ARTICLE IN PRESS+ModelACVD-764; No. of Pages 8

Archives of Cardiovascular Disease (2014) xxx, xxx—xxx

Available online at

ScienceDirectwww.sciencedirect.com

CLINICAL RESEARCH

Percutaneous valvulation of failing Fontan:Rationale, acute effects and follow-up

Valvulation percutanée des circulations Fontan défaillantes :rationnel, effet immédiat et suivi à moyen terme

Sophie Malekzadeh-Milania, Magalie Ladouceura,Laurence Iserina, Younes Boudjemlinea,∗,b

a Centre de Référence Malformations Cardiaques Congénitales Complexes — M3C, HôpitalNecker-Enfants-Malades, Cardiologie Pédiatrique, 149, rue de Sèvres, 75015 Paris cedex,Franceb Hôpital Européen Georges-Pompidou, Assistance Publique des Hopitaux de Paris, Paris,France

Received 26 May 2014; received in revised form 6 July 2014; accepted 22 July 2014

KEYWORDSMelody valve;Percutaneous valveimplantation;Fontan circulation;Failing Fontan;Congenital heartdiseases

SummaryBackground. — Fontan circulation is fragile and unfavourable evolution is frequent. Fontanphysiology largely depends on respiration and gravity. The hypothesis for valvulation is thatvalvulation of the circuit reduces the effects of respiration and the proclive position, andincreases anterograde flow towards the systemic circulation, with increasing exercise capacityand benefits for enteropathy. Because it originates from the bovine jugular vein, the Melody®

valve (Medtronic, Minneapolis, MN, USA) is naturally designed to work in a low-pressure envi-ronment.Aims. — To report our experience of percutaneous valvulation of refractory failing Fontan cir-

culation.

Please cite this article in press as: Malekzadeh-Milani S, et al. Percutaneous valvulation of failing Fontan: Rationale, acuteeffects and follow-up. Arch Cardiovasc Dis (2014), http://dx.doi.org/10.1016/j.acvd.2014.07.050

Methods. — We reviewed all patients who received a Melody valve in Fontan circulation in ourunit.Results. — Four patients were included: two had severe and refractory protein-losing enteropa-thy; one had severe oedema and ascites; and one had very severe lower limb venous

Abbreviations: IVC, Inferior vena cava; NYHA, New York Heart Association; PLE, Protein-losing enteropathy; TCPC, Total cavopulmonaryconnection; TTE, Transthoracic echocardiography.

∗ Corresponding author.E-mail address: younes.boudjemline@nck.aphp.fr (Y. Boudjemline).

http://dx.doi.org/10.1016/j.acvd.2014.07.0501875-2136/© 2014 Elsevier Masson SAS. All rights reserved.

ARTICLE IN PRESS+ModelACVD-764; No. of Pages 8

2 S. Malekzadeh-Milani et al.

insufficiency. The Melody valve was successfully implanted in all patients. Central venous pres-sure and inferior vena cava pressure did not change after valvulation. There were no earlycomplications. At follow-up, no acute or mid-term thrombosis was noted. Two patients hadintracardiac echocardiography 6 and 24 months after valvulation: the Melody valve was foundnot to be functioning in both cases. One patient died 3 months after valvulation; the cause wasunrelated to the procedure.Conclusion. — Percutaneous valvulation of Fontan circulation is technically feasible. More clin-ical studies are needed before considering this treatment as an option.© 2014 Elsevier Masson SAS. All rights reserved.

MOTS CLÉSValve Mélody ;Valvulationpercutanée ;Fontan ;Défaillance duFontan ;Cardiopathiescongénitales

