirrhotic children at pre-LT Doppler ultrasound: They had a smaller extrahepatic internal PV diameter (median, 4.3 mm) than children without cirrhosis (median 5.9 mm; P = 0.047), lower PV velocity (median, 15.5 cm/second; P = 0.007), and a higher ARI (median, 0.9; P = 0.003), and these alterations were correlated with Pediatric End-Stage Liver Disease score (PELD)[59].

There are few published studies concerning native liver hemodynamics in cirrhotic children using Transit Time Flow Measurement (TTFM)[60,61]. The first describes PV flow (mean flow = 13 mL/minute/kg of body weight) and HA flow (mean flow = 13 mL/minute/kg of body weight) in the native liver (total mean flow = 23 mL/minute/kg of body weight) as similar during the LT procedure, with a preservation of HABR in these children[60]. Another study[61] using TTFM revealed that the portal contribution to total hepatic blood flow was markedly decreased in Child-Pugh B and C class native livers compared to the normal liver. Our group demonstrated that children with cirrhosis had a smaller extrahepatic external PV diameter (median, 6 mm) than children without cirrhosis (median 8.5 mm; P = 0.01; perioperative invasive measures). The PV flow into the native liver (measured by TTFM) was significantly lower in children with cirrhosis (median, 12 mL/minute/100 g of liver) than in children without cirrhosis (64 mL/minute/100 g of liver; P < 0.001). Similarly, the total flow into the native liver was lower in patients with cirrhosis (median, 36 mL/minute/100 g of liver) compared to children without cirrhosis (median, 86 mL/minute/100 g of liver; P = 0.002). In contrast, the HA flow was similar in children with cirrhosis (median, 20 mL/minute/100 g of liver) and children without cirrhosis (median 19 mL/min/100 g of liver; P = 0.93, not significant). Thus, HABR did not seem to be functional in children with cirrhosis. Accordingly, pediatric cirrhotic livers were essentially perfused by arterial flow (median, 67.3%), compared to the noncirrhotic liver in which only 21.1% of the flow was delivered by the HA[59]. We also observed that alterations of liver hemodynamic parameters collected by invasive flowmetry were correlated with the clinical degree of cirrhosis (PELD score) and with the histological degree of fibrosis[59].

Concerning portal pressure, the main clinical applications of HVPG measurements include diagnosis, classification, and monitoring of portal hypertension, risk stratification, and identification of candidates for LT[62,63]. This procedure is also feasible in children with chronic liver diseases[64,65]. Miraglia et al[64] measured HVPG in 20 children with chronic liver disease. In all patients with clinical and imaging evidence of portal hypertension that was not due to extrahepatic PV obstruction, the HVPG was increased to ≥ 8 mmHg. A more recent study published by Woolfson et al[65] that included seven pediatric patients with cirrhosis described an HVPG from 10 to 20 mmHg. Our group also demonstrated that cirrhotic children had a higher degree of portal hypertension (median portal venous pressure-central venous pressure gradient, 14.5 mmHg), compared to patients without cirrhosis (median 3.5 mmHg; measures taken perioperatively during the LT procedure)[59].

Systemic hemodynamics in cirrhotic children

We previously described the systemic and cardiac changes observed in cirrhotic adults. Whether similar hemodynamic adaptations occur in cirrhotic children is not well documented.

Cardiac manifestations of end-stage liver disease in the pediatric population: A study comparing 2-dimensional echocardiographic parameters between 22 children with biopsy-proven cirrhosis and 22 age-sex-matched controls showed no significant differences in left ventricular ejection fraction, left ventricular shortening fraction, and left ventricular diastolic dimension. Systolic left ventricular posterior wall thickness was significantly higher in cirrhotic patients[66]. Another study evaluating the cardiovascular status of 40 children with BA listed for LT showed abnormal 2-dimensional parameters (measured by echocardiography) in 70% of patients, compared to well-matched controls. Abnormalities included increased left ventricular wall thickness, left ventricular mass indexed to body surface area (51% increase), and left ventricular shortening fraction (8% increase). Overall, features of cirrhotic cardiomyopathy were observed in most infants (72%); 42% had hyperdynamic contractility and 60% had altered left ventricular geometry[4]. Post-LT pediatric intensive care unit stay and hospital length of stay were significantly longer in the patients with abnormal 2-dimensional parameters[4]. Based on current knowledge, cardiovascular screening is warranted in children with cirrhosis. How abnormal 2-dimensional findings alters patient management is unclear[5]. Cardiac repolarization abnormalities, such as QT interval prolongation on electrocardiogram, can also be observed in children with chronic liver disease. A study on 38 pediatric patients awaiting LT revealed prolonged QT in seven patients (18%)[67]. None of the patients developed ventricular arrhythmia or QT normalized in survivors of LT. Additionally, studies in pediatric populations have demonstrated a correlation between QT prolongation and PELD score[68]. Clinicians should therefore exercise caution in managing children with chronic liver disease and a prolonged QT interval. Furthermore, it is essential to maintain normal electrolyte levels and avoid medications that could further prolong the QT interval in these patients[5].

