Article Text

Download PDFPDF
Original research article
Stenting the Fontan pathway in paediatric patients with obstructed extracardiac conduits
  1. Floris E A Udink ten Cate1,2,
  2. Uwe Trieschmann3,
  3. Ingo Germund1,
  4. Tobias Hannes1,
  5. Mathias Emmel1,
  6. Gerardus Bennink4,
  7. Narayanswami Sreeram1
  1. 1 Department of Paediatric Cardiology, Heart Centre Cologne, University Hospital of Cologne, Cologne, Germany
  2. 2 Department of Paediatric Cardiology, Amalia Children’s Hospital, Radboud University Medical Centre, Nijmegen, The Netherlands
  3. 3 Department of Anaesthesiology and Intensive Care Medicine, University Hospital of Cologne, Cologne, Germany
  4. 4 Department of Cardiac Surgery, Heart Centre Cologne, University Hospital of Cologne, Cologne, Germany
  1. Correspondence to Dr Narayanswami Sreeram, Department of Paediatric Cardiology, Heart Centre Cologne, University Hospital of Cologne, Kerpenerstrasse 62, 50973 Cologne, Germany; n.sreeram{at}


Objectives An unobstructed extracardiac conduit (ECC) is essential for optimal Fontan haemodynamics. We aimed to evaluate the feasibility and results of percutaneous transcatheter stenting of the ECC in paediatric patients with a significant Fontan pathway obstruction.

Methods Our institutional database was searched to identify all Fontan patients who had a stent placed in their ECC. Medical records, cardiac catheterisation data and echocardiographic investigations were reviewed. Vessel diameters were normalised to account for differences in body surface area.

Results Nineteen Fontan patients (age 6.5±3.2 years; male 78.9%) with a significant stenosis of their Dacron ECC graft were identified. Seven patients presented with protein-losing enteropathy (36.8%). An ECC obstruction was suspected on echocardiography in only 6/19 patients (31.6%). The mean minimum diameter of the ECC was 8.3±2.4 mm. A stenosis of >45% was seen in the majority of patients (n=12, 63.1%). Significant correlations between the severity of the ECC obstruction and Fontan pathway vessel diameters were found (all p<0.05). Stenting was successful in all children. The ECC diameter increased significantly after stenting (p<0.0001). An acute clinical benefit of ECC stenting was observed in 18/19 (94.7%) patients. ECC patency was good during a mean follow-up of 1.8±0.9 years.

Conclusions The feasibility and acute results of percutaneous transcatheter ECC stenting are promising and may provide a good alternative to postpone surgery to a later age. The mechanisms contributing to the development of ECC stenoses are likely multifactorial.

  • Fontan – total cavopulmonary connection – extracardiac conduit –stenosis – protein-losing enteropathy

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


The total cavopulmonary connection (TCPC) is a staged surgical palliation used to treat patients with single ventricle physiology.1 An unobstructed blood flow throughout the venous systemic connections and the pulmonary arteries is essential for optimal Fontan haemodynamics.2–5 The extracardiac conduit (ECC) is currently the preferred surgical strategy to establish the Fontan circulation.6 7 In most institutions, the Fontan procedure is performed at a relatively young age, making long-term patency of the ECC in the growing child a continuing concern.8 9 Significant stenosis of the ECC has increasingly been reported both early and late after the Fontan procedure.4 5 10–16 We and others have postulated that stenting the obstructed ECC may provide a promising alternative to surgery.4 5 10–13 However, no study has systematically assessed the results of this procedure, and reported data have been limited to case reports and small case series.4 5 10–13

This study sought to evaluate the feasibility, results and technical aspects of percutaneous transcatheter stenting of the ECC in patients with a significant Fontan pathway obstruction. In addition, we aimed to describe the clinical presentation and the morphological characteristics of these obstructions.


