Notice: Wiley Online Library will be unavailable on Saturday 30th July 2016 from 08:00-11:00 BST / 03:00-06:00 EST / 15:00-18:00 SGT for essential maintenance. Apologies for the inconvenience.
Potential conflict of interest: Nothing to report.
There are approximately 1 million adult patients with congenital heart disease (CHD) in the United States, and the number is increasing. Hepatic complications are common and may occur secondary to persistent chronic passive venous congestion or decreased cardiac output resulting from the underlying cardiac disease or as a result of palliative cardiac surgery; transfusion or drug-related hepatitis may also occur. The unique physiology of Fontan circulation is particularly prone to the development of hepatic complications and is, in part, related to the duration of the Fontan procedure. Liver biochemical test abnormalities may be related to cardiac failure, resulting from intrinsic liver disease, secondary to palliative interventions, or drug related. Complications of portal hypertension and, rarely, hepatocellular carcinoma (HCC) may also occur. Abnormalities such as hypervascular nodules are often observed; in the presence of cirrhosis, surveillance for HCC is necessary. Judicious perioperative support is required when cardiac surgery is performed in patients with advanced hepatic disease. Traditional models for liver disease staging may not fully capture the severity of disease in patients with CHD. The effectiveness or safety of isolated liver transplantation in patients with significant CHD is limited in adults; combined heart-liver transplantation may be required in those with decompensated liver disease or HCC, but experience is limited in the presence of significant CHD. The long-term sequelae of many reparative cardiac surgical procedures are not yet fully realized; understanding the unique and diverse hepatic associations and the role for early cardiac transplantation in this population is critical. Because this population continues to grow and age, consideration should be given to developing consensus guidelines for a multidisciplinary approach to optimize management of this vulnerable population. (HEPATOLOGY 2012;56:1160–1169)
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
As a result of successful reparative surgery for complex congenital heart disease (CHD), approximately 85% of patients with CHD now survive into adulthood.1 Currently, the estimated number of adults with CHD in the United States ranges from 650,000 to 1.3 million, and it is expected that the number of adults with CHD will increase by approximately 5% every year.1, 2 Approximately 1 in 150 adults in the United States has some form of CHD, but the adult healthcare system is ill-equipped to address the needs of these complex patients.1, 3
Hepatic complications are common in patients with CHD either resulting from the primary cardiac defect or from palliative surgical procedures performed in infancy or childhood, or from transfusion or drug-related hepatitis. Given that such patients increasingly require the expertise of a hepatologist, the aims of this review are to examine the pathophysiology and management of hepatic disease resulting from CHD and to address issues related to cardiac surgery and organ transplantation.
There are several known associations between primary liver disease and concomitant CHD defects (Table 1). However, hepatic disease as a result of CHD is more common than cardiac disease associated with liver disease. Several CHD defects may lead to either left or right ventricular failure (Table 2). In these cases, hepatic dysfunction may ensue as a result of the primary cardiac defect or as a result of surgical palliation, especially in patients with single-ventricle physiology (e.g., tricuspid atresia). The mechanisms leading to hepatic dysfunction may be multifactorial (Table 3). As an example, hepatic dysfunction may result from a combination of passive venous congestion of the liver and hypoxia, with the latter being driven by the CHD or concomitant pulmonary disease. Volume overload and low cardiac output may lead to both congestive hepatopathy and hepatic ischemia.
Liver failure and cirrhosis in the neonatal period
25% of patients have some form of CHD, including peripheral pulmonic stenosis and TOF
Aneurysms (basilar artery, internal carotid, cerebral artery) or coarctation of the aorta; skeletal, ocular, facial, renal, and neurodevelopmental abnormalities
Presence of CHD is associated with increased mortality; 10% of deaths are caused by liver disease, and 55% of deaths are caused by cardiovascular complications.
Inflammatory and progressive destruction of the extrahepatic bile ducts
Cholestatic jaundice, biliary cirrhosis, and liver failure; HPS may be common
ASD, VSD, PDA, TOF, right-sided aortic arch, double outlet right ventricle, common atrioventricular canal, dextrocardia, pulmonary stenosis, total anomalous pulmonary venous return, and hypoplastic left heart
Polysplenia, situs inversus, intestinal malrotation, portal vein anomaly, hepatic artery anomaly, and inferior vena cava interruption
Biliary atresia is the most common pediatric etiology for LT; BASM represents a subset of these patients.