Pulmonary circulation: The prevalence of hepatopulmonary syndrome in children with chronic liver disease ranges between 4.6% and 19%[69-71]. Children with features suggestive of hepatopulmonary syndrome (cyanosis, exertion dyspnea, platypnea, digital clubbing) should be evaluated by a cardiologist via contrast-enhanced transthoracic 2-dimensional echocardiography using agitated saline injection. Opacification of the right side followed by opacification of the left atrium within 3-6 cardiac cycles is a positive test result for hepatopulmonary syndrome. The presence of intrapulmonary shunting can also be demonstrated using a technetium-99m-labeled macroaggregated albumin perfusion scan[72]. As in adults, LT is the only proven therapy for hepatopulmonary syndrome.

The prevalence of portopulmonary hypertension in children is not well known[73-75]. Clinical symptoms are nonspecific and often subtle (syncope, dyspnea, reduced exercise tolerance, atypical chest pain). Evidence-based screening recommendations in children need to be established. A systematic screening of all pediatric patients before LT with 2-dimensional echocardiography must be performed to exclude the diagnosis of portopulmonary hypertension, even if echocardiograms do not detect mild portopulmonary hypertension and may underestimate the degree of pulmonary arterial hypertension[76]. Thus, cardiac catheterization is necessary for the diagnosis and further treatment of most patients with portopulmonary hypertension[75]. In our center, cardiac catheterization is systematically performed in children older than 12 years old. The goal of portopulmonary hypertension treatment is to transform a borderline candidate for LT into an acceptable one through aggressive treatment of the pulmonary arterial hypertension. Treatment may include providing supplemental oxygen to maintain saturation above 92%, diuretics to control volume overload, and calcium-channel blockers if the patient has demonstrated good reactivity to vasodilators during cardiac catheterization[77-79]. Continuous prostacyclin improves survival in children with pulmonary hypertension, but it must be given by continuous intravenous infusion[79]. Whether LT cures portopulmonary hypertension in children is unclear[80,81].

Cirrhosis secondary to BA

Liver hemodynamics in children affected by BA: BA is the most frequent indication of LT in children. The cause of BA remains unknown and despite successful portoenterostomy, progressive inflammation and fibrosis of intrahepatic BDs develop in all patients, often leading to biliary cirrhosis[82]. BA is characterized by portal fibrosis, progressively linking portal areas that may dramatically progress to secondary biliary cirrhosis within a few weeks after birth[83]. Rapid progression of liver failure is a well-known complication of children with BA cirrhosis.

Cirrhosis secondary to BA may be associated with severe alterations of liver hemodynamic parameters[84-86]. In this way, Nakada et al[84] demonstrated that there was a good correlation between deteriorating liver function and a decrease in main PV velocity at Doppler ultrasound in children with BA (with a fluctuation in the direction of PV flow). Another study demonstrated that portal hypertension had already developed in most patients with BA at 2-4 months of age (at the time of portoenterostomy)[87]. Several authors have also described an ARI ≥ 1 as a bad prognostic factor for children waiting for a LT (with a death risk), especially in BA cases[86,88-90]. Our group also showed that impaired native liver hemodynamic parameters measured non-invasively by Doppler ultrasound before LT and during the transplantation by TTFM were mostly observed in children with cirrhosis affected by BA[59].

A hypoplastic PV is more frequently observed in children affected by BA. PV hypoplasia is generally defined as an internal PV diameter ≤ 4 mm at pre-transplant Doppler ultrasound[91-93] This definition is based on the work of several authors who have studied the PV diameter by real-time sonography in normal children according to age, weight and height. They all describe normal PV diameters of ≥ 4 mm, while only some newborns have a PV diameter of 3 mm. Indeed, the data found in the literature describe a normal PV diameter of 3-5 mm at birth, 4-8 mm at 1 year, 6-8 mm at 5 years, 6-9 mm at 10 years, 7-11 mm at 15 years[94-96]. According to these values in normal children, PV hypoplasia has been defined in the literature as an internal PV diameter ≤ 4 mm at Doppler ultrasound.

Saad et al[97] analyzed the pathology of the extrahepatic PV in 35 consecutive cases of children with BA with a previous portoenterostomy to evaluate the changes in vessel histology. They described an alteration of PV quality in 80% of cases, with mild to severe fibrosis and thickening of the vessel wall. Our group also showed thickening of the extrahepatic PV wall that is concentrated at the intima and represents a unique characteristic of BA, as it has not been observed in non-BA children with cirrhosis[59].

This characteristic PV hypoplasia (Figure 2) observed in BA patients may lead to specific technical difficulties in PV reconstruction during LT, which increases the risk of post-transplant PV complications, and thus the risk of graft loss[98-100].

Figure 2 Peri-operative pictures of portal vein hypoplasia in children with biliary atresia. LG: Liver graft; PV: Portal vein; SMC: Splenomesenteric venous confluence.