Study design

Our institutional database was searched to identify all Fontan patients undergoing cardiac catheterisation for attempted stent placement in the ECC. The medical records were reviewed for cardiac diagnosis, demographics, age at Fontan completion and clinical presentation. Specific information was obtained about interventional procedural data including conduit size and material, time from Fontan operation, previous ECC replacement and haemodynamic findings at cardiac catheterisation. Outcome data were also obtained. All precatheter echocardiographic investigations were reviewed. A diagnosis of protein-losing enteropathy (PLE) was based on recent recommendations.17 18

This retrospective study was performed according to the local code of research conduct and data protection rules. The Institutional review board granted approval for publication of the data on the condition of strict maintenance of patient anonymity.

Cardiac catheterisation

All procedures were performed under general anaesthesia, after informed parental consent was obtained. The morphology of the Fontan pathway was investigated by angiography, using the anteroposterior (AP) and lateral projections. The minimum diameter of the stenosis and the length of the stenotic segment were measured. Primary balloon dilation of the stenosis was not routinely performed. The appropriate stent was chosen depending on the nominal diameter of the conduit, and the length of the stenotic segment. No attempt was made to stent or to dilate the conduit beyond the nominal diameter at the time of surgical implantation. The particular type of stent chosen was left to the operator’s discretion. The stent was mounted on a 15 mm × 4 cm Cordis Maxi LD balloon (Cordis Cashel, Tipperary, Ireland), and delivered to the desired site through a 11F Mullins long sheath (Cook, Bloomington, Indiana, USA), using serial check injections of contrast delivered through the side arm of the long sheath to optimise stent positioning. Following initial deployment of the stent, the stent was redilated using a high-pressure balloon (Atlas balloons of 16 mm × 2 cm or 18 mm × 2 cm; Bard Peripheral Vascular, Tempe, Arizona, USA), chosen to match the nominal conduit diameter. Coumarin was stopped 1 week before, and restarted 1 day after stenting in all patients. Intravenous heparin was given during (100 units/kg up to a maximum of 5000 units) and for 24 hours following (200 units/kg/24 hours) the procedure.

ECC calculations

All cardiac catheterisation images and cine loops were reviewed. Measurements were performed in both AP and lateral projections on images with contrast (figure 1). The diameters of the inferior vena cava (IVC) and the narrowest point in the ECC were measured in both projections. The length of the stenosis was measured exclusively in the AP view perpendicular to the long axis to maximise consistency, as previously described.19 In addition, the sizes of the right (RPA) and left pulmonary artery (LPA) were obtained in the AP view. After stenting, the size of the ECC was measured again.

Figure 1

Anteroposterior and lateral views of a patient with a mid-ECC stenosis (white arrows) after contrast injection in the inferior vena cava. The locations at which measurements were made are demonstrated (right panels). The narrowest diameter of the ECC (A), length of stenosis (B) and vena cava inferior diameter (C) were measured in both views. The right (D) and left (E) pulmonary artery diameters were obtained in the anteroposterior view. Of note, a typical scooped-out appearance of the left-sided part of the ECC (*) was seen in the majority of Fontan patients with a severe ECC stenosis. ECC, extracardiac conduit.

The ECC was defined as stenosed if an internal diameter reduction of 25% or more was measured angiographically in the AP or lateral projection, or if a pressure gradient was found during catheter pullback.4 The IVC is a very compliant vessel, so that its diameter and shape are depending on patient body position, systemic venous pressure and volume status.20 21 Therefore, percent stenosis was calculated using the nominal diameter of the implanted conduit and the size of the IVC, respectively. Minimum ECC diameter was defined as the lesser of the minimum diameters measured in both angiographic projections.19

All measurements were repeated three times, and the average values were used for analysis. Vessel diameters were normalised by the square root of the body surface area (√BSA (m)) to account for differences in patient sizes, as previously described.22–24

Morphology of ECC obstruction

Special attention was addressed to the location and morphology of the ECC stenosis. For classification purposes, the ECC was divided into (1) a proximal segment, the level of conduit–IVC anastomosis and proximal 1/3 of the ECC; (2) a distal segment, level of the pulmonary artery connection and distal 1/3 of the ECC; and (3) the midconduit segment, defined as the distance between these segments. In addition, the ECC stenosis was classified as a focal or long segment lesion (extending beyond one ECC segment).

