Intermediate-Term Outcome After Intracardiac Repair of Associated Cardiac Defects in Patients With Atrioventricular and Ventriculoarterial Discordance
Background Limited information is available concerning long-term results, especially systemic right ventricular (RV) or tricuspid valvular function, after intracardiac repair of anomalies associated with discordant atrioventricular (AV) and ventriculoarterial (VA) connections (“congenitally corrected transposition of the great arteries”).
Methods and Results We retrospectively reviewed the intermediate-term follow-up of 28 patients, totaling 158 patient-years (median, 60 months), after intracardiac repair involving closure of a ventricular septal defect (VSD) with or without additional surgery. Seven patients had VSD closure alone, 5 had VSD repair with pulmonary stenosis relief, and 16 had VSD closure with conduit insertion between left ventricle and main pulmonary artery. Hospital mortality was 4% (1 of 28 patients; 70% confidence limits, 0.07% to 12%) and the 1-, 5-, and 10-year actuarial survival probabilities were 89%, 83%, and 83%, respectively. Twenty-one of 24 long-term survivors were in New York Heart Association functional class I and 3 were in class II. Sixteen of 24 patients showed increasing tricuspid regurgitation (TR) of more than moderate degree, which occurred within 3 years after surgery in 7 patients. Twelve of 22 patients showed deterioration of RV pump function, mainly (9 of 12 patients) within 3 years postoperatively. The pulmonary to systemic flow ratio at the preoperative cardiac catheter study was significantly (P<.05) higher in patients who developed RV dysfunction (2.3±1.0, mean±SD) than in those with well-maintained RV function (1.4±0.6).
Conclusions Intermediate-term results of intracardiac repair for AV and VA discordance were satisfactory in terms of survival and clinical functional status; however, there is concern about systemic RV dysfunction with development of TR relatively early after the operation. Alternative surgical approaches such as anatomic correction or Fontan repair for cases unsuitable for biventricular repair may improve the long-term results, including ventricular and valvular function.
In patients with AV and VA discordant connection, the anatomic RV maintains the systemic circulation and the tricuspid valve functions as the systemic AV valve. On the other hand, the anatomic left ventricle and mitral valve serve as the pulmonary counterpart. Almost all patients with this cardiac condition have associated cardiac anomalies, including ventricular septal defect, pulmonary stenosis or atresia, Ebstein’s anomaly of the tricuspid valve, and congenital heart block.1 The symptomatic status also varies widely, from predominant cyanosis to severe congestive heart failure, depending on the combination and severity of the associated cardiac defects.2 These symptomatic patients, therefore, may require surgical interventions with palliative procedures or intracardiac repair in early childhood. An intracardiac repair is performed for correction of hemodynamically important coexisting cardiac defects,3 4 5 which in the past usually implied that discordant AV and VA connections were unchanged and that the RV and tricuspid valve still supported the systemic circulation after cardiac surgery.
There has been significant concern regarding long-term prognosis after intracardiac repair for patients with AV and VA discordance. Although there are several studies regarding cardiac function and hemodynamics in this condition, with or without surgical interventions,6 7 8 9 10 the long-term fate of the anatomic RV as the systemic ventricle remains controversial. In the present study, we present the intermediate-term outcome of repair of ventricular septal defect and associated intracardiac defects, especially longitudinal follow-up of the systemic RV and tricuspid valve function, to elucidate risk factors affecting long-term surgical results in this cardiac malformation.
A retrospective analysis was undertaken in 28 patients with AV and VA discordance who underwent intracardiac repair at the Royal Children’s Hospital, Melbourne, Australia, from November 1980 to April 1993. Medical records with reports of surgical operations, cardiac investigations, and postmortem studies were reviewed, and all available cineangiograms and echocardiographic videotapes were reassessed. The diagnosis of AV and VA discordance and associated cardiac anomalies was established by cardiac catheterization, echocardiography, and/or operative findings.
All patients had various associated cardiovascular anomalies, with ventricular septal defect in 28 patients (100%), pulmonary stenosis or atresia in 21 (75%), and tricuspid valve abnormality in 11 (39%) (Table 1⇓). The segmental cardiac arrangement of the 28 patients consisted of situs solitus with AV and VA discordant connection (S,L,L) in 24 and situs inversus with AV and VA discordant connection (I,D,D) in 4 patients.
