The double switch for atrioventricular discordance

INNOVATIVE AND COMPLEX PROCEDURES

The Double Switch for Atrioventricular Discordance William J. Brawn Conventional surgery for atrioventricular discordance usually associated with ventricular arterial discordance leaves the morphologic right ventricle in the systemic circulation. Long-term follow-up results with this approach reveal a high incidence of right ventricular failure. The double switch procedure was introduced to restore the morphologic left ventricle to the systemic circulation. This operation is performed in two main ways: the atrial-arterial switch and the atrial switch plus Rastelli procedure. This double switch approach has been successful at least in the medium term in abolishing morphologic right ventricular failure and its associated tricuspid valve regurgitation. In the atrial-arterial switch group, there is an incidence of morphologic left ventricular dysfunction, sometimes associated with neoaortic valve regurgitation, and the minority of cases need aortic valve replacement. The long-term function of the morphologic left ventricle and the aortic valve need careful surveillance in the future. The atrial-Rastelli group of patients has not in the medium term shown evidence of ventricular dysfunction but will require change on a regular basis of their ventricular to pulmonary artery valved conduits. Semin Thorac Cardiovasc Surg Pediatr Card Surg Ann 8:51-56 © 2005 Elsevier Inc. All rights reserved. KEYWORDS: Congenitally corrected transposition, double switch procedure, Senning procedure, Rastelli procedure

T

he double switch describes a group of procedures including atrial-arterial switch, atrial switch plus Rastelli procedure, and atrial switch with tunneling of the aorta to the morphologic left ventricle for patients with atrioventricular and ventricular arterial discordance or congenitally corrected transposition of the great arteries. This procedure was introduced over the last 20 years or so because of the recognition of the longer-term problem of the conventional repair for congenitally corrected transposition (ie, failure of the systemic morphologic right ventricle associated with tricuspid valve regurgitation). The double-switch procedure restores the morphologic left ventricle to the systemic circulation and may provide better long-term survival and reduced morbidity than the conventional or physiologic repair.

Birmingham Children’s Hospital, Birmingham, United Kingdom. Address reprint requests to William J. Brawn, MD, Diana, Princess of Wales Children’s Hospital, Steelhouse Lane, Birmingham, B4 6NH, United Kingdom.

1092-9126/05/$-see front matter © 2005 Elsevier Inc. All rights reserved. doi:10.1053/j.pcsu.2005.02.003

Background and Problems with the Conventional or Physiologic Repair for Congenitally Corrected Transposition of the Great Arteries The patients for whom a double switch is to be considered have atrioventricular and ventricular arterial discordance, also known as congenitally corrected transposition of the great arteries (CCTGA). CCTGA is a rare condition, comprising about 0.5% of patients with congenital heart problems.1 It can take time for individual surgeons and centers to gain experience in managing these patients and to see the long-term results of surgery. Associated cardiac anomalies are common and include ventricular septal defect (VSD), pulmonary stenosis or atresia, tricuspid valve dysplasia with Ebsteinoid-like abnormalities of the valve, and abnormalities of the conduction system leading, in the natural history of the condition, to complete heart block.2,3 When only these associated cardiac anomalies are repaired, the 51

52 morphologic right ventricle is a physiologic or conventional repair. Several publications from different countries have shown that the morphologic right ventricle in association with developing tricuspid regurgitation can fail in an unpredictable way after this conventional surgery.4 – 8 Failure of the right ventricle associated with tricuspid regurgitation can occur during the natural history of unoperated CCTGA, making a conventional repair difficult or impossible. The reasons why the right ventricle and tricuspid valve deteriorate are poorly understood. Tricuspid valve regurgitation with right ventricular failure occurs commonly after a conventional repair even though the tricuspid valve may seem to be normal with wellpreserved right ventricular function before surgery. In the presence of dysplasia or Ebsteinoid abnormalities of the tricuspid valve, the onset of tricuspid regurgitation and associated right ventricular failure occur more frequently with or without prior surgery. Conventional surgery in the presence of tricuspid valve dysplasia is associated with increasing tricuspid valve regurgitation and right ventricular failure. Progressive tricuspid valve regurgitation can occur in the presence of what was seemingly a normal competent tricuspid valve after conventional surgery. This is thought to be due to alteration in the cardiac hemodynamics such that the lower morphologic left ventricular pressure in the pulmonary circulation in association with a higher systemic right ventricular pressure allows displacement of the ventricular septum toward the morphologic left ventricle, putting tension on the septal structures of the tricuspid valve and causing tricuspid valve regurgitation to develop. This becomes a vicious circle, with more regurgitation creating annular tricuspid valve dilatation resulting in more heart failure and further tricuspid regurgitation. Pulmonary artery banding with the aim of retraining the morphologic left ventricle to allow a double switch to be performed can realign the septum with a higher morphologic left ventricular pressure and can variably decrease the amount of tricuspid valve regurgitation.9,10 A spectrum of clinical scenarios can occur in this group of patients, from the infant with severe morphologic right ventricular failure needing inotropes and ventilatory support, to the patient with gradually increasing tricuspid regurgitation with development of right heart failure, to the rarer older patient who may be asymptomatic with well-preserved cardiac function and tricuspid valve function well into old age. This can make management decisions difficult for patients with CCTGA. General consensus seems to have evolved over the last 20 years or so that because of the unpredictable outcomes and demonstrated failure of the conventional repair in the longer term, a double-switch procedure is the better option when the cardiac morphology allows.

