Left Ventricle Is Better Suited as PulmonaryVentricle in Simple Transposition With SeverePulmonary HypertensionRajesh Sharma, MCh, Shiv Kumar Choudhary, MCh, Anil Bhan, MCh,Rajnish Juneja, DM, Shyam Sunder Kothari, DM, Anita Saxena, DM, andPanangipalli Venugopal, MChCardiothoracic Centre, All India Institute of Medical Sciences, New Delhi, India

Background. The conventional treatment of transposi-tion of great arteries with prepared left ventricle is anarterial switch operation. This, in our experience, doesnot hold for patients with transposition of great arterieswith intact ventricular septum where the left ventriclecontinues to be prepared secondary to severe pulmonaryarterial hypertension without an immediately reversiblecause.

Methods. Ten infants with D-transposition of the greatarteries with essentially intact interventricular septumand severe pulmonary arterial hypertension underwentsurgical treatment. Age ranged from 3 to 6 months (mean,4.2 months). One of these patients had a large ductuswith left to right shunting but the others had no intra- orextracardiac shunt to account for their pulmonary hyper-tension. All 10 had “prepared” left ventricles. The first 4children underwent an arterial switch operation. Un-eventful surgery was followed by prolonged ventilatordependence in all 4 with occurrence of severe pulmonaryarterial hypertension every time weaning from ventilatorwas attempted. This was accompanied by metabolicacidosis and features of right heart failure. Only 1 patient

with large ductus could be extubated and dischargedfrom hospital. Subsequently, the other 6 infants under-went a Senning repair.

Results. There was no early mortality. All patientswere separated from mechanical ventilation within 48hours of surgery without blood gas derangement or heartfailure despite elevated pulmonary artery pressure in all.The child with the arterial switch operation has pulmo-nary artery pressure of 50% systemic 4 years followingrepair; although among the Senning group, 2 patientscontinue to have pulmonary artery pressure more than60% of systemic and 4 have normal pulmonary arterypressure at a mean follow-up of 1 year.

Conclusions. Atrial level repairs seem to perform betterthan arterial level repairs in children having TGA withpersistent pulmonary artery hypertension without a cor-rectable cause. Better tolerance of pulmonary arterialhypertension in this group is probably consequent to thesuperior ability of the left ventricle to tolerate a pressureload in the early postoperative period.

(Ann Thorac Surg 2002;74:1612–5)© 2002 by The Society of Thoracic Surgeons

The normal course of the left ventricle (LV) followingbirth in simple transposition (TGA) is one of invo-

lution of left ventricular mass in response to drop inpulmonary artery pressure (PAP). Rarely, patients withTGA, even in the absence of a sizable ventricular septaldefect (VSD) or patent ductus arteriosus (PDA), mayhave pulmonary arterial hypertension (PAH) beyond theneonatal period possibly secondary to primary pulmo-nary hypertension [1–4]. Although conventional modernday thinking would have us believe that anatomic repairis the treatment of choice for all patients having TGAwith an LV prepared for systemic afterload [5], our ownexperience with this disease entity cautions us fromaccepting this generalization for all categories of patients.

Patients and Methods

PatientsOf 169 patients with TGA and essentially intact ventric-ular septum (IVS) undergoing the arterial switch opera-tion, 4 infants had systemic level pulmonary artery pres-sures at presentation beyond the neonatal period. In thesame time period, 6 infants with TGA IVS and severePAH underwent a modified Senning repair. These 10patients form the basis of this presentation. Age of thesepatients ranged from 3 to 6 months (mean, 4.2 months).Three patients had a small (2 mm) restrictive VSD, and 1patient had a large PDA with left to right shunt.

All patients underwent cardiac catheterization andcineangiography. Pulmonary vascular resistance was notcalculated because of its unreliability in transpositioncomplexes. Left ventricular mass (indexed to body sur-face area) and posterior wall thickness was calculatedechocardiographically by methods described in an earlierstudy [6]. Besides numerical values of these measure-

Accepted for publication June 7, 2002.

Address reprint requests to Dr Sharma, B-404 Adarsh Palace, Block-5,Jaya Nagar, Bangalore, India; e-mail: rsharmacvs@hotmail.com.

© 2002 by The Society of Thoracic Surgeons 0003-4975/02/$22.00Published by Elsevier Science Inc PII S0003-4975(02)03880-8

ments to assess left ventricular preparedness, we reliedsignificantly on the visual appearance of the left ventricleand the interventricular septum on the cross-sectionalechocardiogram. In cross-sectional view, circular leftventricle with the interventricular septum convex to-wards the right ventricle, was considered favorable. Thepreoperative profile of these patients is summarized inthe Table 1.