Assessment of Tricuspid Regurgitation
TR was assessed by color flow mapping, Doppler study, and RV angiography. It was graded qualitatively as no TR, mild TR, moderate TR, or severe regurgitation. Echocardiographically, TR was considered mild if the regurgitant jet crossed less than one third of the left atrium from the tricuspid orifice, moderate if it projected from one third to two thirds of the depth of the left atrium, or severe if it reached beyond two thirds of the left atrium with a significantly wide jet. Angiographically, TR was assessed by use of the criteria for definition of mitral regurgitation described by Grossman11 and was graded as mild (1/4 and 2/4), moderate (3/4), or severe (4/4).
Assessment of RV Pump Function
RV pump function was assessed by echocardiography, cineangiography, and radionuclide angiography. It was evaluated qualitatively from echocardiography and cineangiography and graded as good contraction, mild depression, moderate depression, or severe depression. Quantitative assessment of RV function was obtained from gated equilibrium radionuclide angiographic study in 10 patients, and RV pump function was also graded as good function (RV ejection fraction >50%), mild depression (45% to 50%), moderate depression (35% to 45%) or severe depression (<35%). Grades of good contraction or mild depression were considered adequate RV function in data analysis.
Preoperative and postoperative changes of parameters were tested by use of paired Student’s t test. Variables were compared within subgroups by the unpaired Student’s t test. Risk factor analysis of postoperative RV dysfunction was performed by use of univariate analysis and then by use of multivariate analysis (egret) to variables with a value of P<.10. Time-related survival and freedom from reoperation were estimated using the Kaplan-Meier estimator. Values of P<.05 were considered significant.
Perioperative Clinical Status
Seventeen of 28 patients underwent one or more palliative procedures before the intracardiac repair, including 19 systemic to pulmonary shunts in 13 patients, pulmonary artery band in 2, and repair of coarctation of aorta, reconstruction of pulmonary artery, ligation, or transplantation of major aortopulmonary collateral arteries in 1 each.
Before the intracardiac repair, 21 of 28 patients (75%) were symptomatic, of whom 13 were cyanosed, 5 were in congestive heart failure, and the remaining 3 had a combination of cyanosis and heart failure.
The mean systemic oxygen saturation at the cardiac catheter study before the intracardiac repair measured 89% (SD, 9.2%) and mean pulmonary to systemic flow ratio was 2.0 (SD, 1.0). Preoperative cardiothoracic ratio, hemoglobin concentration, and hematocrit (mean±SD) were 57±6%, 15.8±2.8 g/dL, and 49±9%, respectively; after repair, these clinical parameters were 54±5%, (P<.002), 11.5±0.7 g/dL (P<.001), and 35±2%, (P<.001), respectively.
All 28 patients underwent intracardiac repair, with age at operation ranging from 1 to 13 years (median, 4 years 9 months). Surgical procedures are summarized in Table 2⇓. Intracardiac repair consisted of closure of ventricular septal defect alone in 7 patients, with relief of pulmonary stenosis in 5 and with valved conduit insertion between left ventricle and pulmonary artery in 16 patients. Procedures to relieve pulmonary stenosis consisted of subpulmonary resection in 4 patients, including excision of aneurysmal or accessory endocardial tissue in 3 and pulmonary valvotomy with infundibulectomy in 1 patient. Conduits used to relieve left ventricular outflow obstruction consisted of Hancock valved conduits in 12 patients, Tascon valved conduits in 3, and an aortic homograft in 1 patient. Additional surgical procedures during the intracardiac repair included closure of atrial septal defect in 9 patients and reconstruction of stenotic main pulmonary artery and/or branch pulmonary artery in 7 patients. Two patients with massive tricuspid valve regurgitation preoperatively due to Ebstein’s anomaly or dysplastic valve underwent tricuspid valve replacement with a St Jude valve. Division of an accessory conduction pathway was undertaken in another 2 patients with Wolff-Parkinson-White syndrome and recurrent supraventricular tachycardia resistant to drug therapy.
All but 4 patients had insertion of a permanent pacemaker 3 to 22 days (median, 7.5 days) after the intracardiac repair or simultaneously with the correction in 1 patient. Of 24 patients with implanted pacemakers, 15 implants were prophylactic against the risk of late development of heart block and another 9 were placed for treatment of heart block before or after the repair (5 with congenital and 4 with surgical heart block).
