Systolic ventricular function is typically normal at the time of Fontan but worsens over time (15, 23, 26). In one large cohort of Fontan patients followed for over 20 years, it was found that Fontan failure occurred at a median of 18.1 years after Fontan completion and 46% of these patients had systolic dysfunction. Fontan failure was more commonly associated with a decline in ventricular function than increasing pulmonary vascular resistance (15). About 50% of patients with a Fontan circulation that are listed for transplant have ventricular dysfunction (27–29).

The underlying cause of the decline in systolic function over time in patients with a Fontan circulation is not clear. During the earlier phases of staged palliation, the ventricle is volume loaded and exposed to chronic hypoxemia, which may predispose to dilation and later dysfunction, although studies have demonstrated that ventricular size and hypertrophy improve initially following the Fontan operation (1, 30). Hypoplastic left heart syndrome (HLHS) confers higher risk of systolic dysfunction because the right ventricular morphology is not embryologically or structurally designed to pump against systemic afterload (10, 26). Myocardial fibrosis has also been seen in this population and is associated with lower mean ejection fraction (31). Fibrosis and abnormal myofiber orientation may predispose to mechanical dyssynchrony over time, which has been demonstrated to be associated with reduced systolic function (32–34). Further, the Fontan circulation is a chronically preload deprived state, which may lead to decreased contractility due to chronic lack of fiber stretch and increased muscle stiffness (1).

Patients with a Fontan circulation are particularly affected by systolic dysfunction due to their baseline low cardiac output state. Additionally, preload is further reduced by loss of the normal suction force which occurs with descent of the atrioventricular valve towards the apex of the heart in systole. This force drives blood through the pulmonary vasculature in the normal circulation. The effect of this force becomes relatively more important in patients with a Fontan circulation who do not have a subpulmonary ventricle but is lost in patients with significant systolic dysfunction (35). There has additionally been recent interest in whether atrial function predicts exercise parameters and Fontan outcomes, although atrial function has not been comprehensively studied and the published data is mixed (36–38).

While systolic dysfunction is typically diagnosed by echocardiogram, there are challenges in using echocardiography to assess systolic function in patients with a Fontan circulation, particularly in patients with a systemic right ventricle (39, 40). There is evidence that utilization of cardiac MRI improves risk stratification in patients with a Fontan circulation and thus many centers will use a combination of echocardiography and cardiac MRI to monitor patients for systolic dysfunction over time (40). In a large study of patients with a Fontan circulation who underwent cardiac MRI, lower end-diastolic volume was the strongest predictor of transplant-free survival. For patients with ventricular dilation (end diastolic volume ≥156 ml/BSA), worse global circumferential strain was particularly predictive of death or transplant (41). On multivariable analysis, end-diastolic volume, end-systolic volume, global circumferential strain, and ventriculoarterial coupling ratio by MRI have been shown to be independently predictive of death or transplant. These measurements are thought to be more useful than ejection fraction because ejection fraction does not account for ventricular dilation. Cardiac MRI can also identify cardiac fibrosis which may contribute to dilatation and dysfunction, although additional work is needed to further elucidate the role cardiac fibrosis may have in this process (42). Further, stress MR to identify patients with decreased ability to augment their ejection fraction in response to dobutamine-stress may also help to identify patients at risk of adverse outcomes (43).

Diastolic dysfunction is increasingly being recognized in the Fontan population, affecting approximately 70% of patients, including >80% of patients with a systemic right ventricle (1, 12, 26, 34, 44–46). The single ventricle has multiple possible underlying etiologies for diastolic dysfunction, including dyssynchrony, myocardial scarring, abnormal geometry, increased wall stress, hypoxemia and history of volume loading prior to the Fontan (1, 45). There is evidence of early relaxation abnormalities leading to prolongation of isovolumetric relaxation which reduces early rapid filling and leads to abnormal wall motion and flow (34, 44). This abnormality is then later compounded by a slow reduction in ventricular compliance which also affects late diastolic filling (12, 44).

Increased filling pressures are not well tolerated by patients with a Fontan due to the requirement for passive pulmonary blood flow. Higher filling pressures lead to higher atrial pressures and a lower gradient for forward flow through the pulmonary vasculature. This leads to decreased preload and ultimately decreased stroke volume even without changes in systolic function (45).

Despite the high prevalence of diastolic dysfunction is in this population, it can be difficult to detect, particularly in early stages. Standard echocardiographic Doppler and tissue Doppler parameters are frequently abnormal but have been shown to correlate poorly with clinical outcomes and only modestly with catheterization measurements (46, 47). Cardiac catheterization is considered the standard for measurement of diastolic filling pressures but is invasive and can miss patients with early-stage diastolic dysfunction because it only measures filling pressures in the resting state (45). Speckle tracking echocardiogram has shown some promise in evaluation of diastolic function, but further work is needed to validate this method in single-ventricle patients (48).