Statistical analysis

Variables were summarised using descriptive statistical methods and expressed as mean (SD) or median (range) as appropriate, depending on normality of distribution. Normality of measurements was tested using the Shapiro-Wilk test. Pearson’s correlation testing was performed to investigate associations between ECC obstruction characteristics and vessel diameters. p Values<0.05 were considered statistically significant. SPSS Version 22 (Chicago, Illinois, USA) was used to perform all statistical analyses.


Study group

Nineteen Fontan patients underwent ECC stent implantation between March 2012 and December 2015 (table 1). Two patients developed a large thrombus in their ECC within the first 12 hours following Fontan completion; this was successfully removed surgically. Another patient with PLE underwent a previous surgical redo operation of his severely obstructed ECC (16 mm Gelweave, Vascutek, Scotland, UK). The explanted Dacron graft and the histological findings are shown in figure 2. During follow-up, he once again developed a severe obstruction of his new ECC (18 mm Gelsoft Plus, Vascutek, Scotland, UK) (figure 3).

Table 1

Baseline characteristics of Fontan patients with ECC obstruction

Figure 2

Histological data of the explanted extracardiac conduit. (A) macroscopical view of the 16 mm graft showing thick neointima proliferation covering the entire inner surface of the extracardiac conduit. (B) CD 31 staining of vascular endothelial cells demonstrating endothelialisation of the neointima. (C) The neointima is covered by endothelial cells (*). The inner layer consists of primarily fibrotic tissue (**). Neovascularisation of the intima is also seen (***). The white arrows show newly formed vessels.

Figure 3

Angiographic views of a Fontan patient with unbalanced atrioventricular septal defect and heterotaxy syndrome who developed a severe stenosis of his initial ECC (16 mm) (A) and of his 18 mm ECC graft after reoperation (B). A stent was successfully placed (C). Upper panels: anteroposterior views; lower panels: lateral views. ECC, extracardiac conduit.

Indications for catheterisation

Indications for cardiac catheterisation included PLE (n=7, 36.8%), exercise intolerance (n=7, 36.8%), chronic oedema (n=3, 15.8%), failing Fontan haemodynamics (n=1, 5.3%) and cyanosis (n=1, 5.3%). A significant stenosis of the ECC was suspected on echocardiography in only 6/19 patients (31.6%). A lack of flow pattern variability during respiration, an increased flow velocity or a prominent retrograde ECC flow were specific findings in these patients. These observations were not found in the other 13 patients.

Catheterisation data before TCPC stenting

Baseline clinical characteristics and vessel diameters are presented in table 2. IVC and superior vena cava (SVC) pressure data were obtained in 14/19 (73.7%) and 13/19 (68.4%) patients, respectively (table 3). A minimal pressure gradient (1.0–2.0 mm Hg) across the Fontan ECC was found in 10/13 patients (76.9%). The mean minimum diameter of the ECC in either view was 8.3±2.4 mm (50.8%±13.6%; range: 32.5%–88.9%). Most patients (n=12, 63.1%) had a stenosis of more than 45% (percent stenosis: 25%–35%, n=1 (5.2%); 35%–45%, n=6 (31.6%); 45%–55%, n=6 (31.6%); 55%–65%, n=3 (15.8%); >65%, n=3 (15.8%)).

