Right Ventricular Outflow Tract ReconstructionWith Bicuspid Valved PolytetrafluoroethyleneConduitMasahiro Yoshida, MD, Peter D. Wearden, MD, PhD, Onur Dur, MS,Kerem Pekkan, PhD, and Victor O. Morell, MD

Department of Cardiothoracic Surgery, Children’s Hospital of Pittsburgh, and Department of Biomedical Engineering, CarnegieMellon University, Pittsburgh, Pennsylvania

PED

IAT

RIC

CA

RD

IAC

Background. In general, all conduits available for rightventricular outflow tract (RVOT) reconstruction eventu-ally become stenotic or insufficient. Owing to the lack ofan ideal conduit and with the hope of reducing theincidence of reoperations, we have developed and uti-lized a bicuspid valved polytetrafluoroethylene (PTFE)conduit for the reconstruction of the RVOT. The purposeof this study was to review our early experience with thisconduit.

Methods. From October 2008 to September 2009, wehave implanted bicuspid valved PTFE conduits in 18patients with a median age of 1.7 years (range 6 days to 16years). Their diagnoses include tetralogy of Fallot withpulmonary atresia in 8, truncus arteriosus in 6, congenitalaortic stenosis in 2, transposition of great arteries in 1,and interrupted aortic arch with a ventricular septaldefect in 1. In 16 patients, a complete biventricular repairwas performed. In another 2 cases, the conduit was used

for palliative RVOT reconstruction. The conduit sizes

Pavilion, 5th Flr, Pittsburgh, PA 15224; e-mail: masahiro.yoshida@chp.edu.

© 2011 by The Society of Thoracic SurgeonsPublished by Elsevier Inc

varied from 10 mm to 24 mm in diameter. Three-dimen-sional flow fields obtained from computational fluiddynamics studies were utilized in the conduit designprocess.

Results. There was no surgical mortality or reinterven-tions associated with the PTFE conduit placement in ourseries. At the time of discharge, none of the patients hadany echocardiographic findings consistent with signifi-cant conduit stenosis or insufficiency. During the fol-low-up period of 6.2 � 3.9 months, all patients were aliveand only 3 had more than mild pulmonary insufficiency.

Conclusions. Our bicuspid valved PTFE conduit has anacceptable early performance, with a low incidence ofvalve insufficiency and no conduit stenosis. Certainly,longer follow-up is necessary to fully assess its long-termbenefits.

(Ann Thorac Surg 2011;91:1235–9)

© 2011 by The Society of Thoracic Surgeons

Based on general experience, all conduits available forright ventricular outflow tract (RVOT) reconstruc-

tion eventually become stenotic and/or insufficient, espe-cially in very young patients [1–3]. Owing to the lack of anideal conduit and with the hope of reducing the inci-dence of reoperations, we have developed and utilized abicuspid valved polytetrafluoroethylene (PTFE) conduitusing standard stretch PTFE graft and 0.1 mm thick PTFEmembrane (W.L. Gore & Associates, Flagstaff, AZ) for thereconstruction of the RVOT. The purpose of this studywas to review our early experience with this conduit.

Patients and Methods

From October 2008 to September 2009, 18 patientsunderwent implantation of a bicuspid valved PTFE

Accepted for publication Nov 8, 2010.

Presented at the Fifty-sixth Annual Meeting of the Southern ThoracicSurgical Association, Marco Island, FL, Nov 4–7, 2009.

Address correspondence to Dr Yoshida, Department of CardiothoracicSurgery, Children’s Hospital of Pittsburgh, 4401 Penn Ave, Faculty

conduit in the RVOT at Children’s Hospital of Pitts-burgh. The median age at the time of implantation was1.7 years (range, 6 days to 16 years). Their diagnosesincluded tetralogy of Fallot with pulmonary atresia in8, truncus arteriosus in 6, congenital aortic stenosis in2, transposition of great arteries in 1, and interruptedaortic arch with a ventricular septal defect in 1 (Table1). Thirteen patients had a previous procedure thatincluded a palliative RVOT reconstruction in 3, repairfor truncus arteriosus in 3, repair for tetralogy of Fallotin 3, bilateral unifocalization in 1, Ross procedure in 1,arch reconstruction in 1, and arterial switch in 1.Concomitant procedure at the time of conduit implan-tation included VSD closure in 4, repair for truncusarteriosus in 3, Ross procedure in 1, reconstruction ofthe central pulmonary arteries in 1, aortic valve re-placement in 1, and one-stage repair of tetralogy ofFallot with pulmonary atresia with unifocalization in 1.The diameter of the conduit ranged from 10 to 24 mm.The size of the conduit was determined according tothe patient’s body weight and body surface area (Fig 1).This research study was approved by our Institutional

Review Board.