RésuméContexte. — La circulation de Fontan est fragile avec une évolution emaillée de nombreusescomplications. Sa physiologie est unique et dépend beaucoup de la respiration et de la gravité.L’hypothèse de l’effet favorable de la valvulation est que la mise en place d’une valve dans lecircuit réduit les effets de la respiration et de la position proclive. Ce faisant, la valvulationaugmente le débit antérograde de la circulation systémique améliorant la capacité à l’effort etpeut également diminuer les signes d’entéropathie exsudative. La valve Mélody® (Medtronic,Minneapolis, MN, États-Unis) qui est issue de la jugulaire de bœuf est, de par son origine,naturellement faite pour fonctionner dans un environnement à basse pression.Objectif. — Nous rapportons notre expérience de la valulation percutanée de circuits de Fontandéfaillants et réfractaires au traitement médical conventionnel.Méthodes. — Tous les patients ayant recu une valve Mélody dans une circulation de Fontan ontété revus.Résultats. — Quatre patients ont été inclus : deux patients avaient une entéropathie exsuda-tive sévère et réfractaire aux traitements conventionnels ; un patient était en anasarque avecune ascite importante ; et le dernier présentait de très larges varices invalidantes des mem-bers inférieurs. La Mélody a été implantée chez les quatre patients avec succès. Les pressionsveineuses centrales et caves inférieures n’ont pas été modifiées par la valvulation. Il n’y a paseu de complications précoces ni tardives. Durant le suivi, il n’y a pas eu de thrombose aiguë ouchronique démontrée. Deux patients ont eu une échographie intracardiaque respectivement à6 et 24 mois après valvulation : il n’y avait pas de fonction valvulaire observée. Un patient estdécédé 3 mois après la procédure sans relation avec celle-ci.Conclusions. — La valvulation percutanée est aisément réalisable. D’autres études cliniquessont nécéssaires avant de considérer ce traitement comme option thérapeutique en cas dedéfaillance de circulation de Fontan.© 2014 Elsevier Masson SAS. Tous droits réservés.

B

TwpemuaF(ct[toeio

viarw[

twcmtisc

ackground

he Fontan procedure provides good palliation for patientsith single ventricular physiology [1]. However, this non-hysiological circulation is fragile, and many patients willxperience a variety of end-organ complications and uniqueorbidities [2,3]. Complications are varied and can be

nspecific, such as ventricular dysfunction, arrhythmiasnd global cardiac failure, or more specific to the failingontan, such as liver fibrosis, protein-losing enteropathyPLE) and plastic bronchitis. These complications are theonsequences of a complex physiology involving circula-ory insufficiency, inflammation and lymphatic derangement4,5]. Management strategies are emerging, but currentreatments are usually palliative, with modest results, and

Please cite this article in press as: Malekzadeh-Milani S, et al. Peeffects and follow-up. Arch Cardiovasc Dis (2014), http://dx.d

nce severe complications occur, life expectancy is short-ned [6]. Heart transplantation is the only treatment, buts technically challenging and the outcome is poorer than inther diseases requiring heart transplantation [7].

pch

The use of stents in Fontan circuits has been reported inarious situations, with good results; they have been placedn pulmonary arteries to relieve stenosis and in the extrac-rdiac conduit to treat acute thrombosis [8,9]. Recently, weeported the use of a covered stent to close fenestration,ith a good result and no thromboembolic complications

10].Fontan physiology is unique. Cardiac output in this set-

ing is almost exclusively preload dependent and variesith gravity and respiration [11]. Part of the inferior venaava (IVC) blood flows back into the abdominal compart-ent during expiration. The importance of respiration in

he Fontan circulation has been evaluated in in vitro exper-mental models, but also by computational fluid dynamictudies [12,13]. This phenomenon is emphasized in the pro-live position; decrease of cardiac output is noticed in this

rcutaneous valvulation of failing Fontan: Rationale, acuteoi.org/10.1016/j.acvd.2014.07.050

osition and might be responsible for decreased exerciseapacity, liver cirrhosis and PLE [14,15]. We and othersave hypothesized that valvulation of the Fontan pathway

IN+Model

pmtw

pii

R

FFsmf

C

APiAaafcpagKYror

eP

ARTICLEACVD-764; No. of Pages 8

Melody® in failing Fontan

reduces respiratory variations and decreases the effectsof gravity and the proclive position. As a result, valvula-tion theoretically increases preload and cardiac output, andmight improve exercise capacity and decrease PLE. Thishypothesis is also supported by recent mathematical andin vitro studies [16,17].

There is long-term experience of valved Fontan in the ini-tial period, showing poor results, related to early stenoticdegeneration of the surgically implanted valve. Two recentchanges made us resuscitate this option: availability oftranscatheter valved stents; and a potentially good substi-tute — the Melody® valve (Medtronic, Minneapolis, MN, USA).In the early days, valvulation was done using an aortic valve(mostly human or porcine); because of their origins, thesevalves are thick and are designed to work at high pressures.The Melody valve is a bovine jugular vein sutured in a stent;this valve is thin and is designed to work in a low-pressureenvironment.