Mitochondrial fatty acid oxidation disorders
Abnormalities in fatty acid metabolism
Metabolic acidosis, hypoglycemia, and muscle weakness
Table 2. Major Types of CHD Defects Associated With Hepatic Dysfunction
Single-ventricle physiology after Fontan procedure (e.g., tricuspid atresia)
d-Transposition of the great arteries after atrial switch repair (Mustard, Senning)
Repaired TOF with pulmonary regurgitation
Pulmonary stenosis/pulmonary regurgitation
Secundum atrial septal defect with pulmonary stenosis or pulmonary hypertension
Partial atrioventricular septal defect with tricuspid regurgitation and/or pulmonary hypertension
Left ventricular outflow tract obstruction/coarctation of the aorta
Repaired complete atrioventricular septal defect
Table 3. Selected CHD Defects and Associated Liver Disease
Atrial and ventricular septal defects
ASD: Shunting may lead to right atrial and ventricular enlargement and tricuspid regurgitation as the tricuspid annulus dilates. Pulmonary hypertension occurs (5%), causing reversal of the shunt from right to left and cyanosis (Eisenmenger's syndrome).
Passive congestion/central venous hypertension
Pulmonary hypertension more commonly occurs with a VSD and has negative implications for LT if surgical repair is delayed.
Pulmonary stenosis, large VSD, overriding aorta, and right ventricular hypertrophy
Surgical repair to relieve pulmonary stenosis leads to pulmonary regurgitation, leading to right ventricular dilatation, secondary tricuspid regurgitation, and, ultimately, right ventricular dysfunction. Ventricular-ventricular interaction may lead to left ventricular dysfunction.
Passive congestion/central venous hypertension; hepatic ischemia resulting from low flow state
Failure of the tricuspid valve to delaminate from the right ventricular endocardium
Small, malfunctioning right ventricle results from defective tricuspid valve, and tricuspid regurgitation may lead to right ventricular dysfunction. Residual right ventricular dysfunction is common, even after repair or replacement of the tricuspid valve.
Passive congestion/central venous hypertension
d-Transposition of the great arteries
Right ventricle gives rise to the aorta, and the left ventricle gives rise to the pulmonary artery (atrioventricular concordance but ventriculoarterial discordance)
Atrial switch procedure (Mustard and Senning procedure): Venous blood was baffled to the left ventricle and thence to the pulmonary arteries, and the pulmonary venous blood was diverted through the right ventricle to the aorta. Ventricular failure and secondary tricuspid regurgitation result, because the morphologic right ventricle is not designed to pump at systemic pressures. Obstruction in the vena caval pathway is possible.
Passive congestion/central venous hypertension (avoided by arterial switch procedure)
Several factors may interact to lead to end-stage liver disease. For example, patients with underlying liver disease (e.g., viral hepatitis, alcohol, or obesity) may be more susceptible to liver injury as a result of decreased functional mass.4 In addition, the presence of cardiac disease and subsequent passive congestion may itself predispose the liver to hepatic injury.5 Over time, cardiac cirrhosis (i.e., central vein to central vein bridging fibrosis and nodule formation) may develop and result in portal hypertension (PH) with ascites and varices. Hepatic consequences of passive venous congestion and low cardiac output are discussed further.
Passive Venous Congestion of the Liver/Central Venous Hypertension.
Right ventricular failure is a consequence of several defects and is reflected by hepatic zone 3 sinusoidal dilation and hemorrhagic necrosis. Zone 3 necrosis may also be caused by ischemia. As an example, CHD may be associated with elevated right atrial pressure resulting from left-to-right shunting through a septal defect with secondary pulmonary hypertension, univentricular physiology (e.g., tricuspid atresia), and with a failing systemic ventricle, which is a morphologic right ventricle (Tables 2 and 3). Restrictive physiology in the right ventricle (e.g., with repaired atrial septal defect [ASD] and tetralogy of Fallot [TOF]) also contributes to passive congestion. Narrowing of the venous pathway to the lungs (e.g., Fontan operation; see below) or in the inferior vena cava (after atrial baffle procedures for d-transposition of the great arteries) may contribute to hepatic venous congestion. The most common biochemical abnormalities in passive venous congestion of the liver are elevated indirect bilirubin and prolonged international normalized ratio (INR) with minimal elevations of the aminotransferases.