SurgeryThe arterial switch operation was performed on cardio-pulmonary bypass and deep hypothermia utilizing singleright atrial or bicaval cannulation, depending on theexposure obtained by the former. Details of surgicaltechnique were along the accepted lines [7], with coro-nary transfer being to excised wedges, punch holes, ormedially based trapdoor flaps in the proximal pulmonaryartery. The Lecompte maneuver was performed for allthe patients. Pulmonary artery reconstruction was by agenerous patch of autologous pericardium sutured prox-imally to the deficiency in the sinuses created by excisionof the coronary buttons and distally to the opening ofpulmonary artery bifurcation. All patients had a pulmo-nary artery line inserted just before sternal closure. All 4patients also had a small (3 to 4 mm) fenestration leftbehind in the atrial septum. A tiny, 2 mm VSD was alsoleft open in one of the patients.

The Senning operation was performed utilizing deepor moderate hypothermia using direct caval cannulationby the technique described by Quaegebeur and cowork-ers [8]. A pulmonary artery line was placed in all of thesepatients. A tiny VSD (2 mm) was left open in 2 of thesepatients.

Postoperative CareAll patients were kept paralyzed and sedated until theywere judged to have good cardiac output clinically. They

were then allowed to awaken with continuous PAPmonitoring. Weaning from the ventilator was com-menced once arterial oxygenation was good and thirdspace fluid had been mobilized. PAH crises were treatedby sedation and hyperventilation with 100% oxygen.Nitric oxide was not utilized in any of these patients dueto nonavailability in this time period.

Follow-UpClinical condition of the patients was noted. Ventricularfunction and status of the atrioventricular valves wasobtained by echocardiography, performed at 1 month, 3months, and 6 months after hospital discharge, and thenevery 6 months thereafter. All survivors underwent car-diac catheterization after an interval ranging from 6months to 4 years.

Results

Arterial Switch Group (n � 4)Surgery was uneventful and patients could be weanedfrom cardiopulmonary bypass without difficulty. In theintensive care unit, resting PAP ranged between 30% to100% of systemic arterial pressure. The postoperativecourse was marked by periods of CO2 retention when-ever weaning from ventilator was attempted. This wasaccompanied by bronchoconstriction and elevation ofPAP to systemic or near systemic levels. Two patientswho were extubated with midazolam infusion developedmetabolic acidosis, hypercarbia, and features of rightheart failure (hepatomegaly, ascites) and had to be rein-tubated and mechanically ventilated. Pulmonary regur-gitation, which was judged to be mild when the PAP waslow, was regularly seen to worsen with rise in PAP in thisgroup. Ultimately only the patient with the large ductuscould be weaned from the ventilator. The other 3 patients

Table 1. Profile of Patients

PatientNumber

Year ofSurgery

Age(months)

Peak SystolicPressure(mm Hg)

SystolicLV/AortaPressure

Ratio

FemoralArterial O2Saturation

(on 100% O2inhalation)

LV MassIndex(g/m2)

EchocardiographicLV Appearance

LVPWThickness

(mm)Surgical

ProcedureAorta LV PA

1* 1996 3 90 90 § 1.0 80 90.6 Favorable† 4.5 ASO2 1994 4 104 85 § 0.8 60 76.0 Favorable† 4.2 ASO3 1995 4 84 75 72 0.9 50 94.5 Favorable† 4.3 ASO4‡ 1996 5 92 72 70 0.8 40 83.6 Favorable† 4.7 ASO5‡ 1995 4 110 107 § 1.0 50 94.6 Favorable† 4.8 Senning6‡ 1996 3 82 80 § 1.0 55 84.6 Favorable† 4.4 Senning7 1997 4 94 65 64 0.7 55 79.8 Favorable† 3.9 Senning8 1997 4 102 80 80 0.8 65 84.5 Favorable† 4.2 Senning9 1997 6 105 85 80 0.8 60 94.2 Favorable† 4.8 Senning

10 1998 5 95 74 § 0.75 40 87.6 Favorable† 4.6 Senning

*Patient 1 had patent ductus arteriosus. † On cross-sectional view, circular left ventricle and interventricular septum convex towards the rightventricle. ‡ Patients 4, 5, and 6 had a 2-mm restrictive ventricular septal defect. § In patients 1, 2, 5, 6, and 10, PA was not entered and pulmonaryartery pressure was considered to be equivalent to the LV pressure as there was no echocardiographic LV-PA gradients.

ASO � arterial switch operation; LV � left ventricle; LVPW � left ventricular posterior wall; PA � pulmonary artery; PVRI � pulmonaryvascular resistance index.

1613Ann Thorac Surg SHARMA ET AL2002;74:1612–5 TGA WITH SEVERE PAH

became ventilator dependent and died. Autopsy was notperformed because permission was denied. During thesame time period, a group of 163 neonates with TGA IVSunderwent arterial switch operation within the firstmonth of life. There were 12 early deaths in this group.