Because of the lack of therapeutic options, we sought tooffer alternative palliation for patients with failing Fontandespite maximal medical treatment. Percutaneous valvula-tion of the Fontan circuit was offered to these patients,using a Melody valve. To date, this is the first report on per-cutaneous valve implantation in a Fontan circulation. Wereview the results of percutaneous valvulation of Fontancircuit and report on the acute and chronic effects of thisprocedure.

Methods

Patients who received a Melody valve in a failing Fontancirculation were reviewed in our institution. The patientshad various complications related to the Fontan circulationdespite maximal medical treatment. Exclusion criteria wereanatomical abnormalities responsible for Fontan dysfunc-tion and contraindication to anticoagulation or antiplatelettherapy.

All patients presented with severe symptoms of Fontan

Please cite this article in press as: Malekzadeh-Milani S, et al. Peeffects and follow-up. Arch Cardiovasc Dis (2014), http://dx.d

failure; they were all on maximal medical treatment andwere aware of the lack of other therapeutic options. Thecardiologist in charge of the patient explained the technicaldetails of the valvulation and the potential benefits of the

ccpM

Table 1 Demographic characteristics.

Patient 1 Patient 2

Age (years) 29 14

CHD Tricuspidatresia

Complex D

Palliation FenestratedextracardiacFontan

ExtracardiFontan

NYHA III II—III

Complications PLE, liverfibrosis

PLE

CHD: congenital heart disease; DORV: double outlet right ventricle; NYHventricular septum; PLE: protein-losing enteropathy.

PRESS3

rocedure to the patient and their family. The absence ofedical data regarding percutaneous valvulation in this set-

ing was also mentioned to the patient. Informed consentas obtained from each patient.

Because of their heterogeneity and number, patients areresented as individual cases. Data on patient character-stics, surgical history and clinical status before and aftermplantation are provided for each patient.

esults

our patients had a Melody valve implanted in a failingontan; they all presented with a severe medical conditionecondary to a failing Fontan circulation despite maximaledical treatment. All valves were implanted from the

emoral vein. Results are summarized in Tables 1, 2 and 3.

ase 1

29-year-old woman with tricuspid atresia was referred forLE in 2011. The patient had a pulmonary artery band innfancy followed by a Blalock shunt at the age of 2 years.t the age of 5 years, she had a Fontan operation with anortic homograft between the right atrium and pulmonaryrtery. In 2010, she presented atrial fibrillation and cardiacailure, pleural effusion and ascites, and received a Fontanonversion with a 20 mm extracardiac conduit. The patientresented with hypoalbuminaemia, major ascites, elevatedlpha-1 antitrypsin clearance and polyclonal increase inamma globulins; she was on high doses of diuretics, vitamin

antagonists, beta-blockers and digoxin, and was in Nework Heart Association (NYHA) functional class III. Transtho-acic echocardiography (TTE) showed good systolic functionf the systemic ventricle, dilated IVC with contrast and noeflux in the IVC or portal vein.

Cardiac catheterization was performed under gen-ral anaesthesia. The right femoral vein was punctured.ressures were low (14 mmHg in the total cavopulmonary

rcutaneous valvulation of failing Fontan: Rationale, acuteoi.org/10.1016/j.acvd.2014.07.050

onnection [TCPC]; 10 mmHg in the capillaries). No anatomi-al abnormality was seen on angiography. A Melody valve waslaced in the conduit using a 22 mm Ensemble® (Medtronic,inneapolis, MN, USA) delivery catheter. No prestenting was

Patient 3 Patient 4

21 25ORV PA-IVS PA-IVS

ac FenestratedextracardiacFontan

ClassicalFontan

III IIOedema,ascites

Lower limbvenousinsufficiency

A: New York Heart Association; PA-IVS: pulmonary atresia, intact

ARTICLE IN PRESS+ModelACVD-764; No. of Pages 8

4 S. Malekzadeh-Milani et al.

Table 2 Haemodynamic data.

