tci

ohbctn

1

AT

RIC

CA

RD

IAC

Transcaval Correction of Partial AnomalousPulmonary Venous Drainage Into the SuperiorVena CavaMohamed Nassar, MD, Virginie Fouilloux, MD, Loïc Macé, MD,Bernard Kreitmann, MD, PhD, and Dominique Metras, MD

Service of Cardiothoracic Surgery, Children’s Hospital La Timone, Marseille, France

PED

I

Background. The ideal technique for addressing partialanomalous pulmonary venous drainage into the superiorvena cava (SVC), with or without sinus venosus atrialseptal defect (ASD), is debated. The risk of sinus nodedysfunction, systemic, or pulmonary venous channelsobstruction has led to different techniques being devel-oped. We present our experience with 45 patients oper-ated on using a vertical transcaval approach, withoutatrial or cavoatrial junction incision.

Methods. Between 2001 and 2010, 45 patients (28 fe-males, 17 males, with a mean age of 5 years (range, 8months to 70 years), underwent operations using 1 patchof autologous pericardium, after vertical SVC incisionanterior to the anomalous pulmonary veins: 43 hadassociated sinus venosus ASD, and 6 had associated leftSVC. Access was through sternotomy in 19 and rightposterior thoracotomy in 26. An additional right atrialincision, without crossing the cavoatrial junction, was

used in 2 patients without ASD. Mean cardiopulmonary

2

3

WasEa

Children’s Hospital La Timone, Boulevard Jean Moulin, 13385 Mar-seille, France; e-mail: dmetras@ap-hm.fr.

© 2012 by The Society of Thoracic SurgeonsPublished by Elsevier Inc

bypass time was 76 minutes. Mean cross-clamp time was44 minutes.

Results. No deaths or important morbidities occurred.Mean follow-up was 4.4 years (range, 2 months to 9.3years). All patients had regular echocardiographic exam-ination, electrocardiogram, and midterm 24-hour Holterelectrocardiogram. No new arrhythmias occurred. Allpatients showed unobstructed caval and pulmonary ve-nous flow.

Conclusions. The vertical transcaval approach is a sim-ple, highly reproducible technique for correction of par-tial anomalous pulmonary venous drainage into the SVC.It yields excellent results, with unobstructed pulmonaryand systemic venous flow and without arrhythmia devel-opment. It can also be performed through a cosmeticright posterior thoracotomy.

(Ann Thorac Surg 2012;93:193–6)

© 2012 by The Society of Thoracic Surgeons

The incidence of partial anomalous pulmonary venous(PAPV) drainage has been reported to be about 0.7%

in some autopsy series [1], and the anomalous drainageinto the right side of the heart is usually associated withsinus venosus atrial septal defect (ASD), which is presentin 87% to 97% of reported cases [1,2]. The PAPV connec-ion to the right side of the heart has always been aontroversial topic, not only surgically but also concern-ng embryology and pathology [3, 4].

Several techniques to complete the surgical correctionf the PAPV connection to the superior vena cava (SVC)ave been developed. Since the early surgical correctiony Neptune and colleagues [5] and then Kirklin andolleagues [6] in 1956, several modifications and differentechniques have been proposed. The three basic tech-iques that are commonly in use nowadays are:

. a single-patch or double-patch technique through aright atriotomy traversing the cavoatrial junction[6–9],

Accepted for publication Sept 14, 2011.

Address correspondence to Dr Metras, Service of Cardiothoracic Surgery,

. division of the SVC, baffling the PAPV drainage tothe left atrium, and insertion of the SVC into theright atrial appendage (known as Warden proce-dure) [10–12], and

. longitudinal or transverse transcaval incisions, withor without terminal SVC augmentation [13–15].

e have adopted the longitudinal, vertical transcavalpproach, as reported by Nicholson and colleagues [15],ince 2000 and report our results with that technique. Thethic Committee of our institution approved the study,nd individual consent of the patients was not required.

Material and Methods

Since 2000, all patients referred to our center for surgicalcorrection of PAPV drainage into the SVC (with orwithout sinus venosus ASD) were addressed using thevertical transcaval approach.