Low-Output Cardiac Failure/Hypoxemia.
Patients with CHD are susceptible to ischemic hepatitis because right heart failure elevates hepatic sinusoidal pressure and reduces portal inflow. This results in increased sensitivity to any decrease in hepatic artery flow, resulting from a decrease in cardiac output (e.g., caused by concurrent arrhythmias or hypotension). For example, left ventricular outflow tract obstruction/coarctation of the aorta is associated with hypoperfusion and, in some clinical situations, may lead to hepatic ischemia.6 Chronic hepatic ischemia may also lead to hepatic fibrosis.7
Acute Cardiac Dysfunction
Hepatic disease caused by acute cardiac dysfunction results from a combination of low-output cardiac failure and passive congestion. Often, the clinical presentation may be indistinguishable from primary liver disease. For example, a marked elevation in transaminase levels characteristic of ischemic hepatitis may also be observed in patients presenting with drug-induced or acute viral hepatitis. However, a rapid reduction in aspartate aminotransferase (AST) and alanine aminotransferase (ALT) levels in the setting of an acute decrease in cardiac output/systemic hypotension suggests hepatic ischemia. Acute cardiac dysfunction is more likely to be associated with jaundice and encephalopathy, as compared to chronic or acute on chronic cardiac dysfunction.7 In acute cardiac dysfunction (e.g., ischemic hepatitis), elevations in the thousands of aminotransferase levels within 24 hours and increases in bilirubin and prothrombin time can be observed. A lag in the rise of serum bilirubin may be observed, and the elevation in bilirubin may take a longer time to resolve, as compared to aminotransferase levels. ALT levels are correlated highly with right atrial pressure, free hepatic venous pressure (FHVP), and wedge hepatic venous pressure (WHVP), but not the hepatic venous pressure gradient (HVPG) or cardiac index. Total bilirubin correlates better with HVPG. However, in persons with chronic cardiac dysfunction, a correlation of biochemical parameters with hepatic pressures is not present.
Elevation of transaminases after cardiac surgery occurs more frequently than previously reported, particularly in the setting of right-sided heart failure. Extreme elevations of ALT, AST, and lactate dehydrogenase correlate negatively with postoperative survival.8 In a single-center study that predominantly examined cases of ischemic cardiomyopathy, hepatic centrilobular necrosis, inflammation, and hemorrhage were more common in the acute group. In contrast, centrilobular and periportal fibrosis were more frequent in patients with chronic cardiac dysfunction.7
Cardiac Dysfunction After Fontan Operation
The Fontan procedure, initially described in patients with tricuspid atresia, is the most common procedure in patients with single-ventricle physiology or when biventricular repair is not feasible (e.g., double-inlet left ventricle and hypoplastic left heart).9 In patients with single-ventricle physiology, intracardiac mixing of blood leads to arterial hypoxemia with excessive volume load on a single ventricle. Hence, separation of the pulmonary and systemic circulation is desirable. The Fontan operation allows systemic venous return to the pulmonary arteries bypassing the right ventricle.
The Fontan operation (Fig. 1) may be accomplished by either creating a direct cavopulmonary (sometimes in a staged manner) or atriopulmonary anastomosis. In the first stage (i.e., superior caval-pulmonary anastomosis or bidirectional Glenn procedure), the superior vena cava is connected to the pulmonary arteries. Eventually, the inferior vena cava is also connected to the pulmonary arteries completing the circulation (Fontan completion). In earlier iterations of the Fontan, an atriopulmonary anastomosis was created with the hope that a hypertrophied right atrium would serve as a functional pump. However, it was associated with a risk of atrial dysrhythmias and atrial thrombi.10 More commonly, a cavopulmonary anastomosis is achieved by the use of an intra-atrial tunnel or patch or by utilizing an extracardiac conduit to connect the vena cavae to the pulmonary arteries (Fig. 1).