Atrioventricular (AV) valve failure, defined as requiring valve repair or replacement or development of moderate or greater AV valve regurgitation (AVVR), is seen in approximately 7% of patients with a Fontan circulation at 5 years, 12% at 10 years and 21% at 20 years (20). AVVR in patients with a Fontan circulation may be secondary to annular dilation, leaflet dysplasia, leaflet tethering, leaflet prolapse, or abnormal subvalvar apparatus (49–51). Patients with hypoplastic left heart syndrome are at particular risk for developing AV valve failure, as systemic right ventricular morphology has been found to be an independent risk factor for AVVR (20, 26). There has also been a demonstrated association between older age at Fontan and more severe AV valve regurgitation (26).

AV valve regurgitation is poorly tolerated in this population as the regurgitation leads to an abnormal feedback loop of dysfunction, worsening dilation and worsening AVVR (50). This further decreases cardiac output and increases atrial pressure, which in turn lead to symptoms of low cardiac output or congestion. Significant AVVR also interferes with interpretation of standard indices of ventricular function, such as ejection fraction, and thus practitioners may underestimate the degree of systolic dysfunction in a single ventricle patient with AVVR (52). Furthermore, AVVR interferes with the aforementioned forward sucking force on the pulmonary venous blood during ventricular systole that becomes relatively more important in single ventricle patients (35). Given this, it is unsurprising that a patient with a Fontan circulation with AV valve failure has been found to have more than double the risk of Fontan failure (20).

Multiple imaging modalities can be utilized in the diagnosis and management of AV valve regurgitation in Fontan patients (53). Though 2D transthoracic echocardiography remains the most common imaging modality used, it is less sensitive to the diagnosis of specific etiologies such as leaflet dysplasia and structural abnormalities. The recent development of 3D echocardiography has addressed some of these limitations, by improving assessment of valve morphology, geometry and subvalvar apparatus. Cardiac MRI can evaluate flow and quantify AVVR but can be limited by artifact in patients with stents or devices. Alternatively, computed tomography can be considered in patients with metallic stents or pacemakers that do not allow for cardiac MRI.

While all patients with a Fontan circulation have lymphatic congestion secondary to the obligate increase in central venous pressure, a subset of patients will develop leakage of chyle into the bronchial, pleural, peritoneal, or intestinal cavities. This results in plastic bronchitis, chylous effusions, chylous ascities, or protein losing enteropathy (1, 14). It is estimated that protein losing enteropathy (PLE) effects between 4% and 13% of patients after Fontan (23, 54, 55). Plastic bronchitis (PB) effects comparatively fewer patients, approximately 2%–4%, although the burden of subclinical PB is likely significant (56, 57). Several factors likely contribute to which patients develop lymphatic complications and which lymphatic disease they manifest, including systemic venous pressures, anatomic Fontan obstruction, lymphatic vessel anatomic differences, prior interventions on the thoracic duct, concomitant cardiac dysfunction, increased mesenteric vascular resistance, proinflammatory state, and genetic differences (1, 58–64). A classification scheme to describe neck and thoracic lymphatics abnormalities has been developed which has been shown to be associated with adverse surgical outcomes after Fontan including mortality, duration of effusions and duration of hospital stay (62). While not a statistically significant association, possibly due to a small sample size, plastic bronchitis and need for transplant within 3 years only occurred in patients with the most severe type 4 abnormalities.

These conditions, which collectively can be referred to as lymphatic insufficiency, can lead to malnutrition, edema, immune deficiency, chronic cough, respiratory distress, hypoxemia and eventually death, although outcomes have improved in the past decade (14, 54, 65). PLE is generally diagnosed by using alpha-1 antitrypsin clearance in a 24-hour stool collection or an elevated alpha-1 antitrypsin level with serum hypoalbuminemia, symptoms of edema, and no other known cause. Imaging modalities such as lymphoscintigraphy and dynamic contrast magnetic resonance lymphangiography can also be used to support the diagnosis of PLE (64). Routine monitoring of serum albumin concentration in patients with a Fontan circulation can identify subclinical albumin loss which may indicate deteriorating hemodynamics (1). Plastic bronchitis is typically diagnosed clinically by patient expectoration of an airway cast or after bronchoscopic removal.

Arrhythmias are common in patients with a Fontan circulation. Sinus node dysfunction and supraventricular tachycardia are most common, affecting up to 60%–80% of patients with a Fontan, although exact rates vary depending on type of Fontan procedure (1). Contemporary surgical techniques such as extracardiac and lateral tunnel Fontan appear to have lower prevalence of arrhythmia as compared to the atriopulmonary type Fontan, but longer follow up is needed to evaluate these newer techniques (1, 17). Ventricular tachycardia is much less common, affecting 5%–12% of the population (66–68). It is estimated that approximately 10% of patients with a Fontan circulation will require pacemaker implantation (69).