Patient 1 Patient 2 Patient 3 Patient 4

Baseline (mmHg)CVP 14 17 14 15Wedge pressure 10 12 10 10

Post-valvulation (mmHg)CVP 14 17 19 15

CVP: central venous pressure.

Table 3 Follow-up duration and status.

Patient 1 Patient 2 Patient 3 Patient 4

Follow-up (months) 28 3a 28 24Clinical status No PLE Transient improvement Stable Stable

PLE: protein-losing enteropathy.a

dPsr

ocwc

C

AtPwhuacibpTOgdCsmpcNwap

ia

tOtoagdwas stopped abruptly, making acute adrenocortical insuf-ficiency a possible cause of death.

Sudden death after 3 months.

one. On angiography, the Melody valve was competent.ressures were not modified after valvulation. Compres-ion was done manually. No procedural complications wereeported.

After a follow-up of 28 months, the patient is stable with-ut PLE. Diuretics have been decreased and NYHA functionallass has improved slightly. Normalization of albuminaemiaas observed after 26 months of follow-up. On TTE, ventri-ular systolic function is good and IVC is not dilated.

ase 2

14-year-old boy with complex double outlet right ven-ricle treated as a single ventricle presented with severeLE despite maximal treatment (diuretics, low-molecular-eight heparin, steroids, albumin infusions). The patientad an extracardiac cavopulmonary connection and hadndergone several surgical valvuloplasties of the twotrioventricular valves. The patient had developed manyomplications secondary to steroid treatment, includ-ng avascular necrosis of the hip and peritonitis treatedy antibiotics followed by surgical appendicectomy. Theatient was considered for transcatheter valvulation of theCPC; at this time, he was in NYHA functional class III.n TTE, systolic function of the systemic ventricle wasood, with moderate tricuspid valve regurgitation. IVC wasilated, with reflux in the hepatic veins but not in the IVC.ardiac catheterization was done under general anaesthe-ia. The right femoral vein was punctured. Pressures wereildly elevated (17 mmHg in the pathway; mean wedgeressure 12 mmHg). A Melody valve was positioned in theonduit with a 22 mm Ensemble delivery catheter (Fig. 1).

Please cite this article in press as: Malekzadeh-Milani S, et al. Peeffects and follow-up. Arch Cardiovasc Dis (2014), http://dx.d

o prestenting was needed. Post-valvulation pressuresere unchanged. Compression was done manually withoutny problems. The procedure was uncomplicated. Post-rocedure, PLE improved. The boy was weaned off albumin

F2

nfusion; other medications (diuretics, heparin 1 dose/daynd steroids) were progressively reduced.

Three months after valvulation, the patient was admit-ed with status epilepticus and multiple ionic abnormalities.n admission, echocardiography showed normal flow inhe TCPC and normal heart function. The patient diedf ventricular fibrillation secondary to uncorrected ionicbnormalities while waiting for a cerebral computed tomo-raphy scan. An autopsy was refused and aetiology ofeath remains unknown. At interrogation, corticosteroid

rcutaneous valvulation of failing Fontan: Rationale, acuteoi.org/10.1016/j.acvd.2014.07.050

igure 1. Angiogram showing a competent Melody valve (patient).

ARTICLE IN PRESS+ModelACVD-764; No. of Pages 8

Melody® in failing Fontan 5

Figure 2. Intracardiac echocardiography (patient 3). A. Inferior vena cava (IVC) Doppler, spontaneous breathing, before valvulation. B. valvhing:

ervmu

spaaaivAd

C

AvlTto

IVC Doppler, spontaneous breathing: increase of anterograde flow;anterograde flow; no valvular function. D. IVC Doppler during breat

Case 3

A 21-year-old woman with pulmonary atresia and intactventricular septum came to our attention with ascites andoedema despite high doses of diuretics.

The patient had received a neonatal Blalock shuntfollowed by a Glenn procedure and an intracardiac cavopul-monary connection at the age of 4 years. In 2006, she hada Fontan conversion with a 20 mm extracardiac conduit.Six months later, because of anasarca, cardiac catheteriza-tion was done and the right pulmonary artery was stented.Because of elevated venous pressure, the fenestration wasreopened using a bare-metal stent, with good initial results.Oxygen saturation was 85% after the creation of the fen-estration. In 2011, TTE showed moderate impairment ofsystolic function of the left ventricle (ejection fraction 50%)and a mildly dilated IVC with reflux in the hepatic veins. Thepatient was in NYHA functional class III.