All but one patient survived intracardiac repair (hospital mortality, 4%; 70% confidence limits, 0.07% to 12%). That patient died of massive gastrointestinal bleeding and septicemia 45 days after the operation. There were three late deaths: one was due to acute myocarditis 6 months after the operation; another occurred at 4 months in a child who suffered severe brain damage secondary to postoperative cardiorespiratory arrest; and the third occurred after reoperation to place a pulmonary artery band, 26 months after the intracardiac repair, to “retrain” the left ventricle for double switch conversion. All three patients also had underlying RV dysfunction with development of TR early after surgery.
Follow-up in 24 survivors ranged from 3 to 148 months (median, 70 months). At the last follow-up, 21 patients were in NYHA functional class I and 3 were in class II. The 3 patients in NYHA class II were all long-term survivors (from 112 to 131 months) and were mildly symptomatic, with easy fatigability and dyspnea on exertion. Two of the 3 had recurrent atrial flutter and fibrillation. The 1-month and 1-, 5-, and 10-year actuarial survival after intracardiac repair was 100% (70% confidence limits, 94% to 100%), 89% (78% to 96%), 85% (70% to 94%), and 85% (53% to 99%), respectively (Fig 1⇓).
Excluding pacemaker revisions, seven patients underwent one or more reoperations 3 to 126 months after the initial surgery (Table 2⇑). Two patients underwent pulmonary artery banding to “train” the left ventricle as preparation for double switch conversion because of progressive TR with or without compromised RV pump function during follow-up after intracardiac repair. One of these patients, with severe TR and moderate RV dysfunction, died shortly after the placement of a pulmonary artery band, and the other patient had an ineffective band, despite subsequent tightening of the band. The double switch conversion of Senning operation and arterial switch was successfully performed for one patient with severe RV dysfunction 3 months after the conventional intracardiac repair and tricuspid valve replacement, without prior banding in this case. The 1-month and 1-, 5-, and 10-year freedom from reoperation rates after intracardiac repair were 100% (70% confidence limits, 93% to 100%), 96% (86% to 100%), 80% (62% to 92%), and 72% (36% to 95%), respectively. (Fig 2⇓).
Echocardiographic and/or angiographic assessment of TR was available in 25 patients preoperatively, 21 early (<3 years) postoperatively, and 16 intermediate-term (>3 years) postoperatively. Preoperatively, TR of moderate or severe degree was documented in 9 (7 moderate and 2 severe) of 25 patients (36%). Patients with tricuspid valve abnormalities showed a high incidence of significant TR (6 of 11 patients) compared with patients with normal tricuspid valve (3 of 14 patients). In contrast, TR was demonstrated early postoperatively in 12 of 19 patients (63%), excluding 2 patients who had undergone tricuspid replacement (10 with moderate and 2 with severe TR). Similarly, 10 of 14 patients (71%) had significant TR at late follow-up, 8 (57%) with moderate TR, and 2 (14%) with severe TR. Increases in both frequency and severity of TR were demonstrated in the short-term after intracardiac repair, although neither was significant statistically.
Fig 3⇓ shows the longitudinal change in TR in 24 patients who were assessed serially. Of 24 patients, only 6 (25%), excluding 2 who had tricuspid valve replacement, had competent tricuspid valves during postoperative follow-up. The remaining 16 patients (67%) showed clearly worsening TR of moderate or severe degree. Of these 16 patients, 7 developed TR within 3 years. Three of 5 patients with early deterioration (within 1 year) had late deaths. All patients with abnormal tricuspid valves showed deterioration in tricuspid function, except the 2 who underwent tricuspid valve replacement.
RV Pump Function
Qualitative assessment of pump function of the morphological RV was available in 25 patients preoperatively, 19 early postoperatively (<3 years after intracardiac repair), and 15 at intermediate term (>3 years after repair). Preoperatively, all 25 patients were evaluated as having adequate RV pump function (good or mildly depressed). In contrast, 9 of the 19 (47%) patients evaluated early postoperatively developed RV dysfunction postoperatively (7 moderate and 2 severe) (P<.01 versus before surgery). Similarly, of 14 patients evaluated later postoperatively, 7 (50%) showed RV dysfunction (6 moderate and 1 severe) (P<.01 versus before surgery).