0003-4975/$36.00doi:10.1016/j.athoracsur.2010.11.010

cc

ns(oouchvo

mmacclaaogw

1236 YOSHIDA ET AL Ann Thorac SurgBICUSPID VALVED PTFE CONDUIT 2011;91:1235–9

PEDIA

TR

ICC

AR

DIA

C

Operative TechniqueThe valved conduit was constructed in the operatingtheater at the time of surgery. First, the PTFE conduit isturned inside out, and then the two cusps created from0.1 mm PTFE membrane are sutured to the inside wall ofthe conduit using a running 7-0 polypropylene runningsuture for conduits 14 mm or less in diameter and 6-0polypropylene for larger ones. Finally, the conduit isturned outside in (Fig 2). The shape and dimensions ofthe valve leaflets for each specific conduit diameter areshown in Figure 3. Note that the conduit has a smallnonvalved portion at the bottom, representing 15% ofcircumference of conduit. This allows for a minimalamount of regurgitation, which should prevent thrombusformation at the base of the valve sinuses. It has been ourpractice to place the “valved” portion of the conduitdistally, leaving just a few millimeters of conduit wallbeyond the top margin of PTFE cusps. The proximal endis then trimmed appropriately to match the opening inthe RVOT.

Postoperative Anticoagulation TherapyLow-dose aspirin (1 to 5 mg/kg daily) was started in thehospital and continued for 6 months.

Echocardiographic EvaluationAll patients had an intraoperative transesophageal echo-cardiogram before discharge and during follow-up.These studies were systematically reviewed to assess thepresence of pulmonary regurgitation and to determinethe mean valve gradient. The pulmonary regurgitationwas classified as trivial (grade 1), mild (grade 2), moder-ate (grade 3), and severe (grade 4) according to features ofthe jet, assessed with pulsed flow Doppler and colorDoppler. Conduit stenosis was assessed by measuringthe peak velocity through the valve with continuous-wave Doppler technique.

Statistical AnalysisPreoperative and postoperative data were retrospectivelycollected. Descriptive data for continuous variables arepresented as means � SD or as medians with ranges;ategorical variables are presented as relative frequen-

Table 1. Patient Profile

Demographic Value

Age, years (range) 1.7 � 1.5 (6 days to16 years)

Body weight, kg (range) 20 � 19 (3.4 to 67)Diagnosis

Tetralogy of Fallot 8Truncus arteriosus 6Congenital aortic stenosis (Ross) 2TGA 1IAA with VSD 1

IAA � interrupted aortic arch; TGA � transposition of great arteries;VSD � ventricular septal defect.

ies. Surgical mortality was defined as death within 30

days of operation. Analysis was performed with JMP 8.01for Macintosh (JMP, Cary, NC).

Results

There was no surgical and late mortality nor eventsassociated with the conduit placement at the mean fol-low-up of 6.2 � 3.9 months. All patients were doing wellwithout any need for conduit reinterventions.

Echocardiographic evaluation of the conduit was doneintraoperatively (T1; n � 18), at the time of discharge (T2;

� 18), and during follow-up (n � 11). Echocardiographyhowed the presence of trivial or mild insufficiency in 1794%) at T1, 17 (94%) at T2, and 8 (73%) at T3. There wasnly 1 patient who had moderate insufficiency in theperating theater and at discharge. It was a neonate whonderwent repair of truncus arteriosus with a 12-mmonduit. We believe that the insufficiency was related tois relatively high heart rate, which prevented adequatealve function. On his latest echocardiogram, there isnly mild insufficiency at a lower heart rate.During follow-up examination in 11 patients, there wasoderate insufficiency in 3. Two of them developedoderate insufficiency after undergoing catheter balloon

ngioplasty for peripheral pulmonary stenosis, whichould have damaged the PTFE valve. In the remainingase, the echocardiogram showed that one of the valveeaflets was stuck in the open position. Pressure gradientcross the conduit was 9 � 8 mm Hg at T1, 13 � 8 mm Hgt T2, and 19 � 15 mm Hg at T3 (Fig 4). During follow-up,nly 1 patient, a neonatal repair, had a pressure gradientreater than 30 mm Hg 12 months after surgery, whiche believe is related to rapid somatic growth.