Senning Group (n � 6)All 6 patients had uneventful repair and weaning fromcardiopulmonary bypass. Here, as in the prior group,PAP ranged from 30% to 100% of the systemic arterialpressure and had a tendency to rise whenever ventilatoryweaning was attempted. Bronchospasm also developedalong with elevation of the PAP. There was, however, nometabolic acidosis or features of right heart failure. Therewas no echocardiographic aggravation of pulmonaryvalvular regurgitation at different levels of PAP as op-posed to the ASO group. All 6 could be extubated,despite CO2 retention, with some sedation to combat theagitation that accompanied spontaneous ventilation. All6 patients could be discharged from the hospital.

Follow-UpIn the single survivor following the arterial switch oper-ation, the PAP had dropped to 50% of the systemicarterial pressure at 4 years after surgery. Of the patientswho survived following atrial level repair, 4 had normalPAP, whereas 2 had 60% systemic level PAP at a meanfollow-up of 1 year.

Comment

The arterial switch operation is widely reported as thepreferred treatment for TGA with a prepared LV [5, 7].Our experience with TGA IVS with idiopathic severePAH, however, seems to dictate otherwise. RV failure inresponse to pulmonary dysfunction and severe PAHseemed to be the principle cause of demise in thepatients who had undergone the arterial switchoperation.

In all of the 10 patients included in this study varyingdegree of difficulty was encountered in separating thepatients from mechanical ventilation. Common to thesewas persistently elevated PAP that had a tendency to riseto systemic level whenever the effort of spontaneousrespiration was sought to be enrolled. Respiratory insuf-ficiency was evidenced in all children by way of differentmagnitudes of elevation of carbon dioxide in the bloodgas reports. The differentiating feature between the twogroups was the absence of features of cardiac compro-mise in the children who had undergone the Senningprocedure. We postulate this to be primarily because ofthe superiority of the left ventricle in the pulmonaryposition. Another factor of importance may be unalteredpulmonary valve function following a Senning ratherthan an arterial level repair. Some neopulmonary incom-petence is anticipated in all arterial switch operationswith persistent PAH and would necessarily worsen right-sided hemodynamics.

The right ventricle (RV) is well known to misbehave inthe face of increased afterload, while the LV is known to

function well in the presence of pressure load. Extendingthis corollary to repaired D-TGA with an RV versus anLV as the pulmonary ventricle, it seems logical thatresidual PAH would be better tolerated by the morpho-logic LV.

The paradox of incriminating a ventricle that is capableof supporting the systemic circulation for a lifetime inmost patients [9], for failure to withstand the pulmonaryafterload is glaring. PAP in most of the instances herewas subsystemic and only intermittently systemic. Toexplain this paradox it may be assumed that intermittentdrops in PAP may render the RV less effective to face asystemic level elevation when it does occur. The damag-ing effects of myocardial anoxia sustained during aorticclamping at the time of repair can also be invoked exceptthat the same right ventricle tolerates systemic afterloadmuch better than pulmonary afterload following repair.

Successful arterial level repair has been reported inliterature for neonates with TGA IVS and persistentpulmonary hypertension. Invariably, the postoperativecourse has been extremely stormy and managementinvolved the use of inhaled nitric oxide and even extra-corporeal membrane oxygenation at times [10, 11]. Thesingle denominator for the successful outcome has beenthe subsidence of PAH in these patients all of who wereneonates in the postoperative period. Our patients werewell beyond the neonatal period and although all werestill in infancy, no early subsidence of PAH could beexpected except in the child with the large PDA and leftto right shunt.

Given the extremely poor early results of the ASO inthe subgroup of patients with systemic level PAH, wepropose that the procedure of choice for patients of TGAwho are likely to have important PAH in the periodfollowing repair is an atrial level switch. We are loath tolabel these atrial level repairs in all patients in thecategory of simple TGA with PAH as “palliative” [12–14]for the reason that postoperative PAP elevations may notbe to systemic level and the possibility of regression tonormal level still remains a definite possibility especiallyin the younger patients. In our experience, normalizationof PAP has taken place in 4 of the survivors followingatrial level repair. In those patients where the procedureis likely to be palliative in view of advanced age orseverely elevated PAP, creation of a VSD [13] does notseem to be necessary based on our experience of com-plete freedom from right-sided failure with the LV aspulmonary ventricle. This we attribute to the LV beingperfectly adapted to the situation and not manifestingany failure even in the face of severe PAH, thus obviatingthe need for decompression.

The downside of having a systemic RV obviouslydetracts from the ideal situation of a systemic LV. Failureof the RV to function in the systemic circulation in thelong run is one of the drawbacks of atrial level repair ofTGA [9]. Revision surgery for systemic ventricular failure[15, 16] can still be contemplated if PAP normalizes inthose patients where it is strongly indicated.