As a consequence of surgical palliation, significant liver disease can develop as a result of the interplay of central venous hypertension/passive congestion and hypoxia resulting from left ventricular dysfunction. Development of significant hepatic injury after a Fontan procedure is multifactorial. The determinants of cardiac output are central venous pressure, pulmonary vascular resistance, and systemic ventricular end-diastolic pressure. Over time, a “failure of Fontan physiology” is common. Failure of the Fontan circuit may result from elevated pulmonary vascular resistance, pulmonary thrombi, narrowing and scarring in the Fontan pathway or pulmonary arteries, and failure of the systemic ventricle, which results in elevated pressure in the pulmonary venous atrium. Chronic elevation of central venous pressure is common, and reduced cardiac output from the functioning single ventricle is frequent, particularly as diastolic and systolic dysfunction ensues. The former results from the absence of a subpulmonic pump.11 There is impaired coupling between the ventricles and the arterial system with late ventricular dysfunction.12 Atrial arrhythmias may contribute to this decline with relative hypotension and desaturation. The development of pulmonary venovenous collaterals as pressure “pop-offs” are not uncommon in the adult population and further contribute to hypoxemia. Pulmonary arteriovenous malformations, most often observed after a classic Glenn procedure, also contribute to hypoxemia. Chronic hypoxia resulting from a depressed cardiac output, in addition to the aforementioned changes, may also lead to hepatic injury.11, 13
On histology, hepatic sinusoidal fibrosis (within the space of Disse in a pericellular manner) and marked sinusoidal dilatation are universal, extending from zone 3 toward the portal tract.11, 13 Sinusoidal dilation is associated with higher right atrial pressures, similar to that observed in patients with cardiac cirrhosis.11, 13 In contrast to cardiac failure, the extent of dilation as well as fibrosis is more severe in patients with Fontan circulation.11, 13 After a failed Fontan procedure (at least in patients with cavopulmonary anastomosis), the back pressure on the liver is usually continuous (i.e., nonpulsatile), as opposed to the back pressure secondary to problems such as tricuspid regurgitation (i.e., pulsatile). This continuous high venous pressure may explain why liver dysfunction is frequent after a Fontan procedure.
The exact mechanism for the development of fibrosis with cardiac dysfunction is unknown. Fibrosis in cardiac cirrhosis or after Fontan palliation may develop independent of inflammation and, potentially, driven by repetitive mechanical stretch and compression of sinusoid and other resident cells as a result of passive congestion.14 This, along with hypoxia driven by a low cardiac output, may lead to significant structural and function alteration of the liver parenchyma.
Hepatic complications of a failed Fontan are, in part, related to the duration of follow-up.11, 15, 16 As compared to a duration of less than 5 years, the odds of hepatic complications for a post-Fontan duration of 11-15 years is 4.4 (confidence interval [CI]: 1.1-17.2) and 9.0 (CI: 2.2-36.2), for a duration of 16-20 years.15 Furthermore, the extent of hepatic fibrosis on pathological specimens is strongly correlated with elevated hepatic venous pressures (r = 0.83; P = 0.003), low cardiac index, and ventricular function.11 Hepatic dysfunction correlates best with a low cardiac index and ventricular function. After cardiac transplantation, 1-year actuarial survival is 89% in patients with preserved ventricular function, as compared to 56% in those with failing ventricular function (P = 0.04).17 Progression to cirrhosis may even be observed within 10 years of the initial Fontan surgery.18 The majority of hepatic complications are incidentally discovered. In patients with Fontan circulation followed for a median of 12 years, elevated liver function tests (30% abnormal transaminases and 32% abnormal bilirubin), coagulopathy (58%), hepatomegaly (53%), cirrhosis (26%), and hepatic masses (3%) are recognized.19 PH with gastroesophageal varices may develop, resulting in increased risk of gastrointestinal hemorrhage; hepatocellular carcinoma (HCC) may also develop.
Approach to Hepatic Disease in CHD
Abnormal Liver Biochemical Tests.
Liver function test and coagulation abnormalities (especially protein C deficiency), particularly in patients with Fontan circulation, are common.20, 21 The approach to abnormal liver function tests is similar to other patients with liver disease. However, there are salient features that may be unique to patients with CHD. Besides primary liver disease, elevated tests may be the result of cardiac dysfunction (univentricular or biventricular failure and pulmonary hypertension), related to medical treatment (transfusion-related viral hepatitis, antiarrhythmic drug toxicity, and transfusion-related iron overload) or surgical palliative correction (e.g., stenosed Fontan conduit, constrictive pericarditis, and hepatic vein thrombosis). The pattern of elevation may guide diagnostic testing. A hepatocellular pattern with elevation in aminotransferases results from low flow states and hepatic ischemia. Passive congestion is associated with isolated hyperbilirubinemia and elevated prothrombin time. In symptomatic patients with cholestatic jaundice, ischemic cholangiopathy and pigment stones should be considered. Serum albumin, unless accompanied by protein-losing enteropathy, is preserved until the onset of decompensated cirrhosis.