Fig 4⇓ shows the longitudinal change in RV pump function before and after intracardiac repair in 22 patients in whom serial assessment was available. Of these 22 patients, only 10 (45%) retained adequate RV pump function during postoperative follow-up, and the remaining 12 (55%) showed deterioration. Of these 12 patients, 9 showed deterioration in the early postoperative phase (within 3 years). Particularly, 4 patients with significant deterioration within 1 year postoperatively showed poor prognosis, that is, 2 died and 1 required reoperation (double switch conversion).
Several factors were analyzed to compare the 10 patients who retained adequate RV function with the 12 patients who developed RV dysfunction (Table 3⇓). There was no significant difference in operative factors, including operative age, bypass time, or aortic cross-clamp time. There was also no difference in preoperative arterial oxygen saturation, hemoglobin concentration, or cardiothoracic ratio. However, the pulmonary to systemic flow ratio was significantly (P<.05) higher in patients who developed RV dysfunction postoperatively (2.3±1.0) than in those with adequate RV function (1.4±0.6). In addition, logistic regression analysis on postoperative RV dysfunction was performed for variables including atrial situs, absence of pulmonary outflow obstruction, abnormal tricuspid valve, previous palliation, preoperative TR, symptomatic heart failure, pulmonary to systemic flow ratio >2:1, arterial saturation <90%, hemoglobin concentration >16 g/dL, operative age >5 years, conduit repair, additional operative procedure (excluding closure of atrial septal defect), long follow-up period, age >15 years, and postoperative TR. Although no variables were significant predictors of postoperative RV dysfunction, high pulmonary to systemic flow ratio (P=.079 and odds ratio=6.0) and postoperative TR (P=.074, odds ratio=6.3) were relatively correlated to the postoperative RV dysfunction.
Intracardiac repair of associated anomalies in patients with AV and VA discordance has been widely performed.1 In recent years, double switch repair (combined atrial and arterial switch procedure) has become an alternative for some patients with this congenital cardiac anomaly.12 13 14 Repair of intracardiac defects may involve closure of ventricular septal defects and/or relief of pulmonary outflow obstruction because ventricular septal defect and pulmonary stenosis or atresia are common coexisting cardiac defects. In previous reports, early operative results and time-related survival have been less favorable than those after repair of ventricular septal defect, atrioventricular septal defect, tetralogy of Fallot, or transposition of great arteries.2 15 The extensive experience of the University of Alabama showed that 1-month and 1-, 5-, and 10-year survivals after intracardiac repair were 88%, 80%, 76%, and 68%, respectively, with 11% early hospital mortality (70% confidence limit, 7% to 16%).16 Our results have been similar in terms of early operative results and intermediate-term survival, with 4% hospital mortality and actuarial survival of 100% at 1-month, 89% at 1 year, 85% at 5 years, and 85% at 10 years. Lundstrom et al2 suggested that poor preoperative symptomatic status, especially the presence of heart failure, was a risk factor for death within 6 months of operation. In this study, no patient with heart failure died early postoperatively, but there were two late deaths at 7 months and 26 months. The postoperative functional status at intermediate-term follow-up was good, with 88% of patients in NYHA functional class I and the remainder in class II. Thus, the results of intracardiac repair for associated cardiac defects with AV and VA discordance have been generally acceptable in regard to both intermediate-term survival and functional status.
Perioperative or late-developing complete heart block is one of the characteristic problems in patients with AV and VA discordance and is also considered one of the major reasons for lower survival rates in patients with this condition.1 2 It has been suggested that some postoperative patients have died suddenly as a result of the sudden appearance of complete heart block with ventricular asystole or fibrillation.17 Additionally, postoperative tricuspid valve regurgitation may be associated with the development of complete heart block.4 The prevalence of complete heart block after operation has been reported as 15% to 30%.2 4 5 16 In our study, perioperative complete heart block was seen in 9 of 28 patients (32%). However, 5 of these 9 patients had heart block before surgery. Although positive knowledge of the location of the cardiac conduction system in AV and VA discordance18 may help reduce the incidence of complete heart block as a complication of repair,19 the risk of late development of block will never be completely eliminated because pathological change in the conduction system is an acquired progressive condition.20 Our strategy for this difficult problem, therefore, has been to implant permanent pacemakers prophylactically in patients undergoing repair, even if sinus rhythm is preserved. In the present series, no patient had postoperative sudden death.