Comment

Right ventricular outflow tract reconstruction has beenperformed utilizing different types of conduits, mostcommonly aortic or pulmonary homografts, which havebeen associated with a high incidence of stenosis orinsufficiency requiring reoperations especially in very

Fig 1. The authors use various sizes of conduit from 10 to 24 mm indiameter. This graft shows the correlation between conduit size andbody surface area (BSA). The curved line shows normal pulmonary

valve size by Rawlatt.

PTr

sp“c

totsmr

flflcPRtogpe

1237Ann Thorac Surg YOSHIDA ET AL2011;91:1235–9 BICUSPID VALVED PTFE CONDUIT

PED

IAT

RIC

CA

RD

IAC

young patients [1–3]. Also, exposure to homograft mate-rial has been proven to result in significant allosensitiza-tion, which could negatively affects the results of futurecardiac transplantation.

Over the last couple of years, we have opted to prefer-entially utilize a PTFE bicuspid valved conduit for RVOTreconstruction with the expectation that they will have abetter overall performance than other conduits. Brown andcolleagues [4] reported excellent long-term results using a

TFE monocuspid valve in the RVOT. Also, Ando andakahashi [5] described good long-term results of RVOTeconstruction using PTFE trileaflet valved conduits.

Our conduit was designed with a bicuspid valve in-tead of a trileaflet valve because of one of the author’srior experience with a trileaflet conduit in which theposterior” leaflet did not open (Fig 5) during systole in 2onsecutive patients. It was theorized that the flow along

he lesser curvature of the conduit was not sufficient topen the posterior leaflet, and that is why we modifiedhe valve component of the conduit to be bileaflet with amall posterior “nonvalve” segment. The nonvalve seg-ent only represents 15% of the valve circumference,

esulting in very little regurgitation.To further investigate the general three-dimensional

ow patterns inside curved conduits and to quantifyuid-induced forces important for valve kinematics,omputational fluid dynamics studies were performed.ulsatile blood flow was simulated inside the curvedVOT conduit using the second-order accurate compu-

ational fluid dynamics solver (Fluent 6.3.26; ANSYS, Can-nsburg, PA), which was originally developed for investi-ating the reconstructive surgeries for single ventriclealliation [6, 7]. Two conduits of 14-mm and 22-mm diam-ter were evaluated based on the body surface area, 0.5 m2

Fig 2. Drawings of the steps tomake a bicuspid valved conduit.At first, a polytetrafluoroethylene(PTFE) conduit with standardstretch wall is turned inside out.After holding the graft at both endsusing Kelly’s clamps, two cuspstrimmed from the PTFE sheet of0.1-mm thickness are sutured using6-0 or 7-0 polypropylene sutures.Actually, the author put the top ofthe triangle first and put markingstitches at the middle of both sides.Then, continuous suture is put us-ing the stitch of the top. After com-pletion of suturing two cusps, theconduit is turned back outside in(upper drawings). A special featureof this conduit is having a non-valved portion at the bottom (lowerdrawings).

Fig 3. The design of each valve. Thefigures are as follow: circumference(C) � conduit size � 3.14; nonvalvedportion (nV) � C � 0.15; width ofsinus (WS) � (C � nV ) / 2; heightof sinus (HS) � WS � 0.7; width ofcusp (WC) � WS � 1.2; heightof cusp (HC) � HS � 0.9; and fanof cusp (FC) � HC � 0.2.

