State of the art of the Fontan strategy for treatment of univentricular heart disease

Physiology

In unpalliated univentricular CHD, cyanosis occurs because of mixing of saturated and unsaturated blood in the heart. The SV is also exposed to volume overload as it drains both systemic and pulmonary venous return at the same time. The Fontan strategy reduces volume overload and restores normoxemia. Following PCPC surgery, some volume unloading of the SV is achieved. Following TCPC, the volume load of the SV is further reduced¹⁷. Furthermore, after TCPC, the systemic and pulmonary blood flows are connected in series rather than parallel as before this strategy has been deployed. This comes at the expense of the lack of a ventricle supplying energy to the pulmonary circulation. This is illustrated in Figure 2. The SV provides the energy needed to attain blood flow through the systemic as well as the pulmonary vascular bed and is subjected to increased afterload. After TCPC, central venous pressures are higher than normal. Pulsatility in the pulmonary artery is mostly lost and there is preload insufficiency of the SV. This highly abnormal circulation is called the Fontan circulation. The resulting physiology has been referred to as a “Fontan paradox”, where systemic venous pressure is high in the presence of relative pulmonary artery hypotension¹⁸. This might augment lymphatic outflow and impede lymphatic inflow from the thoracic duct. Several complications of the Fontan strategy have been linked to abnormalities in lymphatic drainage. Because of the multiple inherent hemodynamic challenges of the Fontan circulation, it is generally considered a palliative procedure¹⁹.

Figure 2.

Scheme of pressures in the normal circulation (A) and the Fontan circulation (B). This scheme illustrates the effects of the lack of a prepulmonary pump in the Fontan physiology. Red represents oxygenated blood and blue represents deoxygenated blood. Ao, aorta; CV, caval veins; LA, left atrium; LV, left ventricle; P, pulmonary circulation; PA, pulmonary artery; RA, right atrium; RV, right ventricle; S, systemic circulation; V, single ventricle. This figure has been reproduced with permission from Gewillig and Brown and the British Medical Journal Publishing Group Ltd²⁰.

The aim of this article is to provide an overview of current outcomes, treatment options, and remaining challenges to improve outlook for patients with univentricular heart disease.

Overall survival

Survival following the Fontan procedure has increased dramatically in the past few decades. We will discuss data from recently published reports of large cohorts with long follow-up intervals. An overview of studies assessing survival is presented in Figure 3, obtained from Kverneland et al.²¹.

Figure 3. Survival following Fontan completion.

Each line represents a study assessing survival at multiple time points and is colored by surgical era. Dots represent Kaplan-Meier survival estimates. Studies marked with an asterisk show survival curves for death, transplant, or Fontan revision; others show survival curves for only death. This figure has been reproduced with permission from Kverneland et al. and John Wiley and Sons²¹.

In a recent study from Oceania, perioperative mortality decreased from 8% between 1975 and 1990 to 1% in 2001–2010²². In this cohort, early Fontan takedown occurred in 2% of patients. Ten-year survival among patients discharged with a Fontan circulation was 89% following APC and 97% for both ECC and ILT²³. Survival at 25 years was 76%. This group was composed only of APC patients.

In a retrospective study from the Mayo Clinic of 40 years and 1,052 Fontan patients, overall survival rates were 74% at 10 years, 61% at 20 years, and 43% at 30 years²⁴. Survival was significantly higher in later surgical eras, and 10-year survival was 95% for patients operated on after 2001. Interestingly, patients operated on with the ECC technique showed better overall survival over ILT. It should be taken into account that patients with cardiac defects with worse prognosis were more frequently operated on using the ILT technique.

A Danish national registry described outcomes for SV patients from 1977 to 2009¹. Fifty percent of patients died before Fontan completion. Overall survival improved in later birth eras. Five-year survival for any univentricular CHD increased from 22% in 1977–1989 to 51% in 2000–2009.

One should note that follow-up studies after Fontan procedures reflect results of an earlier surgical era and do not necessarily represent the outlook of Fontan patients undergoing surgery in the current era.

Determinants of survival

Several factors have been associated with survival after the Fontan procedure. Of preoperative factors, male gender²³ and specific CHD diagnosis, most strikingly for HLHS²⁴, have been associated with worse long-term survival. Perioperative risk factors for mortality include APC type procedure, earlier surgical era, older age at procedure, concomitant valve replacement, and prolonged postoperative pleural effusion²⁵. Postoperative factors that affect survival include elevated central venous pressure and lower arterial saturation²⁶; imaging-derived parameters such as lower global longitudinal strain on echocardiography and higher end-diastolic volume measured by magnetic resonance imaging (MRI)²⁵; peak heart rate and peak oxygen uptake during cardiopulmonary exercise testing^25,27,28; and serum levels of sodium, creatinine, and brain natriuretic peptide^25,29,30.