Etiologies of arrhythmia in patients with a Fontan circulation include sinus node injury, disruption of the arterial supply of the sinus node, suture lines interfering with conduction, atrial dilation, hypertrophy leading to elevated pressures, the underlying genetic cause of the congenital heart defect causing altered tissue organization, and fibrosis secondary to injury from cyanosis or bypass resulting in altered conduction (1). Arrhythmias further worsen the limited cardiac reserve in a patient with a Fontan circulation. Patients with arrhythmia are significantly more likely to develop heart failure and in one study were found to have a 6-fold increased hazard of death or transplant (19, 22). Atrioventricular synchrony appears to be particularly important to maintain hemodynamics in patients requiring pacing after a Fontan operation (69).

Given the frequency of arrhythmia in Fontan patients, Holter monitoring is recommended every 1–2 years in adolescents and adults, and should particularly be considered in patients experiencing signs and symptoms of Fontan failure (1).

A variety of catheter-based interventions can be used to improve the Fontan circulation. In patients with obstruction within the Fontan pathway or pulmonary arteries, stent angioplasty may address obstruction and improve lymphatic failure (61, 73). Patients without a patent Fontan fenestration with PLE have been shown to benefit from creation of a fenestration, although this intervention comes at the expense of worsening hypoxemia and the risks of systemic embolization of venous thrombosis (54, 65, 74).

Lymphatic interventions for patients with lymphatic disease can also be attempted. Thoracic duct embolization and selective lymphatic embolization have been used to successfully treat patients with lymphatic insufficiency (75–77). Our institution prefers to use selective lymphatic embolization as it leaves the thoracic duct intact for other possible interventions in the future (64). Further, thoracic duct embolization can actually worsen abdominal symptoms in patients with multicompartment lymphatic disease. If selective lymphatic embolization is not successful in these patients with multicompartment involvement, there has been early promise in transcatheter thoracic duct decompression, although this procedure requires considerable technical expertise (78). Finally, in atriopulmonary Fontan patients with persistent atrial arrhythmias contributing to heart failure symptoms, catheter ablation has been demonstrated to be safe and effective at reducing arrhythmia burden, although approximately half of patients will experience a documented episode of recurrent atrial arrhythmia post procedure and close to one third will require repeat ablation (79).

Surgical interventions should also be considered in a patient experiencing Fontan circulatory failure. These include Fontan takedown, which is typically used in patients in the early postoperative period, Fontan revision, used in patients with Fontan obstruction, and Fontan conversion to a lateral tunnel or extracardiac Fontan, which is used in patients with an atriopulmonary type Fontan (80). Fontan takedown is accompanied by a very high early mortality, though it may be the only option for patients who are intolerant of Fontan physiology but may tolerate Glenn physiology until a transplant can be obtained. Fontan revision for surgical relief of Fontan obstruction can improve lymphatic dysfunction and Fontan conversion is often performed in conjunction with atrial arrhythmia surgery or pacemaker implantation to improve flow dynamics, lower incidence of arrhythmia and improve functional status (65, 81). The main drawback of surgical revision in this population is the high upfront risk. Fontan conversion has an early mortality at approximately 5%–10% with significant variability in outcomes among centers (81–83). In an analysis of the Society of Thoracic Surgeons Congenital Heart Surgery Database (STS-CHSD), Fontan revision had the highest reported mortality among all adult congenital cardiac operations (83). Risk factors for worse outcomes in patients undergoing Fontan revision include older age, presence of any STS-CHSD defined preoperative risk factor including preoperative mechanical circulatory support, diabetes, hepatic dysfunction, hypercoagulable state, renal dysfunction, complete heart block, pacemaker, implantable ICD and prior cardiothoracic operation, and concomitant pulmonary artery reconstruction (82).

Surgical innominate vein reimplantation to the lower-pressure pulmonary venous atrium (“turn down”) is another strategy that has been described in Fontan patients with lymphatic dysfunction, although it will worsen cyanosis (84, 85). Alternatively, in certain subsets of patients with a borderline second ventricle, biventricular conversion with possible recruitment of the systemic ventricle before conversion could be considered (86). Finally, in single ventricle patients with significant AV valve regurgitation, surgical intervention on the valve is an option. There is a movement in the field toward identifying the mechanism of AV valve insufficiency with earlier valve repair, perhaps prior to Fontan, and more aggressive valve replacement in specific situations (87–89). Although these procedures can be successfully performed, there remains a question of durability of repair (90).