LNdatmwiicctcopop

vs

tpsdcth

ectpis

1238 YOSHIDA ET AL Ann Thorac SurgBICUSPID VALVED PTFE CONDUIT 2011;91:1235–9

PEDIA

TR

ICC

AR

DIA

C

and 1.35 m2, and the corresponding cardiac outputs, 1.2/min and 2.4 L/min, respectively. Blood was chosen asewtonian fluid with a viscosity of 3.7 l e-3 N-s/m2 and aensity of 1,060 kg/m3. Physiologic RVOT flow waveformnd blunt flow profile (typical for ventricle) was assigned athe inlet of the conduit in agreement with the clinical

easurements. Computational domain of each conduitere discretized using 50,000 tetrahedral elements. Accord-

ng to our simulations, about one diameter distal from thenlet, blood flow skewed toward the major curvature of theonduit. Hence, the flow velocity at the lesser curvature ofonduit was slower than that at the major curvature. Fur-hermore, flow profile was almost symmetrical along theurvature axis, indicating balanced opening forces exertedn each leaflet during the systole (Fig 6). These resultserfectly agree with the author’s clinical experience in theperating room. Similar bioengineering studies toward im-roved designs are ongoing.Yamagishi and associates [8] reported a unique PTFE

alved conduit design in 2007. It incorporated bulginginuses that can generate diastolic vortex flow between

Fig 4. Pressure gradients of conduit in the operating room (OR), atdischarge, and at clinic by echocardiography.

Fig 5. Echocardiography the author experienced in Kobe, Japan. Thestuck valve at the bottom of tricuspid valved conduit is shown (ar-rowhead). It was considered that flow velocity at the lesser curvature

of the curved conduit was slower than at the major curvature.

he conduit wall and the leaflets, and hence potentiallyrovide better long-term results than a conduit withoutinuses. Future modification to our bicuspid valved con-uit may include the creation of sinuses. Our ongoingomputational fluid dynamics investigation will identifyhe benefit of sinus geometry on the RVOT conduitemodynamics and guide its clinical implementation.Our bicuspid valved PTFE conduit has an acceptable

arly performance, with a low incidence of valve insuffi-iency and no conduit stenosis. The lack of allosensitiza-ion is another benefit of this conduit, as well as possiblyroviding a reliable target to deliver percutaneously

mplantable valves. Certainly, longer follow-up is neces-ary to fully assess its long-term benefits.

References

1. Kaza AK, Lim HG, Dibardino DJ, Del Nido PJ, Mayer JE,Pigula FA. Long-term results of right ventricular outflow tractreconstruction in neonatal cardiac surgery: options and out-comes. J Thorac Cardiovasc Surg 2009;138:911–6.

2. Schreiber C, Sassen S, Kostolny M, et al. Early graft failure ofsmall-sized porcine-valved conduits in reconstruction of theright ventricular outflow tract. Ann Thorac Surg 2006;82:179–86.

3. Shebani SO, McGuirk S, Brawn WJ, et al. Right ventricularoutflow tract reconstruction using Contegra® valved conduit:natural history and conduit performance under pressure. EurJ Cardiothorac Surg 2006;29:397–405.

4. Brown JW, Ruzmetov M, Vijay P, Rodefeld MD, TurrentineMW. Right ventricular outflow tract reconstruction with apolytetrafluoroethylene monocusp valve: a twelve-year expe-rience. J Thorac Cardiovasc Surg 2007;133:1336–43.

5. Ando M, Takahashi Y. Ten-year experience with handmadetrileaflet polytetrafluoroethylene valved conduit used for pul-monary reconstruction. J Thorac Cardiovasc Surg 2009;137:124–31.

6. Pekkan K, Kitajima HD, Yoganathan AP, et al. Total cavopul-monary connection flow with functional left pulmonary arterystenosis: angioplasty and fenestration in vitro. Circulation2005;112:3264–71.

7. Wang C, Pekkan K, Yoganathan AP, et al. Progress in the CFDmodeling of flow instabilities in anatomical total cavopulmo-nary connections. Ann Biomed Eng 2007;35:1840–56.

8. Miyazaki T, Yamagishi M, Kado H, et al. Expanded polytet-rafluoroethylene valved conduit and patch with bulging si-nuses in right ventricular outflow tract reconstruction. J Tho-

Fig 6. Newtonian pulsatile blood flow simulated inside the 14-mmconduit for a cardiac output of 1.2 LPM using second-order compu-tational fluid dynamics solver (Fluent 6.3.26). Time-averaged veloc-ity contours shown during systole indicated lower velocity at thelesser curvature of the conduit.

rac Cardiovasc Surg 2007;134:327–32.