Operations were initiated through a median sternot-omy or a cosmetic posterior right thoracotomy. All weredone under normothermic cardiopulmonary bypass(CPB) with direct cannulation of the SVC above theinsertion of the highest anomalous pulmonary vein, with

the azygos vein well controlled and cannulation of the

0003-4975/$36.00doi:10.1016/j.athoracsur.2011.09.042

Tc

tn(dF

194 NASSAR ET AL Ann Thorac SurgTRANSCAVAL CORRECTION OF PV DRAINAGE INTO SVC 2012;93:193–6

PEDIA

TR

ICC

AR

DIA

C

inferior vena cava, both cavae with right-angled cannulas(Fig 1).

After verification of the anatomy and the start of CPB,application of aortic cross-clamp, and introduction ofcold antegrade cardioplegia, the SVC is vertically incisedjust anterior to the anomalous pulmonary veins (Fig 2).

he cavoatrial junction is always respected, and is neverrossed.

In a patient with a sinus venosus ASD, a patch is usedo divide the cavopulmonary channel in such a man-er that the SVC is directed to the right atrium

anterior to the patch) and the anomalous veins areirected toward the left atrium (posterior to the patch;ig 3). Appropriate sizing of the patch with a flat final

Fig 1. Three anomalous pulmonary veins (PV) enter the terminalpart of the superior vena cava (SVC) with the cannulation site highenough above the site of drainage of the pulmonary veins. (IVC �inferior vena cava; RA � right atrium.)

Fig 2. After a vertical incision of the superior vena cava, the orificeof the abnormal pulmonary veins (PV) is well seen. (RA � right

atrium; SVD � sinus venosus defect.)

aspect is crucial to avoid obstruction of either channel. Thepatch (autologous pericardium treated with glutaralde-hyde) is sewn using a running polypropylene 5–0 suturethat fixes the patch to the inferior and medial edges of thedefect, then to the inner aspect of the SVC, rejoining thehighest pulmonary vein orifice. The lateral edge (the inci-sion site) is closed in sandwich between the two edges ofthe caval incision (Fig 4).

The sinus venosus ASD was absent in 2 patients, andso a small right atriotomy was initiated in addition to thecaval incision to create an ASD by removal of the floor ofthe fossa ovalis. A larger patch was inserted to baffle theanomalous veins to the left atrium using a runningsuture, taking advantage of both incisions.

Fig 3. The patch is sutured to the inferior and medial edge of thesinus venosus atrial septal defect and in the terminal superior venacava anterior to the pulmonary vein (PV), dividing an anterior cavalchannel and a posterior pulmonary channel. (RA � right atrium;SVD � sinus venosus defect.)

Fig 4. After finishing the suture of the patch at the site of incision(*) in the sandwich technique, the suture stands well above the

cavoatrial junction.

TiiSiIdisncdadlt

taaaw

195Ann Thorac Surg NASSAR ET AL2012;93:193–6 TRANSCAVAL CORRECTION OF PV DRAINAGE INTO SVC

PED

IAT

RIC

CA

RD

IAC

Air was removed and aortic clamp release was done inthe standard manner, followed by weaning of the CPB.

Regular follow-up of all patients included a physicalexamination, chest roentgenogram, electrocardiogra-phy, and echocardiography. A midterm or long-term24-hour Holter electrocardiogram was obtained inmost patients.

Results

Between 2000 and 2009, 45 patients (17 males and 28females), with a mean age of 5 years (range, 8 months to70 years), were operated on using the transcaval ap-proach. Of these, 22 patients were addressed through amedian sternotomy and 23 through a cosmetic rightposterior thoracotomy. There were 2 anomalous veins in20 patients, 3 in 24 patients, and 1 patient had 4 anoma-lous veins draining into the SVC.

Persistent left SVC was found in 6 patients. PAPVdrainage without a sinus venosus ASD was found in 2patients who had a small patent foramen ovale (PFO). Forthose with a sinus venosus ASD, the defect was largeenough so that only 1 patient required enlargement ofthe defect to avoid obstruction of either channel. Fourpatients had PFO in addition to the sinus venosus ASD,and in all cases, closure of the PFO was accessiblethrough the caval incision.

Only 1 patient required augmentation of the site ofincision of the SVC to avoid SVC obstruction because theterminal SVC, associated with the presence of a left SVC,was estimated to be too small.

The mean aortic cross-clamp time was 44 minutes, witha mean CPB time of 76 minutes.

The mean stay in the intensive care unit was 1.6 days.No deaths were recorded. Morbidity included pneumo-thorax in 2 patients, pericardial effusion in 4, postcar-diotomy syndrome in 1, and a transient dysfunction ofthe right phrenic nerve in 1 that returned to normal 1month later.