Cardiac catheterization was done under general anaes-thesia. The right femoral vein was punctured. Meanpressures were low (14 mmHg in the TCPC; 10 mmHg inthe capillaries). Angiography showed reflux in the IVC. AMelody valve was placed in the extracardiac conduit belowthe fenestration without previous prestenting. Moderate

Please cite this article in press as: Malekzadeh-Milani S, et al. Peeffects and follow-up. Arch Cardiovasc Dis (2014), http://dx.d

elevation of central pressure to 19 mmHg was noted aftervalvulation, but the patient received a high dose of con-trast dye. Oxygen saturation increased from 85% to 90% aftervalvulation. Valvular function was observed on intracardiac

fgvN

ular click. C. IVC Doppler in apnoea, post-valvulation: continuous valvular closure during expiration.

chocardiography (Fig. 2); indeed, a valvular click wasecorded on Doppler measurements. The opening of thealve correlated with respiration. Compression was doneanually without any difficulties. The procedure was

ncomplicated.No clear benefit of valvulation was reported (other than

tabilization of clinical status) during the follow-up. Theatient had a second haemodynamic evaluation 6 monthsfter valvulation (Fig. 3). Pressures were unchanged. Intrac-rdiac echography showed no thrombus in the Melody valvend no gradient across the valve. On two-dimensionalmages, valvular motion of the leaflets was recorded but thealve was significantly regurgitant on colour flow Doppler.fter 28 months of follow-up, the patient is stable withoutegradation of status.

ase 4

25-year-old woman with pulmonary atresia and intactentricular septum was referred because of very severeower limb venous insufficiency with large vein varicosis.he patient had neonatal anastomosis and a right ventricle-o-pulmonary artery opening, and had a classical Fontanperation at the age of 2 years. A pacemaker was implanted

rcutaneous valvulation of failing Fontan: Rationale, acuteoi.org/10.1016/j.acvd.2014.07.050

or sinus dysfunction at the age of 24 years. TTE showedood systolic function of the left ventricle, competent mitralalve and dilated IVC without reflux. The patient was inYHA functional class II.

ARTICLE IN+ModelACVD-764; No. of Pages 8

6

Fat

twittcVcdbosVffvrv

iawrbeLaac

D

Twerero[

ceoh

bsp[ibvitohHdthPcnpfsi

Itm(towtetht

ceflIwomsretol2olw

igure 3. Intracardiac echocardiography (patient 3). Six monthsfter valvulation: no valvular click; retrograde flow during expira-ion.

Cardiac catheterization was done under general anaes-hesia after puncture of the right femoral vein. Pressuresere low (15 mmHg in the pathway; 10 mmHg in the capillar-

es). A covered Cheatham-platinum stent was implanted inhe IVC to create a safe landing zone, followed by implan-ation of a Melody valve mounted in the 22 mm deliveryatheter. Post-valvulation, pressures remained unchanged.alve function was tested with variation of the respiratoryycle in the declive and proclive positions using intracar-iac echocardiography. The proclive position was achievedy moving the table until the patient was lying at an anglef up to 30◦. The patient was not intubated but was deeplyedated and ventilated with non-invasive positive pressure.alve function was highly dependent on respiratory cycle,unctioning at high frequency (≥ 30 cycles/min) and dys-unctioning below this level. Gravity alone did not improvealve function but it improved valve function at high respi-atory frequency. Venous insufficiency did not improve afteralvulation.

Twenty-four months after valvulation, cardiac catheter-zation was done for haemodynamic evaluation and valvularssessment using intracardiac echocardiography. Pressuresere low (12 mmHg). No gradient across the valve was

ecorded. Intravascular echocardiography showed no throm-us formation. There was no evidence of valvular function,ven in the proclive position and with a high respiratory rate.eaflets could not be seen. The Melody valve appeared like

non-valved conduit. Retrograde flow was seen in the IVCnd no valvular click was recorded. No calcification of theonduit was seen on fluoroscopy and cine angiography.

iscussion

he Fontan operation provides good palliation for patientsith univentricular hearts [1]. Surgical techniques havevolved over the past few decades, leading to betteresults and the long-term survival of these patients. Nev-

Please cite this article in press as: Malekzadeh-Milani S, et al. Peeffects and follow-up. Arch Cardiovasc Dis (2014), http://dx.d

rtheless, many of these patients will develop a wideange of complications [2—4], most of which are the resultf complex physiology and are incompletely understood4,5]. Some therapies have been developed to treat these

e

md

PRESSS. Malekzadeh-Milani et al.

atastrophic complications, but efficacy is low and lifexpectancy is reduced [6]. Once the Fontan has failed, thenly therapeutic option is heart transplantation, but with aigh mortality rate [7].