Complications of PH.
The development of ascites is driven by cardiac status (i.e., right heart failure), dysfunction of surgical palliation, including constrictive pericarditis, or narrowing of the Fontan circulation or protein-losing enteropathy, sinusoidal hypertension resulting from either coexisting liver disease (e.g., viral hepatitis) or cardiac cirrhosis, or portal vein thrombosis (PVT). Measurement of the serum ascitic albumin gradient and total protein on evaluation of the ascitic fluid may help differentiate between a cardiac or liver etiology for ascites.22 Often, measurement of HVPG and a transjugular liver biopsy is needed to determine whether the cause of ascites is cardiac, hepatic, or combined. Cannulation of the portal vein provides the most accurate assessment of portal pressures.7
Management of complications of PH (e.g., ascites, PVT, or variceal bleeding) follows standard guidelines.22-24 However, there are certain differences in the management of PH in CHD patients. Transjugular intrahepatic portosystemic shunt is not recommended in the presence of high right-sided pressures, given the risk for shunt dysfunction as well as the potential for abrupt increase in preload.23 Second, gastric variceal bleeding is harder to manage, because variceal obliteration with tissue adhesive may be associated with a risk of systemic emboli, given the presence of potential right-to-left intracardiac shunts.23 Finally, one may need to consider ruling out the presence of varices before the initiation of anticoagulation in CHD patients with cirrhosis.
Imaging and Liver Masses
Reticular (i.e., peripheral diffuse patchy enhancement during portal venous phase imaging) or zonal enhancement (i.e., altered enhancement of the liver periphery) is commonly observed on computed tomography scans. Zonal enhancement correlates with lower hepatic vein pressures and a lower likelihood of cardiac cirrhosis, whereas reticular enhancement is associated with extensive hepatic fibrosis. Hypervascular nodules, defined as intense vascular blushes observed during arterial phase imaging, are observed in patients with high Fontan venous pressures.11 Selected postmortem pathology reveals that these nodules are focal nodular hyperplasia (FNH).25
The presence of hypervascular nodules should be critically evaluated, especially in patients with cirrhosis. There have been increasing reports of HCC in Fontan patients with cardiac hepatopathy and correlates with the duration of the Fontan circuit (Fig. 2).11, 25, 26 In contrast to FNH, HCC may be associated with an elevated alpha-fetoprotein (AFP). The incidence of HCC in patients with CHD is likely to increase in the future, because patients survive longer.25 In the presence of cirrhosis, serial monitoring is with AFP and imaging every 6 months, with biopsy when imaging is not diagnostic.27 The risk of needle-track seeding is 2.7%.28 An arterial hyperenhancing lesion with washout of the contrast on the portal venous phase, or a mass associated with an AFP >200 ng/mL, would warrant treatment as an HCC. The use of magnetic resonance imaging to better characterize the lesions may be limited by the presence of cardiac pacemakers. Pacemakers also limit the treatment of tumors with radiofrequency ablation. Because the risk of cirrhosis increases with duration of Fontan circulation, it may be reasonable to start HCC surveillance at 10 years after Fontan completion or earlier, if there is imaging or clinical evidence of cirrhosis.
Areas of Uncertainty
Hepatic Dysfunction With CHD and Risk for Cardiac Surgery.
Nonliver transplant surgery in patients with cirrhosis can be associated with significant risk of mortality.29 The interaction between the presence of liver disease and repair of the cardiac defect is unclear. Among patients with chronic liver disease undergoing cardiac surgery (none with CHD), patients with disease of mild severity (Child-Pugh A) did well; high morbidity and mortality were observed in more advanced liver disease.30 On the other hand, in two small studies of children with cirrhosis undergoing cardiac surgery, morbidity and mortality were not inconsequential. The limited number of cases and the population characteristics preclude generalizability to adults.31, 32
Significant pulmonary hypertension and/or right heart failure may exist in patients with CHD, leading to perioperative hemodynamic instability and thus suboptimal outcomes.32 Cirrhotic patients have decreased effective circulating arterial volume, which may be further reduced by impaired venous return resulting from tense ascites and diuretic therapy.29 Postoperatively, low cardiac output may reduce hepatic perfusion, but judicious perioperative support may lead to better outcomes.29, 32 Laparoscopic procedures (e.g., cholecystectomy) may need to be avoided, given that increased intra-abdominal pressure resulting from procedural pneumoperitoneum may decrease the passive venous flow in a Fontan circulation. Whether a lower goal for insufflation (e.g., 10-12 mmHg) would be permissive for procedures is unknown.33
There are no data to predict outcomes in adult patients with CHD and liver disease undergoing cardiac surgery. It is unknown whether early heart transplant may avoid the need for liver transplant in patients that have evidence of hepatic congestion, but have not yet progressed to cirrhosis. Furthermore, it is also unclear whether liver disease, even in the absence of cirrhosis, portends increased surgical mortality. These concerns warrant further investigation.