The follow-up period was a mean 5.4 years (range, 4months to 10.4 years) and was complete. All patientswere regularly followed up and were asymptomaticthrough the follow-up period.

Echocardiography confirmed the absence of any resid-ual ASD. No obstruction of the pulmonary venous drain-age was noted. No echocardiographic signs of caval flowobstruction were detected, except for 1 patient who had agradient of 6 mm Hg across the cavoatrial junction in theimmediate postoperative period that completely disap-peared during follow-up.

All patients, excluding a 70-year-old woman in atrialfibrillation preoperatively, were in sinus rhythm at thelast report by their cardiologist. No incidence of newarrhythmia was detected. A 24-hour Holter echocardiog-raphy in 20 patients confirmed the absence of any ar-rhythmia. Systematic Holter monitoring is planned for all

others. a

Comment

Since the early reported techniques for PAPV drainageand sinus venosus ASD repair, the risk of complicationhas always been the motive for innovation. The mostfearful complications are arrhythmia and obstruction ofthe caval or pulmonary venous channel.

It is the complexity of the anatomic site of the anomalythat makes such a repair challenging. Injury to the sinusnode can result in a serious arrhythmia and subsequentlythe need for permanent pacing. The arterial supply to thesinus node is highly variable [16], and injury to thefeeding artery can have the same results. Theoretically,procedures trying to avoid crossing the cavoatrial junc-tion, such as the Warden procedure and the transcavaltechniques, are believed to have a minimal incidence ofarrhythmia. That is true for the transcaval techniques,with no incidence of new arrhythmia in all reviewedliterature [13–15]. Although the Warden procedure issupposed to yield the same result, the first report byWarden and colleagues [10] reported a 10% incidence ofnew arrhythmia that decreased on follow-up to 2.5%.Other reports of the same procedure have a variableincidence of arrhythmia from 0% to 23% [17].

Of interest were the results of two reports:

● the atrial flap technique by Takahashi and col-leagues [18], who reported an 93% incidence ofarrhythmia in the immediate postoperative pe-riod that decreased to 57% at the time of dis-charge, although according to their technique, thecavoatrial junction was always respected; and

● the Toronto series of 171 patients who received alateral right atriotomy crossing the cavoatrialjunction, after which Alsoufi and colleagues [1]reported a paradoxic 0% incidence of newarrhythmia.

he risk of SVC obstruction after surgical correction isnherent in all techniques, and that was the reason for thencorporation of another patch to enlarge the terminalVC. A literature review revealed a highly variable

ncidence of SVC obstruction with different techniques.yer and colleagues [19] compared the single-patch andouble-patch technique, both through atriotomy cross-

ng the cavoatrial junction, and found a statisticallyignificant difference in favor of the double-patch tech-ique. Results of the Warden procedure and its modifi-ations have shown a low incidence of SVC obstructionespite a more complex technique that sometimes needs

patch when the appendage is small and the SVCistant, and they all emphasized the importance of a

arge caval-auricular anastomosis with resection of anyrabeculation that may cause future obstruction [10–12].

With the transcaval approach, no incidence of obstruc-ion was reported whether or not the terminal SVC wasugmented [13–15]. Although we felt it was necessary tougment the terminal SVC of a patient who had a SVC ofn abnormally small caliber and an associated left SVC,e believe that the terminal SVC receiving caval flow

nd PAPV drainage flow with this anomaly is usually

p

196 NASSAR ET AL Ann Thorac SurgTRANSCAVAL CORRECTION OF PV DRAINAGE INTO SVC 2012;93:193–6

PEDIA

TR

ICC

AR

DIA

C

large enough to allow an adequate division in two parts,provided the patch is adequately sized and inserted, andwe therefore think that this approach is always possible.

Obstruction of the pulmonary venous return aftersurgical correction is uncommon, with an incidence ofabout 0% in most of the reviewed techniques [6–15].However, some experiences have reported an incidenceas high as 50% with the single-patch technique [18].

It is of some interest that the approach through a rightosterior thoracotomy (for cosmetic purposes) [20] allows

an excellent view of the area, as was done in 26 of ourpatients. The SVC and the cavoatrial junction are theclosest structures using this approach in children.