Indications for the Double-Switch Procedure The indications, timing, and type of double switch vary according to the morphology of the heart, the clinical state of

W.J. Brawn the patient, and the patient’s age. Patients with CCTGA can be divided into two main groups: those without associated cardiac anomalies and those with associated cardiac anomalies. The algorithm shown in Fig 1 illustrates possible management pathways and the preferred pathways leading to a double switch that we like to use at our institution. In patients without associated anomalies, in the presence of new or increasing tricuspid regurgitation with or without morphologic right ventricular failure we would advocate pulmonary artery banding to retrain the morphologic left ventricle and later perform an atrial and arterial switch procedure. Our cutoff for retraining the morphologic left ventricle is about 15 years of age, and although pulmonary artery banding can be considered in older patients, we have no success in retraining such a patient, nor have we been able to reduce tricuspid regurgitation by pulmonary artery banding enough to clinically improve the patients. We are not in favor of retraining patients over 15 years of age. Other options, such as tricuspid valve repair or replacement or later cardiac transplantation, would in our view have to be considered. In the group of patients with associated anomalies, tricuspid valve dysplasia with tricuspid regurgitation and congestive cardiac failure is an indication for atrial-arterial switch with or without pulmonary artery banding before surgery to retrain the morphologic left ventricle depending on the degree of pulmonary hypertension and hence the degree of preconditioning of the morphologic left ventricle before surgery. If the morphologic left ventricle is developing systolic pressures of 80% systemic with well-preserved ventricular function, then we would be prepared to go straight to an atrial-arterial switch without the need for pulmonary artery banding. Most patients with pulmonary atresia and pulmonary stenosis are palliated with a systemic shunt in infancy and come forward for consideration of management of their increasing cyanosis in the first few years of life. These patients would go forward for an atrial switch in association with a Rastelli procedure, which involves placing a valved conduit from the morphologic right ventricle to the pulmonary arteries. A subgroup of these patients with CCTGA and a VSD and pulmonary stenosis with fairly well-balanced circulation where the systemic oxygen saturations are in the 80s are relatively asymptomatic, and in this group of patients we would not advocate surgical intervention until they became symptomatic by virtue of increasing cyanosis. It seems to us that the risk of the operation, the atrial switch with a Rastelli operation, the need for conduit change in the future, and the risk of heart block with pacemaker placement outweigh the risk of the patients having a good quality of life with stable hemodynamics. The algorithm for the management of congenitally corrected transposition (Fig 1) shows the pathways to cardiac transplantation. At each of the stages it may be necessary to consider such an option depending on the patient’s ventricular function. In patients with pulmonary atresia, it is not possible to tunnel the VSD to the aorta when the VSD is in an inlet position between the mitral and tricuspid valves; a systemic or venous shunt with a later Fontan procedure may be

The double switch for AV discordance

53

Figure 1 Algorithm for the management of congenitally corrected transposition

necessary. In our experience this has been minority of patients, and we have only seen two patients for whom it was not possible to perform an atrial switch with a Rastelli operation. Rarely, atrioventricular discordance is associated with double outlet of the right ventricle, which is associated with a VSD. In these patients, an atrial switch with tunneling of the VSD to the aorta can be performed.