Table 2

Clinical characteristics and vessel diameter of Fontan patients

Table 3

Characteristics of ECC obstruction in 19 Fontan patients

Morphological classification of Fontan pathway stenosis

Five different morphological obstruction patterns were identified: mid-ECC stenosis (n=3, 15.8%; figure 1), a whole tunnel stenosis (n=1, 5.3%; figure 3), mid-distal stenosis (n=13, 68.3%; figure 4A), proximal stenosis (n=1, 5.3%; figure 4B) and a distal stenosis (n=1, 5.3%; figure 5). A focal stenosis was observed in 5/19 (26.3%) patients. These patients had either a distally, mid or proximally located ECC stenosis.

Figure 4

Different types of extracardiac conduit obstruction were encountered. Mid-distal stenosis (A) and a proximal stenosis (B).

Figure 5

A patient with a near-total occlusion of the extracardiac conduit (A). Balloon dilation (B) was performed prior to stent implantation (C).

Two additional observations are important for conceptualising a clinical framework for the pathogenesis of an ECC obstruction. First, a typical scooped-out appearance of the left-sided or right-sided part of the ECC was seen in the majority of patients (figure 1), suggesting that compression by surrounding structures (in particular the adjacent native atrium) may contribute to the development of a stenosis. Furthermore, a circumferential reduction of the internal graft diameter was observed on the angiographic images of ECC obstructions, typically extending beyond the actual stenosis (figures 1, 3 and 4).

ECC stenting

ECC stenting was successful in all patients. Twenty-one stents were implanted in 19 patients. A second stent was implanted during the same procedure in two patients. Stents deployed included Palmaz P4014 (n=17) and P308E (n=3) stents (Cordis, Miami, Florida, USA), or a CP stent (n=1) of 39 mm length (NuMed, Hopkinton, New York, USA). Primary balloon dilation of the ECC stenosis prior to stent implantation was performed in only one patient with near-total occlusion of the ECC (figure 5). The conduit diameter increased significantly after stent placement (AP view: 8.6±2.5 mm vs 16.8±1.3 mm, p<0.0001). Concomitant interventional procedures were performed in five patients (LPA stenting n=3; aortic arch dilation n=1; RPA dilation n=1; coil closure of a collateral n=1). None of the following complications was observed: damage to the conduction tissue, in particular the sinus node; compression of a pulmonary vein. None of the implanted stents crossed the midline, and the lower end of the stent was above the level of the hepatic venous confluence (in the majority of patients the stent was entirely supradiaphragmatic) avoiding potential stent compression by liver tissue.

Correlations between Fontan pathway and mid-distal ECC stenosis diameters

Most patients were found to have a significant mid-distal ECC narrowing (n=13, 68.3%). Therefore, relationships between mid-distal ECC measurements and vessel diameters were explored. Strong significant correlations between the normalised mid-distal stenosis diameter in the AP view and normalised IVC diameter in AP (r=0.734, p=0.004), normalised RPA diameter (r=0.762, p=0.002) and RPA/LPA ratio were found (r=0.623, p=0.023). Similarly, moderate to strong significant relationships between the normalised mid-distal stenosis diameter in the lateral view and the normalised lateral IVC diameter (r=0.668, p=0.013) and normalised RPA diameter (r=0.630, p=0.021) were found. These findings indicate that several geometric characteristics of the Fontan pathway, such as IVC diameter, RPA size and disparity between the pulmonary arteries sizes, might be relevant factors in the development of a mid-distal type of ECC stenosis. Moreover, the longer the time between Fontan completion and TCPC stenting, the higher the severity of the ECC stenosis was (r=−0.574, p=0.04). Patients with a severe stenosis were those with a greater length of stenosis (r=−0.669, p=0.012). The relationship between each pair of tested variables was linear.

Outcome and follow-up after stenting

An acute clinical benefit of ECC stenting was observed in 18/19 (94.7%) patients. Patients with exercise intolerance reported a greater exercise capacity after stenting. No recurrent episodes of oedema occurred in the patients with chronic oedema. They also experienced a significant increase in their exercise capacity. The cyanosis normalised in one of the patients after the procedure (SaO2 increased from 80% to 95%). Serum albumin levels and disease severity initially improved in 6/7 (85.7%) of patients with PLE.