In conclusion, the vertical transcaval technique re-ported by Nicholson and colleagues [15] has by far theleast incidence of complications reported when com-pared with all other procedures. For us it has led toexcellent and reproducible results, and we considerit as our routine and basic approach. In addition,this technique is simple, highly reproducible, and canalso be completed through a cosmetic right posteriorthoracotomy.

References

1. Alsoufi B, Cai S, Van Arsdell G, et al. Outcomes after surgicaltreatment of children with partial anomalous pulmonaryvenous connection. Ann Thorac Surg 2007;84:2020–6.

2. Attenhofer Jost CH, Connolly HM, Danielson GK, et al.Sinus venosus atrial septal defect long-term postoperativeoutcome for 115 patients. Circulation 2005;112:1953–8.

3. Van Praagh R, Corsini I. Cor triatriatum: pathologic anatomyand a consideration of morphogenesis based on 13 postmor-tem cases and a study of normal development of the pulmo-nary vein and atrial septum in 83 human embryos. AmHeart J 1969;78:379–405.

4. Webb S, Kanani M, Anderson RH, et al. Development of thehuman pulmonary vein and its incorporation in the morpho-logically left atrium. Cardiol Young 2001;11:632–42.

5. Neptune WB, Bailey CP, Goldberg H. The surgical correctionof atrial septal defects associated with transposition of thepulmonary veins. J Thorac Surg 1953;25:44–50.

6. Kirklin JW, Ellis FH, Wood EH. Treatment of anomalous

pulmonary venous connections in association with inter-atrial communications. Surgery 1956;39:389.

7. Ohmi M, Mohri H. A single pericardial patch technique forrepair of partial anomalous pulmonary venous drainageassociated with sinus venosus atrial septal defect. AnnThorac Surg 1988;46:360–1.

8. Kouchoukos N, Blackstone EH, Doty DB, Hanley FL, KarpRB. Atrial septal defects and partial anomalous pulmonaryvenous connection. In: Kouchoukos N, Blackstone E, Doty D,Hanley F, Karp R, eds. Kirklin/Barratt-Boyes Cardiac Sur-gery. 3rd ed. Philadelphia: Churchill Livingstone; 2003:715–52.

9. Backer CL, Mavroudis C. Paediatric cardiac surgery. 3rd ed.Philadelphia: Mosby; 2003:283–97.

10. Trusler GA, Kazenelson G, Freedom RM, et al. Late resultsfollowing repair of partial anomalous pulmonary venousconnection with sinus venosus atrial septal defect. Journal ofThoracic and Cardiovascular Surgery 1980;79:776–81.

11. Warden HE, Gustafson RA, Tarnay TJ, et al. An alternativemethod for repair of partial anomalous pulmonary venousconnection to the superior vena cava. Ann Thorac Surg1984;38:601–5.

12. Atsushi N, Toshikatsu Y, Koji K, et al. Partial anomalouspulmonary venous connection to the superior vena cava.Ann Thorac Surg 2006;82:978–82.

13. Gaynor JW, Burch M, Dollery C, et al. Repair of anomalouspulmonary venous connection to the superior vena cava.Ann Thorac Surg 1995;59:1471–5.

14. Victor S, Nayak VM. Transcaval repair of sinus venosusdefect. Using a butterfly-shaped patch. Tex Heart Inst J1995;22:304–7.

15. Nicholson IA, Chard RB, Nunn GR, et al. Transcaval repairof the sinus venosus syndrome. J Thorac Cardiovasc Surg2000;119:741–4.

16. Baskett R, Ross DB. Superior vena cava approach to repair ofsinus venosus atrial septal defect. J Thorac Cardiovasc Surg2000;119:178–80.

17. Futami C, Tanuma K, Tanuma Y, et al. The arterial bloodsupply of the conducting system in normal human hearts.Surg Radiol Anat 2003;25:42–9.

18. Takahashi H, Oshima Y, Yoshido M, et al. Sinus nodedysfunction after repair of partial anomalous pulmonaryvenous connection. J Thorac Cardiovasc Surg 2008;136:329–34.

19. Iyer AP, Somanrema K, Pathak S, et al. Comparative study ofsingle- and double-patch techniques for sinus venosus atrialseptal defect with partial anomalous pulmonary venousconnection. J Thorac Cardiovasc Surg 2007;133:656–9.

20. Metras D, Kreitmann B. Correction of cardiac defects

through a right thoracotomy in children. J Thorac Cardio-vasc Surg 1999;117:1040–2.