Isolated Cardiac Transplantation.
In a single-center study, isolated cardiac transplantation after failure of Fontan procedures was associated with a 63% 1-year and 57% 5-year survival, which is approximately 12%-15% lower than the 1- and 5-year survival after isolated cardiac transplantation for other indications. Persons undergoing cardiac transplantation after Fontan had a lower survival (albeit not statistically significant) than persons undergoing a Glenn procedure.34 Deaths within 7 days of cardiac transplantation were caused by hemorrhage, sepsis, and multiorgan failure. Whether these were related to undiagnosed cardiac cirrhosis is unknown.34
The Model for End-Stage Liver Disease (MELD) score is used to both quantify the severity of liver disease as well as prioritize patients for organ allocation in the United States, with a higher MELD score portending a poor outcome.35 It is possible that hepatic synthetic dysfunction among patients with CHD may not be accurately captured by mathematical modeling. For example, INR and albumin do not correlate with degree of histological severity in patients with CHD, and hence traditional models (e.g., Child–Turcotte–Pugh [CTP] or MELD) may be inadequate.11 This gap in knowledge needs to be studied, because decisions regarding transplantation (either liver transplantation < or combined heart-liver transplantation [CHLT]) may hinge on these scores. In patients with cirrhosis of other etiologies undergoing open-heart surgery, the MELD score, CTP, American Society of Anesthesiologists class, and age are predictors of outcome.36 The mathematical risk model (http://www.mayoclinic.org/meld/mayomodel9.html) may be helpful in risk stratification in patients evaluated for repair of CHD defects. It should, however, be emphasized that derivation of the model did not include any patient that underwent surgery for CHD, and hence the above models are extrapolated with caution to patients with CHD.
Keeping these limitations in mind, Fig. 3 provides a guide for initial triage of patients with CHD and liver disease needing surgery. It represents our approach to taking care of patients with CHD and liver disease, though this has not been subject to prospective evaluation. It simply serves as a starting point for discussion in a multidisciplinary environment in evaluating the needs of the patients.
Referral to a transplant center should be initiated once there are signs of decompensated liver disease. However, the effectiveness or safety of isolated LT in patients with significant CHD is limited in adults. Among children, LT has been successfully performed in selected candidates with CHD; approximately 18% of pediatric LT candidates have some form of CHD.37 Most of the LT experience with CHD is in infants and children, and caution is warranted in extrapolating results to adults. In a selected cohort with biliary atresia requiring LT, there were no perioperative cardiac complications, although the cardiac lesions (e.g., ventricular septal defect [VSD], ASD, and pulmonary stenosis) were incidental. The overall recipient and graft survival at 1 and 5 years were both 100%.38
Manzoni et al. compared outcomes after LT in patients with cirrhosis and CHD (61% with Alagille syndrome) and a group undergoing LT without CHD (86% with biliary atresia).37 In this group of patients with “mild” deficits attributable to their CHD, rates of mortality (7% versus 8%), recovery, and retransplantation were similar. However, patients with CHD that required corrective cardiac surgery and patients with liver masses were excluded. Furthermore, only a subset of CHD defects with mild severity were included (e.g., patent foramen ovale and pulmonary artery stenosis). A separate analysis by the same group demonstrated that living donor LT can be safely performed in hemodynamically stable patients with small- to large-sized ASD. However, once again, it must be borne in mind that these were less-severe defects. Hence, the available evidence as to the efficacy of LT is patients with severe CHD remains sparse.