The physiology of the Fontan circulation is unique. It haseen termed paradoxical, in that systemic venous hyperten-ion (mean pressure > 10 mmHg) occurs simultaneously withulmonary arterial hypotension (mean pressure < 15 mmHg)18]. The most important determinant of cardiac outputs preload [11], which is determined by the amount oflood returning in the heart; it is affected by systemicenous return and pulmonary vascular resistance. Focus-ng on systemic venous return, some studies have reportedhat infradiaphragmatic venous flow is markedly dependentn respiration and gravity in Fontan patients, with gravityaving an adverse effect on portal and IVC flow [12,13].aemodynamic studies on patients with failing Fontan haveemonstrated that one of the main components of the symp-oms of chronic Fontan failure is liver fibrosis and portalypertension secondary to elevation of venous pressure.art of the circulatory dysfunction in failing Fontan mimicsirculatory dysfunction in portal hypertension, with splanch-ic arterial vasodilatation because of reduced hepaticerfusion, followed by arterial hypotension, renal hypoper-usion and activation of the renin-angiotensin-aldosteroneystem, resulting in fluid retention. Preservation of the livers thus a key issue in the treatment of Fontan [5].

Valved conduits have been used in Fontan surgery [19].ndeed, homografts were used to reduce backward flow inhe true Fontan operation, which was later abandoned byost centres. The bovine jugular vein is used surgically

Contegra®; Medtronic, Minneapolis, MN, USA) and percu-aneously (Melody) for the treatment of right ventricularutflow tract reconstruction and in the tricuspid position,ith good results [20—22]. The valve is thin, coming from

he venous system, and is designed to work in a low-pressurenvironment, making it a good candidate for valvulation ofhe Fontan circulation. Experimental studies on this valveave demonstrated that little or no reflux is needed to closehe valve [23].

Our hypothesis was that valvulation of the failing Fontanircuit with the Melody valve would reduce the adverseffects of gravity and respiration by reducing retrogradeow and thus increasing anterograde net flow and preload.

n the four patients studied, valvulation was straightforwardithout complications. Acutely, no pressure changed wasbserved. Despite valve functioning, no clinical improve-ent was noticed early after valvulation. PLE improved or

tabilized in two patients, but it is not known if this waselated to the procedure. The last two patients had cath-terization 6 months and 24 months after implantation ofhe Melody valve. Deterioration of the valve was observedn intracardiac echocardiography: at 6 months, valvulareaflets were observed, but with regurgitation; and at

years from implantation, no valvular function was noticedn intracardiac echocardiography. The reasons for valvu-ar disappearance are unknown. We found similar behaviourhen we implanted the Melody valve in the IVC in a chronic

rcutaneous valvulation of failing Fontan: Rationale, acuteoi.org/10.1016/j.acvd.2014.07.050

xperimental model [24].This study was conducted to find an alternative treat-

ent for chronic venous insufficiency, a frequent vascularisease. The results of the study were disappointing, unlike

IN+Model

D

Ydt

R

[

[

[

[

[

[

[

ARTICLEACVD-764; No. of Pages 8

Melody® in failing Fontan

other acute studies, in that at autopsy, the valve leafletswere completely embedded in the vascular wall, leavingan opened non-valved conduit [25]. We hypothesized thata valve needs to be constantly functioning to keep working.Similar results have been reported in the clinical settingwhen the Melody valve was inserted in the tricuspid posi-tion in patients with unfavourable haemodynamic conditions[26]. Interestingly, intracardiac echocardiography providedus with very interesting information. The role of respira-tion should be noted. We showed that the closure of thevalve was only present when the respiratory rate was ≥ 30cycles/min, which is higher than the normal respiratory rateat rest in children and adults. Gravity alone did not affectvalve function. This, of course, was tested with patientsdeeply sedated, using non-invasive ventilation. We agreethat this is not a physiological situation, but we do thinkthat it mimics closely what happens in the normal breathingsituation. Moreover, it allowed us to test various variables,such as respiratory rate, inspiration/expiration time ratioand level of positive expiratory pressure.