Surgical Techniques Cardiopulmonary bypass with core cooling down to 22°C nasopharyngeal or 18°C nasopharyngeal if circulatory arrest is anticipated is routinely used. The heart is arrested with cold crystalloid cardioplegia and iced slush irrigation throughout the procedure. Circulatory arrest is used intermittently, in particular in cyanotic patients where there is often a lot of open heart return, making visualization of the atrial pathways difficult. We use Aprotinin infusion during these operations and tissue fibrin glue to help with haemostasis. Our unit is more familiar with the Senning procedure11,12 to perform the atrial component of the double switch. The superior vena cava is cannulated high beneath

the innominate vein to allow for room at the superior vena caval right atrial junction for placement of the pulmonary venous suture line. The inferior vena cava is cannulated as low down as possible and somewhat anterior to preserve the Eustachian valve. The Eustachian valve can then be incorporated into the atrial systemic venous pathway. When the heart is dissected from previous adhesions in situations where there has been previous surgery, the heart is cannulated, bypass is started, the aorta is cross clamped, and the heart cardiopleged.

The Atrial-Arterial Switch On bypass, the interatrial groove is developed and the baffles for the Senning reconstruction are formed. The right atrium is opened longitudinally just anterior to the cristaterminalis, and the incision is extended inferiorly and parallel to the crista to reach the Eustachian valve. The interatrial septum is opened adjacent to the mitral valve annulus, and incisions are extended superiorly and then inferiorly. The superior incision is cut back deeply into the limbus and into the root of the superior vena cava. The incision usually exits through the previously delineated interatrial groove. The atrial septum is

W.J. Brawn

54 separated from the right pulmonary veins hinging on the wall of the right atrium to create a posterior wall to the systemic venous chamber. The pulmonary veins are inspected, and often the right pulmonary veins are cut back to enlarge the opening of these for the pulmonary venous atrium. This incision also relocates the pulmonary venous pathway more laterally away from the systemic venous channel. Once the venous baffles are created, if there is a VSD present it is closed through the mitral valve. The VSD closure sutures from the morphologic right ventricular surface of the VSD are placed superiorly to avoid the superiorly placed conduction system on the morphologic left ventricular side.13 There can be traction on the crux of the heart, particularly when the heart is in a mesocardiac or dextrocardiac position; this may create temporary of permanent heart block even if the sutures are carefully placed. We generally use a smooth VSD patch in case there is neoaortic valve regurgitation because hitting a rough patch might cause hemolysis. At this point, the arterial switch is performed. The aorta is transected above the entrance of the coronary arteries, and in our experience these have been in the facing sinuses adjacent to the rightward sided pulmonary artery. Careful dissection may be necessary between the two great vessels in particular when a pulmonary artery band has been placed. Large cuffs of aorta are taken with the coronary arteries as they are removed from their respective sinuses; these are mobilized and relocated to medially hinged flaps created in the proximal part of the transected pulmonary artery. Often, the proximal pulmonary artery is enlarged, and it may be necessary to reduce the size by tailoring what becomes the noncoronary sinus by resection of tissue from the noncoronary sinus of the neoaorta. Likewise, the area where the pulmonary artery band has been placed may be somewhat smaller than necessary, and the relocation of the coronary arteries with aortic wall tissue may satisfactorily enlarge the new aorta where they are inserted. The defects in the old aorta are repaired with heterologous material, such as pulmonary homograft patch or bovine pericardium. Before reconstructing the new aorta, consideration has recently been given to tailoring the aortic root. Neoaortic valve regurgitation of mild variety is not uncommon after this procedure, and we have recently tailored the root by circumferential 4-0 PDS suture tied down around the appropriately sized Hegar dilator placed in the new aortic root. Care is taken to tailoring in the sinuses, particularly in the noncoronary sinus if it is aneurysmal proximal to a previously placed pulmonary artery band. In general, if the vessels are strictly side by side, the pulmonary arteries are left behind the aorta; when the vessels are in a more anterior posterior position the pulmonary arteries are relocated anterior to the aorta. At this point the aorta is reconstructed, and the Senning procedure is completed. The posterior wall of the systemic venous chamber is reconstructed starting the suture line at the posterior lip of the left atrial appendage then passing superiorly and inferiorly. It is important to leave plenty of space superiorly between the posterior margins of the superior vena cava and the suture line because the tricuspid valve is set deeply (ie, more than in a usual dextro transposition). Often, the morphologic left atrium is markedly dilated, and it is important to gather