Follow-up after stenting was 1.9±0.8 years (range: 0.5–3.0 years). One patient was lost to follow-up (moved abroad), and one patient with PLE died of a fulminant bacterial sepsis 7 months after stenting. This patient showed no improvement following ECC stenting. PLE recurred in the majority of patients (5/6, 83.3%). Only one patient experienced a continued remission of his PLE at up to 2.5 years following ECC stenting. All non-PLE Fontan patients had an ongoing improvement in their clinical and functional status. A re-stenting procedure was necessary in one patient (5.8%) with PLE, 6 months after initial stent placement. He developed a large thrombus in his ECC during a septic shock episode, despite the use of coumarin. Due to his critical illness, we decided to relieve the thrombotic obstruction with an additional stent and lysis therapy. He survived with a patent ECC.

One patient underwent angiographic re-evaluation 2.8 years following stent implantation. A patent-stented ECC without intima peel formation was seen (supplement 1).


Clinical presentation of a Fontan pathway obstruction

Significant obstruction of the ECC may develop early or during midterm follow-up after Fontan completion. All patients in this cohort had clinical signs and symptoms of a ‘failing’ Fontan circulation. ECC obstruction was not suspected on echocardiography in most patients, probably due to the distal location of the ECC obstruction in the majority of patients.

Clinical relevance of an ECC obstruction

A number of cross-sectional studies have investigated the haemodynamic impact of geometrical features of the TCPC, such as diameter, connection angle and distance between vessels, on Fontan energetics and flow dynamics.2 23 24 Yoganathan and co-workers demonstrated that indexed power loss (iPL), a flow-independent resistance index, seems a clinically useful parameter to assess the energetic efficiency of the Fontan pathway.2 23–26 Elevated iPL is associated with increased TCPC resistance resulting in energy dissipation, impaired flow distributions, decreased diastolic filling and limited ventricular preload reserve.2 23–28 The power loss in the TCPC is higher when the diameter of the Fontan pathway is smaller, resulting in less efficient Fontan haemodynamics.23–26 Recently, an inverse relationship between iPL and vessel diameter was observed in Fontan patients with a significant stenosis of their lateral tunnel (LT).23 27 The LT or ECC carries most of the total systemic return (59%±15%).1 Therefore, it is not surprising that the significant ECC obstruction found in our patients resulted in clinical signs and symptoms of a ‘failing’ Fontan circulation. Relief of the ECC obstruction was associated with an increased exercise performance, normalisation of cyanosis and chronic oedema, and an initially favourable response on PLE disease activity.

Aetiology of ECC obstruction

The tendency for intima peel formation and lack of growth of the prosthetic material used to create the TCPC are potential disadvantages of the ECC Fontan.8 10 20 24 A reduction in the internal diameter of the ECC due to neointima lining is observed in most grafts.8 20 24 29 30 A small number of studies have reported a 14%–18% decrease of the internal ECC diameter during follow-up.10 31 Nonetheless, there is a subgroup of Fontan patients who develop a significant ECC obstruction.4 5 10–16 Although several reports have suggested that the thrombogenicity and biochemical aspects of the graft material (Dacron or GoreTex, Gore & Associate, Flagstaff, Arizona, USA), the absence of graft growth, the diameter of the implanted conduit in relation to patient size, longitudinal growth of autologous vessels above or below the graft and changes in the shape of the conduit may affect ECC patency during follow-up, risk factors for the development of an ECC obstruction are not well defined.8 20 24 29 30

Histological examination of an explanted graft demonstrated a thick layer of fibrosis covered by endothelial cells. Excessive neointima proliferation of the entire graft lumen has previously been reported in stenosed ECC Dacron grafts.16 The reduction in ECC diameter typically extended beyond the actual stenosis. Moreover, a scooped-out appearance of the left-sided or right-sided part of the ECC was seen in the majority of patients. These findings suggest that (1) the endothelialisation process of the Dacron graft might be characterised by a hyperplastic endothelial response, inherent to the graft material; (2) extrinsic compression and stretching may distort the original shape of the ECC graft. In particular, the lie of the conduit alongside the wall of the native atrium, and the necessary curved course the conduit takes, may render it susceptible to compression from the atrium, particularly with serial increases in atrial diameter with age.