Medical options include the use of many of the drugs used in biventricular heart failure, but with less evidence to support the practice. A recent systematic review of the evidence for medical therapy in prevention of Fontan circulatory failure identified only 9 studies which included only 267 Fontan patients total (91).

Diuretics are frequently used in systolic and diastolic dysfunction for symptomatic control. While they may reduce the filling pressure and improve symptoms of heart failure, they also decrease cardiac output and are often associated with electrolyte disarray and potential worsening of renal dysfunction (1). Reverse remodeling agents including renin-angiotensin-aldosterone system blockers and β-blockers are frequently used in clinical practice, albeit with little data to support the benefit of these drugs in single-ventricle patients with ventricular dysfunction or AV valve regurgitation (12, 45, 92–94). In one cohort study of patients after Fontan, 10 weeks of enalapril administration did not alter baseline hemodynamics, exercise capacity, or diastolic function and there was a suggestion of a worsening of short term exercise capacity (93). In another small cohort treated with ACE inhibitors for 6 months, no difference was seen in ventricular ejection fraction (94). Data demonstrating improved ventricular function, reversal in left-ventricular hypertrophy, and improved ventricular volume overload in patients with two-ventricle circulations with mitral or aortic valve regurgitation are often extrapolated to single ventricle patients, but there have been no studies specifically examining this effect in the patients with AV valve regurgitation and a Fontan (95–97). Use of beta blockers has also been examined in patients with a Fontan with heart failure, but a randomized control trial of beta blockade in children demonstrated that use of beta blockers may actually be harmful in patients with a systemic right ventricle (98). There is some evidence for a role for high dose spironolactone in failing Fontan patients specifically with PLE, although again the data remain limited to small series (65, 99, 100).

Modulators of pulmonary vascular resistance have also been used in this population, with some evidence supporting the practice. Phosphodiesterase type 5 inhibitors are of particular interest given their ability to decrease PVR, improve ventricular end-diastolic elastance, decrease arterial elastance, and increase ventricular end-systolic elastance, which in turn may improve preload and ventriculoarterial coupling (101). Small studies using sildenafil have demonstrated improved exercise capacity, improved exercise hemodynamics, and improved ventricular performance (102–106). Recent results from the larger FUEL trial demonstrated no significant change in peak oxygen consumption in patients with a Fontan circulation treated with udenafil. However, there was a statistically significant improvement in submaximal exercise measures including oxygen consumption, work rate, and ventilatory efficiency at the anaerobic threshold (107). Bosentan has also been examined in small scale trials and has shown mixed results in its effect on exercise capacity and New York Heart Association functional class (108–110). Finally, a single study of inhaled iloprost prior to exercise stress test demonstrated improved peak oxygen pulse and peak oxygen consumption and was particularly beneficial in patient with impaired exercise function at baseline (111). One subpopulation of patients with Fontan failure that may benefit from pulmonary vasodilation in particular are patients with lymphatic failure (65, 112).

If mediators of pulmonary vascular resistance are not effective, patients with protein losing enteropathy have several additional medical options which may be effective in some patients. Midodrine, an alpha-1-receptor agonist, has shown some promise, with a recent case series of four patients with refractory PLE demonstrating symptomatic improvement leading to deferral of transplant evaluation in two patients and transplant delisting in one patient (113). It is hypothesized that midodrine works by its effect on tone in the smooth muscle cells in the lymphatic vessels, though additional work is needed to confirm this hypothesis (113, 114). Additional medical treatments for patients with PLE include oral budesonide and octreotide, but use may not reduce the need for transplant over the long term (65, 115, 116).

Antiarrhythmics for control of atrial arrhythmias can improve systolic dysfunction and AV valve regurgitation. However, use of antiarrhythmic medications in patients with Fontan has been associated with a higher risk of recurrence and more adverse events compared to interventions such as catheter ablation or Fontan conversion (1, 117). In patients with a pacemaker, employing a pacing mode with atrioventricular synchrony has been demonstrated to improve Fontan hemodynamics (69). In Fontan patients with heart block or intact conduction with significant dyssynchrony, cardiac resynchronization therapy with dual-site ventricular pacing has shown mixed results though recent data has demonstrated improved indices of ventricular function that are less dependent upon ventricular anatomy and geometry. These improved indices also appear to correlate with improved New York Heart Association classification (118).

Finally, exercise training as therapy for Fontan patients should be considered, as it has been examined in a number of small studies with some demonstrating an increase in peak oxygen uptake, an increase in cardiac output, or an increase in quality of life with training (119). More study is needed to determine the optimal training type and program. However, if there is significant Fontan circulatory failure it is unlikely that exercise training alone will prevent the need for advanced heart failure therapies.