Cardiac complications

Although survival following Fontan completion has increased over time, the Fontan strategy has been associated with important morbidity, most likely related to extensive surgical procedures and the highly abnormal circulatory state after these procedures. In recent cohorts, event-free survival has ranged from 59 to 81% at 15 years^23,31.

The most commonly reported event is cardiac arrhythmia. Supraventricular tachycardia (SVT) contributes significantly to late mortality in Fontan patients³². SVT has been associated with Fontan type; the highest risk is for APC type, followed by ILT, and the lowest is for ECC^23,33. Results may have improved for the ILT technique after introduction of the prosthetic modification³⁴. The incidence of SVT increases during follow-up, and after 20 years of follow-up, 10 to 60% of patients have experienced some form of SVT^34,35.

Failure of the Fontan circulation can occur even in long-standing uncomplicated Fontan circulations. The definition of Fontan failure varies but generally includes excessive constitutional limitations of the Fontan physiology, abnormal parameters of hemodynamic function, poor functional status, or the presence of Fontan sequelae³⁶. Fontan failure occurs in 2 to 13% of patients, depending on definitions and follow-up period³⁶. A 56% 25-year freedom of Fontan failure for APC connections has been reported²³.

Extracardiac complications

Abnormalities of coagulation. Blood flow can be slow in the Fontan circuit because of the absence of a prepulmonary pump. This promotes coagulation. Furthermore, coagulation factor abnormalities have been described in the SV patient, even before PCPC palliation³⁷. Thromboembolic events are more common following APC surgery than TCPC³⁷. The prevalence of thrombi is particularly high in those who develop atrial arrhythmias³⁸. Up to 65% of thrombi are detected within the first year following TCPC (early thromboembolic events)³⁹. Early thromboembolic events probably relate to perioperative factors, such as extracorporeal membrane oxygenation (ECMO) use and altered hemodynamics. Beyond 10 years after Fontan completion, incidence of late thromboembolism steadily increases³⁹. Reports of the incidence of thromboembolism are complicated by different definitions and methods of detection. Studies report incidences of late thromboembolism of between 1 and 25%^31,40. Silent thrombi are detected with transesophageal echocardiography in up to 33% of patients⁴¹. Thromboembolic events have been reported to account for 8% of Fontan deaths⁴².

Liver function abnormalities, cirrhosis, and even hepatocellular carcinoma have been reported following Fontan surgery⁴³. The elevated systemic venous pressures encountered in the Fontan circulation lead to chronic venous congestion in the liver. However, the exact mechanism for Fontan-associated liver disease remains unknown, as does how best to monitor progression^44,45. The American College of Cardiology recently provided a position statement on the subject, advising laboratory and imaging screening at least every 3–5 years in children and at least every 1–3 years in adult Fontan patients⁴⁶. Preventive strategies need to be developed.

Protein-losing enteropathy (PLE) is a devastating complication of the Fontan circulation in which loss of protein in the gastrointestinal tract occurs, leading to low albumin levels, edema, pleural effusions, and ascites. Incidences of between 3 and 29% have been reported^24,31,35,47. It is thought to be caused, in part, by impairments in the lymphatic drainage. Hepatoduodenal lymphatic connections exist in some patients as a normal anatomic variant, which might induce competition in lymphatic flow in the presence of elevated venous pressure, diverting lymphatic flow to the gastrointestinal tract.

Plastic bronchitis is a serious pulmonary complication of the Fontan circulation in which large gelatinous casts are formed in the airways. It is thought to be related to abnormal lymphatic drainage directly into the airways, resulting in cast formation. Reported incidences range from 0.5 to 4%^48–50.

Chronic kidney disease is estimated to be present in up to 50% of adult Fontan patients and is associated with adverse outcomes^30,51. With the increasing availability of cystatin C-determined glomerular filtration rate (GFR), a muscle mass-independent estimate, the prevalence of clinically significant renal dysfunction appears to be lower⁵². GFR estimates based on serum creatinine concentrations probably overestimate renal function in Fontan patients⁵¹.

Psychological, psychiatric, and cognitive defects have been described in Fontan patients. Cerebral MRI has shown morphological differences in some cerebral structures of Fontan patients compared with those of healthy controls^53,54. Interestingly, the pituitary gland, supplied by a portal venous system similar to the liver, appears to be enlarged following Fontan surgery⁵⁵. The relevance of this possible congestion on the endocrine system remains unclear.

Re-interventions

Many patients require additional surgical and catheter-based interventions following Fontan completion. Twenty-year freedom of re-operation following TCPC procedures in recent eras ranges from 86 to 92%⁵⁶. In older cohorts, higher re-operation rates have been reported^24,57,58. The most common surgical re-intervention procedures, in order of incidence, are pacemaker implantation in 9 to 23% of patients^23,24, Fontan revision or conversion in 3 to 18% of patients^9,24,59, and atrioventricular (AV) valve repair in 1 to 14% of patients^24,59,60.