All these liver hemodynamic findings of children affected by BA contribute to the fact that perfusion of the native liver may become critical at a very young age. Indeed, they may have a reversed portal flow because of severe portal hypertension and PV hypoplasia, and an associated reversed arterial diastolic flow secondary to increased intra-hepatic vascular resistances. Apart from increased intra-hepatic resistances, several factors may diminish arterial blood supply to the liver in patients with advanced BA cirrhosis: (1) Acute variceal bleeding may cause hypovolemia and hypotension[101-103]; (2) Hypovolemia may be secondary to other conditions such as gastroenteritis; and (3) Splanchnic and systemic vasodilatation may be responsible for reduced arterial blood volume[104]. All these factors can precipitate the occurrence of ischemic liver necrosis. In a large series of children with BA cirrhosis, liver ischemic necrosis was reported to occur in 6% of patients[105]. This entity can precipitate liver failure and require urgent LT[106]. Due to precarious hepatic blood supply in such BA patients, cautious use of diuretic treatment, aggressive treatment for dehydration, and close monitoring of portal and arterial flows at Doppler ultrasound are recommended[106].

Concerning liver microcirculation in children affected by BA, there are some published studies describing HA anomalies[107-109]. Stowens et al[107] first described the hyperplastic and hypertrophic changes in branches of the HA in the intrahepatic portal areas of patients with BA. Ho et al[109] subsequently analyzed wedge liver biopsies in 11 cases of BA, and they indeed described arteriopathy manifesting as hyperplasia and hypertrophy of the HA (with an hypertrophy of the tunica media) from the trunk of the common HA to its peripheral branches supplying the entire biliary tree.

On the other hand, intrahepatic PV was reconstructed from serial histologic sections of surgical liver specimens from patients with BA, and a remarkable decrease in the number of PVs was already apparent as early as the second month of age, as compared to an ordinary liver[92]. Additionally, our group demonstrated the existence of morphological alterations of intrahepatic PV with critical narrowing of the intralobular PV lumens[59]. Therefore, BA is characterized by the presence of specific micro- and macro-vascular liver hemodynamic changes at a very young age.

Systemic hemodynamics in children affected by BA: BA is an isolated finding in 80%-90% of infants and occurs in association with other congenital abnormalities in up to 20% of all cases. A spectrum of congenital heart defects occurs in children with BA splenic malformation, with a reported incidence of 45%. The major abnormalities include Tetralogy of Fallot, hypoplastic left heart syndrome, aortic arch anomalies, and complex cardiac defects[110,111]. Additional anomalies, like atrial septal defect, ventricular septal defect, patent ductus arteriosus, patent foramen ovale, pulmonary artery stenosis, and dextrocardia with situs inversus, have also been reported. Interrupted inferior vena cava with azygous continuation commonly occurs, with an incidence of 65% in children with polysplenia[112]. Although the functional significance of this anomaly is minimal, it is important information when LT is considered. Children with BA splenic malformation also have a higher incidence and earlier development of hepatopulmonary syndrome[69,113].

Besides the description of the possible cardiac malformations associated with BA, there are very few publications about systemic hemodynamic disturbances in children affected by BA, and it is unknown if cardiac structural, functional, and electro-physiological abnormalities exist in infants with BA. Only one study has shown abnormal 2-dimensional echocardiographic parameters in 70% of BA patients listed for LT[4].

FUTURE RESEARCH PERSPECTIVES?

Future research should aim to better characterize hepatic and systemic hemodynamics in cirrhotic children, notably through non-invasive Doppler ultrasound, as pretransplantation monitoring. Integrating these parameters into the PELD score could improve transplant prioritization for children with poor liver perfusion. Further studies are needed to explore post-transplant flow hemodynamics—particularly the effects of splenic artery ligation on HA and PV flows—to clarify HABR in a recently implanted liver graft. These insights could support the use of intraoperative flow measurements and modulation to optimize graft function and patient outcomes after pediatric LT.

CONCLUSION

Numerous studies of cirrhotic adults have demonstrated that portal hypertension and cirrhosis are associated with hemodynamic changes in hepatic and systemic circulation, and these findings had important clinical repercussions on the pretransplant management of these patients. However, knowledge regarding the corresponding hemodynamic alterations in children with cirrhosis is limited. A better understanding of these liver and systemic hemodynamic features could contribute to a better clinical management of cirrhotic children. In this review, we demonstrated that liver hemodynamics are impaired in cirrhotic children, with a severe native liver hypoperfusion. These hemodynamic alterations seem to be correlated to the clinical severity of cirrhosis and to the histological severity of fibrosis. Cirrhosis in children may therefore also be considered as a hemodynamic disease of the liver. These hemodynamic characteristics must be included in the management of cirrhotic children and in the evaluation of the degree of emergency for LT. In contrast to adults, cirrhotic children do not seem to develop hyperdynamic circulatory syndrome. However, it is important to acknowledge that many studies on pediatric cirrhosis are small and retrospective, which limits the generalizability of their results. Nevertheless, these studies still provide valuable insights into the disease's hemodynamic features. Further studies including larger cohorts of patients are necessary to better understand systemic hemodynamics in cirrhotic children before, during, and after liver graft implantation.