The most important geometric parameter that influences the haemodynamic performance of the TCPC, in the absence of LT or ECC stenosis, is the size of the pulmonary arteries.26 32 The minimum cross-sectional area of the pulmonary arteries PAs significantly correlated with cardiac output and iPL in Fontan patients. This impact becomes even stronger under exercise conditions.26 28 Interestingly, we found a strong correlation between RPA size and stenosis diameter of patients with a mid-distal type of Fontan pathway obstruction. These data suggest that PA morphology (smaller PA size, PA disparity) may significantly affect the susceptibility to ECC stenosis development.

Clinical implications

Our institutional strategy for surgical management of single ventricle patients has changed. Currently, only polytetrafluoroethylene conduits are implanted. Furthermore, 16 mm grafts are no longer in use. Insititutional policy is to implant an 18 mm diameter conduit at Fontan completion. We are prepared to delay Fontan completion to allow for this. Our current policy is to undertake MRI evaluation of the conduit at 12-month post- ECC Fontan completion, and angiographic investigation at 24 months. Further follow-up policy needs to be evaluated based on the findings at these time points.

Study limitations

This was a retrospective observational study. There was no control group and the number of patients was relatively small. Nakata or McGoon indices were not calculated, potentially limiting the comparability of geometric TCPC parameters with previous published reports. As this was a retrospective study, MRI was not considered as a technique for routine follow-up of this cohort of patients. No exercise tests were done before and after stent implantation to quantify exercise performance. However, despite these potential limitations we were able to demonstrate that the development of an ECC stenosis in Fontan patients with a Dacron graft is likely multifactorial, with PA morphology being a potential relevant risk factor.


This study shows that a significant obstruction of the ECC may develop early or during midterm follow-up after Fontan completion. A significant Fontan pathway stenosis should be excluded in all Fontan patients presenting with PLE or failing Fontan haemodynamics. Percutaneous transcatheter implantation of a stent in the obstructed ECC is feasible and safe. Patency during short-term follow-up is good. This technique may provide a good alternative to early reoperation.

Key messages

What is already known on this subject?

Most Fontan patients with a significant extracardiac conduit (ECC) obstruction undergo redo surgery. Percutaneous transcatheter stenting of the obstructed ECC may provide a promising alternative.

What might this study add?

This study demonstrates that stenting an obstructed ECC is feasible and safe, and is associated with excellent acute results and short-term stent patency. Of clinical relevance, an ECC obstruction was not suspected on echocardiography in the majority of patients. Furthermore, the development of a conduit obstruction seems not related to the type of graft material used, but it depends on the geometrical characteristics of the Fontan pathway.

How might this impact on clinical practice?

Percutaneous transcatheter stenting of an obstructed ECC in Fontan patients provides a promising alternative to early redo surgery. Echocardiographic monitoring alone does not seem to reliably exclude the presence of a severe Fontan pathway obstruction, suggesting that all Fontan patients should undergo additional imaging studies during follow-up.


The authors would like to gratefully thank Professor Dr J. Fries, Institute of Pathology, University Hospital Cologne, for kindly providing us with the images and histological analysis of the explanted extracardiac conduit.



  • Contributors All authors contributed to drafting and editing the manuscript. All authors contributed to interpretation of results, and critically reviewed and commented on the report. FU, UT, IG and NS collected and analysed the data.

  • Funding This work was supported by the Stiftung KinderHerz, Essen, Germany.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.