tissue from the left atrial wall onto the posterior edge of the systemic venous baffle. We have not supplemented the posterior baffle of the systemic venous atrium. The anterior wall of the systemic venous atrium is then reconstructed starting inferiorly at the Eustachian valve and working superiorly. Depending on the local anatomy, the coronary sinus can be allowed to drain into the systemic or the pulmonary venous atrium. The suture line is continued superiorly along the margins of the cut edge of the inter-atrial septum between the mitral and tricuspid valves onto the cristaterminalis around the atrial appendage. Occasionally, when the superior vena caval pathway is narrowed, then the anterior aspect of the suture line is supplemented with an additional piece of foreign material, such as bovine pericardium or pulmonary Homograft patch. We think this is because the tricuspid valve is set more deeply in congenitally corrected transposition and because the angulation from the superior vena cava back to the opening in the tricuspid valve is more acute. The pulmonary venous pathway is then reconstructed, usually directly anastomosing the free edge of the right atrium to the cut edge of the right pulmonary veins. Bypass rewarming is usually started at this time, and usually the heart can be de-aired, and the cross clamp can be removed so that the heart can be reperfused while the pulmonary arteries are reconstructed. Sometimes, because of previous surgery, the pulmonary arteries are tight across the aorta when the Le Compte maneuver has been performed. We then divide the right pulmonary artery to let in an extension tube, usually a devalved Contegra Conduit (Medtronic Ltd., Minneapolis, MN) to lengthen the pulmonary arteries and relieve tension of the pulmonary bifurcation on the posterior aorta and coronary arteries. A full compliment of pacing wires are placed, a left atrial pressure line is placed, and usually the skin only is closed to prevent the heart being from compressed by the closure of the sternum. The chest is usually closed formally within 1 or 2 days.

The Atrial Switch with Rastelli Procedure The Senning procedure is used for the atrial component of this procedure. Because of the high incidence of mesocardia or dextrocardia in situs solitus or levocardia in situs inversus in patients with associated pulmonary atresia or pulmonary stenosis and large VSD, the free right atrial wall between the atrioventricular groove and the pulmonary veins is often reduced in size. In these cases, it may be necessary to supplement the pulmonary venous reconstruction with a patch of material such as pulmonary Homograft or bovine pericardium to restore volume to the new pulmonary venous chamber as it flows around the outside of the systemic venous chamber. This prevents distortion and tension on the atrioventricular groove and ventricles, which would occur with a direct anastomosis of the narrowed free margin of the right atrium directly to the cut edge of the pulmonary veins. An alternative is to use a Shumaker modification,14 suturing the edge of the right atrium to in situ native pericardium. Having

The double switch for AV discordance created the atrial incisions, the completion of the Senning procedure is deferred until the VSD has been closed and the distal end of the valved conduit has been sutured to the pulmonary arteries. To do this, an incision is made in the morphologic right ventricle between coronary arteries being careful to stay well away from the aorta so that the incision does not impinge on the aortic valve. The ventriculotomy is enlarged, and through this the VSD is visualized and baffled over to the aorta with a Dacron or Gore-Tex patch sutured with interrupted mattress sutures with Teflon pledgets. Occasionally, if the suture line is straightforward, a continuous suture is used superiorly around the infundibulum and the aorta. Usually, there are marked trabeculae and infundibular folds that support the aortic valve, and it is important to be careful to close any potential defects beneath these trabeculae that might result in a residual VSD. A valved conduit reconstruction is then performed between the right ventriculotomy and the pulmonary arteries. Depending on the distance between the ventriculotomy and the pulmonary artery, which is usually long, we tend to use a Hancock valved conduit (Medtronic Ltd.), a Contegra valved conduit (Medtronic Ltd.), or an aortic Homograft. Usually, the graft is placed to the left side of the aorta, but it may be placed on the right side in extreme dextro position of the heart in situs solitus or extreme levo position in situs inversus. Although we try to get the majority of the conduit away from the back of the sternum, this is not always possible, and usually part of the conduit lies behind the sternal incision. The ventriculotomy suture line to the conduit is reinforced with Teflon pledgeted mattress sutures. These pass inside outward in the heel of the graft and outside in toward the toe. The atrial component of this double switch is completed, and, as for the atrial-arterial switch procedure, the same allowances are made for chest closure, and the chest may not be closed until a few days after surgery if low cardiac output is a problem. A Gore-Tex membrane is placed behind the sternum to facilitate safer sternal opening when the conduit needs changing.