In adults, significant cardiopulmonary disease is a relative contraindication to LT, and the presence of significant pulmonary hypertension is associated with poor outcomes.39 In pediatric patients with hypoxemia resulting from intrapulmonary shunting, as in adult patients with hepatopulmonary syndrome, a Pa02 value <50 mmHg is associated with significant mortality (33%). The decrease in systemic vascular resistance after reperfusion may lead to further intracardiac shunting (right to left), leading to hypoxia.37 Most adult patients with failing Fontans have significantly elevated right atrial pressures. At our center, however, we consider right atrial pressures greater than 15 mmHg as a relative contraindication to isolated LT.
The potential for air embolism during the LT procedure (leading to either pulmonary embolism or paradoxical emboli and cerebral infarction) and the risk of infective endocarditis need to be considered.29, 37 Patients with CHD have tenuous hemodynamics and may be at a higher risk for hypotension, arrythmias, and bleeding. Postoperatively, hepatic congestion resulting from right-sided failure may occur, and one must bear in mind that the eventual progression of cardiac disease may ensue.38
There are limited data on CHLT in patients with congenital heart disease.40-42 The procedure has been performed in selected cases at a handful of centers and may be an option for heart transplant candidates with cirrhosis or for patients with liver failure or HCC secondary to cardiac cirrhosis. The most common indication for CHLT in the United States is amyloidosis (30%).42 Hemochromatosis (13%), hepatitis C virus (13%), and cardiac cirrhosis (11%) are the other common causes for LT in patients with cardiac disease; congenital heart disease (13%) and idiopathic dilated cardiomyopathy (13%) are the other common causes for a heart graft in patients with cirrhosis. Hence, there is only limited experience with transplanting persons with CHD and liver disease. In addition, the severity of cardiac dysfunction among the above-described cases is not known. Overall survival of patients receiving heart transplants in the United States for CHLT is 83% (3 months), 74% (1 year), and 64% (5 years), respectively. However, this excellent survival may be driven by the unique characteristics of the population. Most patients undergoing CHLT have amyloidosis, and these patients are often young to middle-aged with normal liver synthetic function and minimal coagulopathy.41 The risk of the procedure is often determined by the cardiac disease, rather than the liver disease. At our center, we have performed CHLT for 3 patients with complex CHD and cardiac cirrhosis (MELD range, 10-15) with 100% survival (range, 8 months-4 years).
In patients with failed Fontans who have had multiple transfusions, there is the risk of sensitization to donor antibodies, which makes receipt of a suitable organ challenging. The multiple sternotomies and cardiac procedures greatly increase the technical complexity of the cardiac transplant. Transplanting the liver before the heart may serve to absorb donor-specific antibodies, which can cause cardiac rejection, but places the liver at increased risk of ischemia in the absence of adequate cardiac function. In the 3 patients with CHD and cardiac cirrhosis undergoing CHLT, all of the patients were sensitized to donor antibodies; though there were episodes of acute cellular rejection, there were no episodes of antibody-mediated rejection.
Patients listed for CHLT often get transplanted based on their cardiac status, rather than the MELD score. Wait-list mortality for the average candidates listed for the CHLT dual waiting list (cardiac status 2 and MELD scores of 20-29) approximates the waiting-list mortality of those with status 1 or a MELD score higher than 30.40
After CHLT, lower immunosuppression levels are tolerated with a lower risk of graft rejection related to induction of partial tolerance.41, 42 In 93% of patients undergoing CHLT at the Mayo Clinic, both surgeries were completed in a single stage without perioperative mortality.41 As compared to a control group undergoing heart transplant alone, rejection rates were lower and pulmonary embolism was higher in the CHLT group, but survival was similar between the two groups.
Significant strides have been made in reducing mortality in patients with CHD. However, the long-term sequelae of palliative procedures in early childhood are not yet fully realized, and an increase in morbidity attributed to liver disease, especially with the associated and potentially increased risk of HCC, is expected over the lifespan of this vulnerable population. Whether earlier cardiac transplantation in patients without cirrhosis can prevent the subsequent development of liver-related complications or impede progression to advanced fibrosis remains unclear. Theoretically, earlier cardiac transplantation may be beneficial, given that the hemodynamic improvement with additional cardiac surgeries is often limited. Furthermore, each additional surgery is associated with higher transfusion requirements with negative implications for future transplants and may increase the complexity of cardiac transplantation. We anticipate that the hepatologist will play a central role in the multidisciplinary management of these complex patients. It is hoped that the formation of consortia may allow for better elucidation of the nature and frequency of complications as well as clarification of the optimal management in this vulnerable population.