Study limitations

This series is very limited. Although no spectacular clinicalimprovement was demonstrated, three patients were sta-ble, stabilized or had improvement in PLE for almost 2 years.This apparent stability could be considered a positive end-point because of the natural evolution of failing Fontan. Thedifferent nature of the complications and the heterogene-ity of the patients make definitive conclusions very difficultto draw. The exact effect of the valve remains unknown.Additional evaluation (i.e. magnetic resonance imaging andexercise testing) would have been interesting, to evaluatemore accurately the effect of valvulation. Patients refusedto have additional testing. The risk of the procedure is verylow in patients with an extracardiac TCPC made of a Goretexconduit. Besides questioning the effectiveness of the proce-dure, long-term complications may occur, making cautiousfollow-up of value. In particular, endocarditis and thrombo-sis are of concern, as with all cardiac valves.

Conclusions

Once Fontan fails, few treatments are useful, apart fromheart transplantation, with its own long-term problem. Inlight of this, we offered percutaneous valvulation to patientswith failing Fontan. Despite modifications of the echocar-diography — particularly modulation of hepatic flow — nospectacular clinical benefit was observed. Valve functionwas not obvious in implanted patients in the long term,but assessment remains difficult, requiring invasive assess-ment by intracardiac echocardiography in our cases. Patientselection remains very important. At this stage, this strat-egy is not recommended in all cases of failing Fontan; itshould probably be limited to patients with demonstrated

Please cite this article in press as: Malekzadeh-Milani S, et al. Peeffects and follow-up. Arch Cardiovasc Dis (2014), http://dx.d

back flow in the abdominal venous system. Systemic reviewof detailed echocardiographical findings in supine and pro-clive positions should be obtained in this population beforeconsidering such therapy.

[

PRESS7

isclosure of interest

. Boudjemline is a proctor for Medtronic. The other authorseclare that they have no conflicts of interest concerninghis article.

eferences

[1] Khairy P, Poirier N, Mercier LA. Univentricular heart. Circula-tion 2007;115:800—12.

[2] Gewillig M. The Fontan circulation. Heart 2005;91:839—46.[3] Mondesert B, Marcotte F, Mongeon FP, et al. Fontan circulation:

success or failure? Can J Cardiol 2013;29:811—20.[4] Goldberg DJ, Shaddy RE, Ravishankar C, Rychik J. The fail-

ing Fontan: etiology, diagnosis and management. Expert RevCardiovasc Ther 2011;9:785—93.

[5] Hebson CL, McCabe NM, Elder RW, et al. Hemodynamic pheno-type of the failing Fontan in an adult population. Am J Cardiol2013;112:1943—7.

[6] Schumacher KR, Cools M, Goldstein BH, et al. Oral budesonidetreatment for protein-losing enteropathy in Fontan-palliatedpatients. Pediatr Cardiol 2011;32:966—71.

[7] Backer CL, Russell HM, Pahl E, et al. Heart transplantation forthe failing Fontan. Ann Thorac Surg 2013;96:1413—9.

[8] Mets JM, Bergersen L, Mayer Jr JE, Marshall AC, McElhinneyDB. Outcomes of stent implantation for obstruction of intrac-ardiac lateral tunnel Fontan pathways. Circ Cardiovasc Interv2013;6:92—100.

[9] Tanase D, Ewert P, Eicken A. Plastic bronchitis: symp-tomatic improvement after pulmonary arterial stenting in fourpatients with Fontan circulation. Cardiol Young 2013:1—3,http://dx.doi.org/10.1017/S1047951113002163.

10] Malekzadeh-Milani S, Ladouceur M, Bajolle F, BonnetD, Boudjemline Y. Closure of Fontan fenestration withthe use of covered stents: short- and mid-term resultsin a cohort of 50 patients. Cardiol Young 2014:1—6,http://dx.doi.org/10.1017/S1047951114000894.