Postoperative Management Patients undergoing the double-switch procedure have low cardiac output postoperatively; this is not surprising with these long and complex procedures. There is potential for compression of the ventricles behind the sternum, particularly in patients with meso or dextro cardiac where the venous structures lie posterior to the ventricular structures on which may be placed a valved conduit behind the sternum. The median times at our institution were 149 minutes for cardiopulmonary bypass, 131 minutes for aortic cross clamp, and 43 minutes for circulatory arrest.10 Inotrope support and ventilatory management are maintained until the cardiac output improves, and the ventricular function is monitored by sequential echo evaluation. At the end of the operation, in the operating room the cardiac function and reconstructed pathways are checked for patency and lack of obstruction by epicardial or transoesophageal echo. Careful echo evaluation of cardiac function is maintained in the ICU postoperatively.

55 When the cardiac function is stabilized, the chest is closed in the ICU, and the patient is weaned from inotropic and ventilatory support. One patient has recently required postoperative ECMO support for 4 days at our institution.

Alternatives to the Double-Switch Procedure In CCTGA with pulmonary stenosis, other suggested methods of repair include a one and a half operation with limited relief of left ventricular outflow tract obstruction and a bidirectional Glenn procedure.15,16 Closure of a VSD with morphologic left ventricle to pulmonary artery limiting valved conduit is placed to maintain morphologic left ventricular pressures, about 50% of the systemic pressures. The morphologic left ventricular pressures are maintained at approximately 50% systemic to maintain alignment of the ventricular septum and to prevent development of tricuspid valve regurgitation and deterioration in morphologic right ventricular function in the longer term. This may be particularly applicable in situations where the atrial switch with the Rastelli procedure is not possible because of the position of the VSD. Finally, a Fontan type procedure may have to be considered if it is not possible by virtue of the position of the heart, location of the VSD, or hypoplasia of the ventricle with or without straddling of the AV valves to be able to perform a septation; however this is an uncommon scenario.

Results of the Double Switch and Recommendations Although the popularity of the double switch has grown over the last 20 years or so, only relatively early results for this type of repair have been published in the literature, particularly from North America, Japan, and Europe.10,17–24 In our group of 54 patients, the early mortality was 5.6%, with two out of three patients who died being in cardiac failure preoperatively. Seven of 54 developed new complete heart block, and there were two late deaths. The survival in the Birmingham, UK series was 94% at 1 year and 90% at 4 to 9 years. Freedom from reoperation of all patients was 94% at 1 year, 85% at 5 years, and 76% at 9 years. There were no reoperations for tricuspid valve regurgitation. Preoperative tricuspid regurgitation, which was often severe, was universally improved by this technique. The reoperations were for the conduit change in the Rastelli group, and this will be a continuing necessity. Six patients had new signs of left ventricular dysfunction (LVD), of whom two had new aortic valve regurgitation in the atrial arterial switch group. Others have reported this problem, although not to this degree. New aortic regurgitation is specific to the atrial arterial switch group where the old pulmonary valve becomes a new aortic valve. All patients in the atrial-arterial switch group had some degree of new aortic valve insufficiency. This was usually mild, but in four patients it was moderately severe, and

W.J. Brawn

56 aortic valve replacement was required in two patients. This has been reported in other series. More recently, we have been concerned about the deterioration in the morphologic left ventricular function in patients who have undergone atrial-arterial switch after training of the morphologic left ventricular with previous pulmonary artery banding. In 44 patients receiving an atrial-arterial switch, 11 patients required pulmonary artery banding to retrain the morphologic left ventricle. There were 32 survivors in the untrained group and 10 survivors in the trained group. Freedom from significant LVD at 2-year follow-up was greater in the untrained group (P ⫽ .01) (Birmingham Children Hospital, Birmingham UK, unpublished data). The algorithm shown in Fig 1 summarizes our thoughts on management for this difficult group of patients. We feel that the morphologic left ventricle replaced in the systemic circulation is of benefit for the majority of the patients with this condition; hence, our preferred management is highlighted with red arrows. A longer follow-up is necessary to show whether problems of new aortic regurgitation and LVD in the atrial-arterial switch group will limit the applicability of this operation to patients with CCTGA. The need for conduit change in the Rastelli group of patients will increase over. There is no doubt, however, that the problem of right ventricular dysfunction and tricuspid regurgitation has been solved by the double switch.