The need to make this compromise of shifting fromarterial switch as the strategy for all TGA with severe

1614 SHARMA ET AL Ann Thorac SurgTGA WITH SEVERE PAH 2002;74:1612–5

PAH would not arise if effective treatment of PAH ex-isted. The much heralded advent of inhaled nitric oxide(iNO) has sadly, in our and in other’s experience, not metwith any benefit to the late presenters of congenital heartdisease with severe PAH [17]. Postoperative manage-ment of residual PAH is, therefore, fraught with thepossibility of RV failure in response to either sustained orepisodic elevation of PAP. In this situation, where effec-tive predictable treatment of PAH is not existent, itbehooves us to devise surgical strategies that provide thepatient with congenital heart disease with PAH, whopresents late, with a better chance of survival than thatexist with accepted methodologies, until better methodsof medical management of PAH become available. Theshift from arterial switch to atrial level repairs for pa-tients of simple TGA with severe PAH would be one suchstep.

References

1. Newfeld EA, Paul MM, Muster AJ, Idriss FS. Pulmonaryvascular disease in complete transposition of the greatarteries:a study of 200 patients. Am J Cardiol 1974;34:75–82.

2. Lakier JB, Stanger P, Heymann MA, Hoffman JL, RudolphAM. Early onset of pulmonary vascular obstruction in pa-tients with aorto-pulmonary transposition and intact sep-tum. Circulation 1975;51:875–80.

3. Hawker RE, Freedom RM, Rowe RD. Persistence of fetalpattern of circulation in transposition of the great arteries.Hopkins Med 1974;134:107–17.

4. Kumar A, Taylor GP, Sandor GG, Patterson W. Pulmonaryvascular disease in neonates with transposition of the greatarteries and intact ventricular septum. Br Heart J 1993;69:442–5.

5. Castaneda AR, ed. Anatomic correction of transposition ofthe great arteries at the arterial level. In Sabiston DC Jr,Spencer FC, eds. Surgery of the chest. Philadelphia: WBSaunders, 1990:1435–46.

6. Iyer KS, Sharma R, Kumar K, Bhan A, Kothari SS, Saxena A,Venugopal P. Serial echocardiography for decision making

in rapid two-stage arterial switch operation. Ann ThoracSurg 1995;60:658–64.

7. Castaneda AR, Jonas RA, Mayer JE Jr, Hanley FL. D-transposition of great arteries. In EDITORS’ NAMES? Car-diac Surgery of the Neonate and Infant. Philadelphia: WBSaunders, 1994:409–38.

8. Quaegebeur JM, Rohmer J, Brom AG. Revival of the Senningoperation in the treatment of transpositon of the greatarteries. Preliminary report on recent experience. Thorax1977;32:517–24.

9. Turina MI, Siebenmann R, von Segesser L, Schonbeck M,Senning A. Late functional deterioration after atrial correc-tion for TGA. Circulation 1989;80(3, Pt 1):1162–7.

10. Luciani GB, Chang AC, Starnes VA. Surgical repair oftransposition of the great arteries in neonates with persistentpulmonary hypertension. Ann Thorac Surg 1996;61:800–5.

11. Chang AC, Wernovsky G, Kulik TJ, Jonas R, Wessel DL.Management of the neonates with transposition of the greatarteries and persistent pulmonary hypertension. Am J Car-diol 1991;68:1253–5.

12. Lindesmith GG, Stiles QR, Tucker BL, Gallaher ME, StantonRE, Meyer BW. The Mustard operation as a palliative pro-cedure. J Thorac Cardiovasc Surg 1972;63:75–80.

13. Stark J, deLeval MR, Taylor JFN. Mustard operation andcreation of VSD in two patients with transposition of thegreat arteries, intact ventricular septum and pulmonaryvascular disease. Am J Cardiol 1976;38:524–7.

14. Levinsky L, Srinivasan V, Gingell RL, Choh JH, Pieroni DR,Fisher J, Subramanian S. Senning repair with ductal decom-pression. Palliative approach to d-TGA and irreversiblepulmonary vascular disease. Am Heart J 1983;106:409–12.

15. Mee RB. Severe right ventricular failure after Mustard orSenning operation. Two stage repair: pulmonary arterybanding and switch. J Thorac Cardiovasc Surg 1986;92:385–90.

16. Poirier NC, Mee RB. Left ventricular reconditioning andanatomical correction for systemic right ventricular dysfunc-tion. Semin Thorac Cardiovasc Surg Pediatr Card Surg Annu2000; 3:198–215.

17. Sharma R, Raizada N, Choudhary SK, et al. Does inhalednitric oxide improve survival in operated congenital heartdisease with severe pulmonary hypertension? Indian Heart J2001;53:48–55.

1615Ann Thorac Surg SHARMA ET AL2002;74:1612–5 TGA WITH SEVERE PAH