Fontan conversion, from APC to TCPC, can improve functional status and exercise tolerance in the failing Fontan circulation⁶¹. AV valve repair after Fontan completion is considered in patients with moderate to severe regurgitation, but survival following successful repair remains inferior to that of patients without prior AV regurgitation^62,63.

For several reasons, re-interventions by catheter may be required. A fenestration in the atrial tunnel, inducing a right-to-left atrial shunt, can be created during surgery to decrease systemic venous pressure, increase ventricular preload, and improve cardiac output at the cost of lower arterial saturation. This fenestration sometimes closes spontaneously or can be closed via catheter at a later time⁶⁴.

Hemodynamically significant obstruction in the Fontan pathway may occur, most commonly in the left pulmonary artery^57,65. In the absence of a prepulmonary pump, this can severely affect the Fontan circulation, and obstructions are routinely dilated or stented.

Systemic to pulmonary venous collaterals can produce a right-to-left shunting which often worsens in time. Coiling of these collaterals is routinely performed in some centers, although no survival benefit has been demonstrated⁶⁶.

Aortopulmonary collaterals are common in the Fontan circulation. They increase pulmonary blood flow but induce a volume overload on the SV and might increase pulmonary artery pressure, limiting flow from the caval veins⁶⁵. During exercise, aortopulmonary flow increases, possibly augmenting loading conditions of the ventricle⁶⁷. A large aortopulmonary collateral burden has been associated with worse short-term outcomes⁶⁸. No clear consensus regarding the long-term effects and management of these collaterals exists⁶⁵.

Catheter ablation of an arrhythmogenic substrate is common. Long-term success rates vary between 15 and 72%^69,70.

The reported incidence of catheter-based interventions (that is, excluding diagnostic cardiac catheterization without intervention) varies heavily; 3 to 65% of patients require at least one additional catheter intervention following Fontan completion^{31,56,57,59,71}. The most common catheter interventions are fenestration closure (10 to 64% of patients with a fenestration require catheter-based fenestration closing^56,57,59,72), occlusion of veno-venous or aortopulmonary collaterals (incidence of 10 to 20%)^57,59, and stenting and dilation of (all types of) obstructions in the Fontan pathway (incidence of 6 to 19%)^57,59,73.

Other outcomes and functional status

Most studies report a diminished quality of life in Fontan patients compared with healthy controls^74–78. Physical and emotional functioning are the most severely affected domains^74,79. Low perceived health status can lead to unnecessary restrictions in daily life. Furthermore, increased rates of developmental disorders and lower intelligence scores have been reported in the Fontan population^53,76.

Fontan patients have a moderately decreased exercise capacity compared with healthy controls⁸⁰. Mean peak oxygen uptake ranges from 61 to 74% of predicted values^80–82. A small fraction of Fontan patients have normal exercise capacity⁸³. Exercise capacity in the Fontan patient has been shown to decrease over time^81,82. Exercise capacity is predictive of hospital admissions, quality of life, and late mortality^25,28,79,82. Exercise training can be done successfully in Fontan patients and can improve quality of life, functional class, and health perception in a short-term follow-up^84–87. Whether exercise training has a role in optimizing long-term outcome is currently not clear^84,86–89. Resistance training can be used to increase muscle mass. In the Fontan patient, this could augment peripheral venous return, augment ventricular preload, and improve cardiac output⁹⁰. Similarly, a benefit of inspiratory muscle training has been demonstrated⁹¹.

Despite high morbidity and suboptimal outcomes, most patients with a well-functioning Fontan circulation manage to lead fulfilling lives, are employed, may attain academic achievements, can participate in sports, and are able to successfully carry pregnancy to term^92–94.

Assessment techniques

Fontan patients are routinely assessed for health and functional status. Diagnostic cardiac catheterization has been standard practice in the pre-TCPC evaluation, as it provides excellent anatomic and necessary hemodynamic information regarding the pulmonary artery pressure, pulmonary vascular resistance, and end-diastolic SV pressure⁹⁵. There is recent interest in omitting cardiac catheterization in the pre-TCPC assessment for low-risk SV patients^95–97. No consensus regarding this policy has been reached, as long-term outcomes are currently unavailable. Catheter-based interventions are discussed above.

Cardiac magnetic resonance imaging (CMR) is routinely performed during follow-up after TCPC, particularly to assess ventricular size and function and to quantify large vessel flow, including the amount of collateral flow^98–102. Death and (being listed for) heart transplantation have been associated with higher end-diastolic volume index (EDVi) (>125 mL/m²) as assessed with CMR in adolescents with a Fontan circulation^103,104. Combined with a computational fluid dynamics approach, CMR might provide very useful information on the Fontan circulation and can aid in the evaluation of modifications in treatment strategies^104,105.