References 1. Ferencz C, Rubin JD, McCarter RJ, et al: Congenital heart disease: Prevalence at livebirth. The Baltimore-Washington Infant Study. Ann J Epidemiol 121:31-36, 1985 2. Becker AE, Anderson RH: Atrioventricular discordance in pathology of congenital heart disease. Butterworth postgraduate series. 1981, pp 225-240. 3. Losekoot TG, Becker AE: Discordant atrioventricular connexion and congenitally corrected transposition, in Anderson RH, MacCartney FJ, Shinebourne EA, Tynan A, (eds): Paediatric Cardiology. Edinburgh, Churchill Livingstone, 1987, pp 867-888 4. Yeh TJ, Connelly MS, Coles JG, et al: Atrioventricular discordance: Results of repair in 127 patients. J Thorac Cardiovasc Surg 117:11901203, 1999 5. Sano T, Riesenfeld T, Karl TR, et al: Intermediate-term outcome after intracardiac repair of associated cardiac defects in patients with atrioventricular and ventriculoarterial discordance. Circulation 92:II272II278,1995 (suppl) 6. Termignon JL, Leca F, Vouhé PR, et al: ‘Classic’ repair of congenitally corrected transposition and ventricular septal defect. Ann Thorac Surg 62:199-206, 1996 7. Van Son JA, Danielson GK, Huhta JC, et al: Late results of systemic

8.

9.

10.

11. 12.

13.

14. 15.

16.

17.

18.

19.

20. 21.

22.

23.

24.

atrioventricular valve replacement in corrected transposition. J Thorac Cardiovasc Surg 109:642-653, 1995 Hraska V, Duncan BW, Mayer JE, et al: Long-term outcome of surgically treated patients with corrected transposition of the great arteries. J Thorac Cardiovasc Surg 129:182-191, 2005 Acar P, Sidi D, Bonnet D, et al: Maintaining tricuspid valve competence in double discordance: A challenge for the paediatric cardiologist. Heart 80:479-483, 1998 Langley SM, Winlaw DS, Stumper O, et al: Midterm results after restoration of the morphologically left ventricle to the systemic circulation in patients with congenitally corrected transposition of the great arteries. J Thorac Cardovasc Surg 125:1229-1241, 2003 Senning A: Surgical correction of transposition of the great vessels. Surgery 45:966-980, 1959 Jonas RA, Mee RB, Sutherland HD: Reintroduction of the Senning operation for transposition of the great arteries. Med J Aust 2:260-262, 1980 De Leval MR, Basto P, Stark J, et al: Surgical technique to reduce the risks of heart block following closure of ventricular septal defect in atrioventricular discordance. J Thorac Cardiovasc Surg 78:515-526, 1979 Shumaker HBI: A new operation for transposition of the great vessels. Surgery 50:773-777, 1961 Mavroudis C, Backer CL, Kohr LM, et al: Bidirectional Glenn shunt in association with congenital heart repairs: The 1 ½ ventricular repair. Ann Thoracic Surg 68:976-981, 1999 Mavroudis C, Backer CL: Physiologic versus anatomic repair of congenitally corrected transposition of the great arteries. Pediatr Card Surg Annu Semin Thorac Cardiovasc Surg 6:16-26, 2003 Ilbawi MN, Deleon SY, Backer CL, et al: An alternative approach to the surgical management of physiologically corrected transposition with ventricular septal defect and pulmonary stenosis or atresia. J Thorac Cardiovasc Surg 100:410-415,1990 Yagihara T, Kishimoto H, Isobe F, et al: Double switch operation in cardiac anomalies with atrioventricular and ventriculoarterial discordance. J Thorac Cardiovasc Surg 107:351-358, 1994 Di Donato RM, Troconis CJ, Marino B, et al: Combined Mustard and Rastelli operations: An alternative approach for repair of associated anomalies in congenitally corrected transposition in situs inversus (IDD). J Thorac Cardiovasc Surg 104:1246-1248, 1992 Imai Y: The Double Switch operation for congenitally corrected transposition. Adv Cardiac Surg 9:65-86, 1997 Reddy VM, McElhinney DB, Silverman NH, et al: The double switch procedure for anatomical repair of congenitally corrected transposition of the great arteries in infants and children. Eur Heart J 18:1470-1477, 1997 Imamura M, Drummond-Webb JJ, Murphy DJ Jr, et al: Results of the Double Switch operations in the current era. Ann Thoracic Surg 70: 100-105, 2000 Karl TR, Weintraub RG, Brizard CP, et al: Senning plus arterial switch operation for discordant (congenitally corrected) transposition. Ann Thoracic Surg 64:495-502, 1997 Devaney EJ, Charpie JR, Ohye RG, et al: Combined arterial switch and Senning operation for congenitally corrected transposition of the great arteries: Patient selection and intermediate results. J Thorac Cardiovasc Surg 125:500-507, 2003