D

1239Ann Thorac Surg YOSHIDA ET AL2011;91:1235–9 BICUSPID VALVED PTFE CONDUIT

DISCUSSION

PED

IAT

RIC

CA

RD

IAC

DR JOHN W. BROWN (Indianapolis, IN): I would like tocompliment Dr Yoshida and the group at Pittsburgh for thisingenious RV-PA conduit construction. I have seen several otherconduit constructions, and they were difficult for me to figureout how they proportioned the leaflet size. The Yoshida conduitshould be inexpensive to construct and could more easily beafforded in the developing world. I like your idea of verticalcommissures. This reminds me a little bit of the folding mono-cusp that Graham Nunn reported at the AATS 2 or 3 years ago.And I have noticed with our experimental bovine bicuspidvenous valved conduits that when you orient the leaflets in avertical fashion, you see the absence of the lesser curve stasisthat you described.

My first question is how you came up with the width andheight of the leaflets and the width of the gap at the bottom?How did you come up with the leaflet size and dimensions?

DR YOSHIDA: How?

R BROWN: You have a chart in the manuscript, but how didyou come up with the dimensions?

DR YOSHIDA: Basically I considered the bottom side of theconduit is slower than the major curvature. So even if we put thecommissure vertically, the bottom one must be slower, thevelocity slower. That might cause thrombus and also somestagnant flow. So that is why I designed this conduit, which hasthe nonvalved portion at the bottom. It causes some regurgita-tion but I think for right ventricular outflow, some regurgitationis acceptable.

DR BROWN: I totally agree, but how did you come up with thewidth of the leaflets and the height of the leaflets?

DR YOSHIDA: The basic design of the leaflet was developedthrough trial and error over the years at Kobe Children’sHospital in Japan.

DR BROWN: Of the 0.1 PTFE leaflets inside the conduit, howdid you come up with the dimensions, because they are inter-esting. I just want to understand. Did you pull this out of the airor did someone teach you? How did you come up with thedimensions of the leaflets themselves? I would like to try this,but I don’t understand how long and wide to construct theleaflets.

DR YOSHIDA: Actually, in Japan from my experience of trialand error, we determined this figure, and we have some fearhow wide the gap should be. I can show you also in my slide.Actually, I don’t remember, maybe 12% of the circumference. Ithink 12%.

DR BROWN: I don’t understand whether there was an easy way

to determine leaflet dimensions when you turn the conduit

inside out. You obviously put the leaflets on the outside of theconduit and then turn it right side out. There must have beensome trial and error that came into the construction.

My last questions are, why was there progressive regurgita-tion in the latest follow-up in 25% of the patients? What was thecause of the worsening regurgitation at last follow-up? Are theleaflets not opening all the way or closing all the way? Why doyou think the regurgitation increases with time?

DR YOSHIDA: Thank you for that question. There were 3 casesthat had moderate insufficiency. Two of them developed after acatheter intervention for peripheral pulmonary stenosis. Maybecatheter intervention touched one of the leaflets, causing insuf-ficiency. But the other case, that case had no history of catheterintervention. The conduit has no bulging sinus, it’s just a straightgraft. And this may be a disadvantage, since a cusp can touch thewall of the conduit and maybe become immobile. We routinelygive aspirin for anticoagulation to avoid stuck leaflet. And we arethinking about putting bulging sinus to avoid leaflets touching aconduit wall.

DR BROWN: I think a sinus would facilitate leaflet closure. Thenext question I have is, what is the smallest conduit that you cando this?

DR YOSHIDA: Ten millimeters.

DR BROWN: In your manuscript it was 10 mm, but wouldn’t itbe nice to have a valved Sano conduit? Can you make a conduitsmaller than 10 mm?

DR MORELL: My comment is that Dr Yoshida developed thisconduit using trial and error. We are working with Carnegie-Mellon trying to get some engineers working on this valvedconduit to see if we can make it work more optimally to preventlate insufficiency.

DR BROWN: Well, I really like the idea, and I actually think youcould probably take a smaller one, split the conduit, put yourtwo leaflets in, and sew it back together, and maybe you couldhave a valved Sano conduit for hypoplastic left heart syndrome.

DR YOSHIDA: Actually, once we tried to make a 6-mm conduitwith a monocusp, but that was also cusp touching the conduitwall in the early phase. So I would not recommend so far.

DR BROWN: I very much enjoyed your presentation, and this isan innovative technique. Obviously, your follow-up is onlymonths long, and we need longer follow-up. I look forward toseeing your follow-up at future meetings. Thank you.

DR YOSHIDA: Thank you, Dr Brown.