11] Gewillig M, Brown SC, Eyskens B, et al. The Fontan circulation:who controls cardiac output? Interact Cardiovasc Thorac Surg2010;10:428—33.

12] Liu J, Qian Y, Sun Q, Umezu M. Use of computational fluiddynamics to estimate hemodynamic effects of respiration onhypoplastic left heart syndrome surgery: total cavopulmonaryconnection treatments. Sci World J 2013;2013:131597.

13] Vukicevic M, Chiulli JA, Conover T, Pennati G, Hsia TY, Figli-ola RS. Mock circulatory system of the Fontan circulation tostudy respiration effects on venous flow behavior. ASAIO J2013;59:253—60.

14] Ghaferi AA, Hutchins GM. Progression of liver pathol-ogy in patients undergoing the Fontan procedure: chronicpassive congestion, cardiac cirrhosis, hepatic adenoma,and hepatocellular carcinoma. J Thorac Cardiovasc Surg2005;129:1348—52.

15] Hsia TY, Khambadkone S, Redington AN, Migliavacca F, Dean-field JE, de Leval MR. Effects of respiration and gravity oninfradiaphragmatic venous flow in normal and Fontan patients.Circulation 2000;102:III148—53.

16] Santhanakrishnan A, Maher KO, Tang E, Khiabani RH, John-son J, Yoganathan AP. Hemodynamic effects of implanting aunidirectional valve in the inferior vena cava of the Fontan cir-culation pathway: an in vitro investigation. Am J Physiol Heart

rcutaneous valvulation of failing Fontan: Rationale, acuteoi.org/10.1016/j.acvd.2014.07.050

Circ Physiol 2013;305:H1538—47.17] Vukicevic M, Conover T, Jaeggli M, et al. Control of respiration-

driven retrograde flow in the subdiaphragmatic venous returnof the Fontan circulation. ASAIO J 2014.

IN+ModelA

8

[

[

[

[

[

[

[

[

[26] Bentham J, Qureshi S, Eicken A, Gibbs J, Ballard G, Thomson

ARTICLECVD-764; No. of Pages 8

18] de Leval MR. The Fontan circulation: a challenge to WilliamHarvey? Nat Clin Pract Cardiovasc Med 2005;2:202—8.

19] Fontan F, Baudet E. Surgical repair of tricuspid atresia. Thorax1971;26:240—8.

20] Bonhoeffer P, Boudjemline Y, Saliba Z, et al. Percutaneousreplacement of pulmonary valve in a right-ventricle topulmonary-artery prosthetic conduit with valve dysfunction.Lancet 2000;356:1403—5.

21] Breymann T, Thies WR, Boethig D, Goerg R, Blanz U, KoerferR. Bovine valved venous xenografts for RVOT reconstruc-tion: results after 71 implantations. Eur J Cardiothorac Surg

Please cite this article in press as: Malekzadeh-Milani S, et al. Peeffects and follow-up. Arch Cardiovasc Dis (2014), http://dx.d

2002;21:703—10 [discussion 10].22] Roberts PA, Boudjemline Y, Cheatham JP, et al. Percutaneous

tricuspid valve replacement in congenital and acquired heartdisease. J Am Coll Cardiol 2011;58:117—22.

PRESSS. Malekzadeh-Milani et al.

23] Qui Y, Quijano RC, Wang SK, Hwang NH. Fluid dynamics ofvenous valve closure. Ann Biomed Eng 1995;23:750—9.

24] Boudjemline Y, Bonnet D, Sidi D, Bonhoeffer P. Is percutaneousimplantation of a bovine venous valve in the inferior vena cavaa reliable technique to treat chronic venous insufficiency syn-drome? Med Sci Monit 2004;10:BR61—6.

25] Gomez-Jorge J, Venbrux AC, Magee C. Percutaneous deploy-ment of a valved bovine jugular vein in the swine venoussystem: a potential treatment for venous insufficiency. J VascInterv Radiol 2000;11:931—6.

rcutaneous valvulation of failing Fontan: Rationale, acuteoi.org/10.1016/j.acvd.2014.07.050

J. Early percutaneous valve failure within bioprosthetic tri-cuspid tissue valve replacements. Catheter Cardiovasc Interv2013;82:428—35.