Echocardiographic strain measurements have been shown to predict survival in the Fontan population and predict length of hospital stay following TCPC^104,106. Assessment of ventriculo-arterial (VA) coupling may be another important parameter, as it is independent of the (often impaired) ventricular load. VA coupling has been shown to be suboptimal in some Fontan populations¹⁰⁷. Currently, VA coupling has not been associated with long-term outcomes in the Fontan population.

Lymphangiography could play an important role in the management of Fontan complications with a suspected lymphatic pathogenesis, such as PLE and plastic bronchitis. In patients with PLE or plastic bronchitis, increased diameters of major lymphatic vessels have been noted¹⁰⁸. Abnormal lymphatic depositions in the lungs and liver have been visualized in patients with plastic bronchitis and PLE, respectively^109,110.

Medical therapy

Anticoagulation, in the form of anti-platelet drugs or vitamin K antagonists (VKAs), is commonly indicated considering the increased risk for thromboembolic events, as discussed above in the “Extracardiac complications” section¹¹¹. A meta-analysis by Alsaied et al. showed that both acetylsalicylic acid and VKA were equally effective in preventing thromboembolic complications¹¹². However, if international normalized ratio (INR) is not properly controlled, outcomes on VKA are worse compared with acetylsalicylic acid¹¹³. Novel oral anticoagulants (NOACs) do not require frequent monitoring and have mostly outperformed VKA in the adult population. Thirty-day outcomes following NOAC initiation show no major adverse events in the adult CHD population¹¹⁴. However, no NOAC agent currently has US Food and Drug Administration approval for use in children.

Medical prevention of circulatory failure in the Fontan circulation

Various medications have been assessed in the management of Fontan failure. No studies have shown benefit of angiotensin-converting enzyme (ACE) inhibitor therapy on survival, ventricular function, or cardiopulmonary exercise outcomes^115,116.

Vasodilator drugs have been used to lower pulmonary vascular resistance^117,118. Sildenafil has increased ventricular function, exercise capacity, and New York Heart Association (NYHA) status after 6 weeks of follow-up^119,120. The effects of bosentan, an endothelin antagonist, in the Fontan population have varied^121–126. No long-term survival benefit of vasodilator therapy has yet been demonstrated¹¹⁹.

Mechanical circulatory support for the failing Fontan circulation

The failing Fontan circulation can be supported by mechanical assist devices. Despite increasing use and the development of novel devices specifically for the pediatric and CHD population, experience in this population is still limited¹²⁷. Mechanical support devices are mostly used as a bridge to transplant in the failing Fontan¹²⁸. Recent reports showed a 60% 12-month survival in 48 Fontan patients with a ventricular assist device, proving viability of longer mechanical circulatory support^129,130. A total biventricular artificial heart, the SynCardia, has been used to bridge a failing Fontan patient to transplant¹³¹. A registry of mechanical circulatory support specifically for SV patients has been initiated¹³². Currently, mechanical circulatory support in Fontan patients is associated with worse survival compared with mechanical circulatory support patients with a biventricular circulation^133–135.

Cardiac transplantation

Cardiac transplantation is the only treatment that truly corrects Fontan physiology, and it is employed in the failing Fontan circulation. In large cohorts, 1.6 to 3.6% of patients ultimately underwent cardiac transplant^23,24. Survival following cardiac transplantation in Fontan patients is generally worse compared with other types of CHD^136,137. Five-year survival ranges from 60 to 67%^136–138.

Surgical innovations

Continual efforts are made to improve the surgical techniques used in Fontan surgery. Recently, a Y-shaped graft was proposed for the connection of the inferior vena cava to the left and right pulmonary artery¹³⁹. Theoretically, this graft is more energetically favorable and provides better distribution of hepatic blood flow between the left and right pulmonary artery, distributing “hepatic factors” that may prevent the formation of intrapulmonary collaterals more equally. Worse energetic performance and pulmonary flow distributions in comparison with ECC connections have been noticed in practice^140,141.

Fontan completion without cardiopulmonary bypass, particularly with the ECC technique, is an attractive option. However, experience is still limited and reported rates of conduit replacement and outcomes following off-pump procedures differ across centers^142–145.

Less-invasive surgical approaches such as lateral thoracotomy have been described in this population¹⁴⁶. Hybrid procedures, which combine transcatheter and surgical approaches, have been implemented in the initial management of HLHS¹⁴⁷. Long-term outcomes are favorable, and some centers have adopted this hybrid approach as the standard for selected patient populations¹⁴⁸.