TR is recognized as another risk factor for poor long-term prognosis in patients with AV and VA discordance.15 It is well known that significant TR develops in many patients with this anomaly both during the natural history and after surgery,2 3 4 15 especially in patients with coexisting tricuspid valve abnormalities.2 In the present study, 60% of postoperative patients displayed significant TR during intermediate-term follow-up, and all patients with abnormal tricuspid valves showed deterioration of TR during postoperative follow-up, except for two who had early valve replacement. Of greater importance, this deterioration of TR was demonstrated during the relatively early postoperative period. Westerman et al4 reported that TR has been apparent immediately after repair in patients without preoperative TR and speculated that it might be associated with the development of complete heart block. However, our study showed that deterioration of TR occurred even in patients without complete heart block. All patients in the present series underwent closure of the ventricular septal defect through the right atrium or via a left ventricular incision for conduit anastomosis. Therefore, operative damage to the tricuspid valve is unlikely to be responsible for the deteriorating TR. A major cause of this phenomenon may be poor tolerance of systemic pressure by the tricuspid valve itself, especially when the valve is dysplastic or abnormal or with the changes in afterload that result from closure of the ventricular septal defect. Additionally, changes in RV geometry, especially dilatation after surgery, may contribute to increased TR.
Huhta et al15 demonstrated that TR altered the long-term outcome of patients with AV and VA discordance and recommended tricuspid valve replacement in patients with hemodynamically significant TR. One patient who underwent tricuspid valve replacement in our series showed good long-term results with satisfactory RV function, although the other patient required reoperation for early progressive RV function (double switch conversion). Our results demonstrate that TR does not improve spontaneously, even when volume overload is eliminated by intracardiac repair.
Long-term RV function has been a major concern in patients with AV and VA discordance or after atrial switch for transposition of great arteries. Although there are several studies regarding RV function in AV and VA discordance, many of them studied patients without associated cardiac anomalies or before surgery, and few studies have been made after intracardiac repair of coexisting cardiac anomalies.6 7 8 9 10 Some reports have shown that the morphological RV can function well over many decades in patients with isolated AV and VA discordance or with minimal associated anomalies.8 9 10 On the other hand, Graham et al7 demonstrated that RV function was maintained well during childhood but that RV dysfunction was more common in adolescence or adult life. Thus, systemic RV function during long-term follow-up remains controversial. Our results showed that RV function was well maintained in all patients before intracardiac repair. However, 55% of patients developed RV dysfunction, even with normalized hemodynamics, after intracardiac repair. This finding is clearly different from previous studies of patients with isolated AV and VA discordance and suggests that patients with AV and VA discordance are at high risk of development of systemic RV dysfunction after intracardiac repair of ventricular septal defect with or without associated left ventricular outflow obstruction.
The cause of this phenomenon is not evident. However, our data suggest that two factors are relevant: preoperative high pulmonary flow and postoperative development of TR. The pulmonary to systemic flow ratio was significantly higher in patients who developed RV dysfunction than in those who retained adequate RV function. Postoperative significant TR was also very frequent, with more than 80% of patients developing RV dysfunction. In contrast, there was no difference in the incidence of preoperative TR between patients who later developed RV dysfunction and those who did not. Decreased RV pump function with significant TR suggests that the systemic RV is functioning with afterload mismatch, that is, elevated afterload due to an inappropriate compensation against the volume overload, with or without impaired myocardial contractility. Sano et al21 reported that compensation for volume overload was not sufficient in the morphological RV, in which afterload mismatch occurred easily, compared with the morphological left ventricle in a study of patients with univentricular heart. In addition, Graham et al22 described the long-standing, volume overload–induced impairment of myocardial contractility in patients with tricuspid atresia long term after systemic to pulmonary shunt. Preoperative excessive volume overload of high pulmonary flow may more easily induce afterload mismatch when the systemic RV is suffering from additional volume overload due to TR postoperatively.
Double Switch Operation
There are now a number of reports of anatomic correction in AV discordance with a combined atrial switch and Rastelli-type repair12 13 14 or combined arterial and atrial switches (so-called “double switch” procedure).14 Imai et al14 reported such corrective procedures in 18 patients with AV discordance and documented satisfactory postoperative biventricular function with significant reduction of TR as well as excellent early operative results. The double switch repair is likely to be a better surgical option than conventional intracardiac repair, at least from the viewpoint of long-term systemic ventricular and AV valvular function, although it is a complex and time-consuming procedure and the long-term results are still unclear. One patient in the present series underwent combined arterial and atrial switch procedure because of severe systemic RV dysfunction shortly after conventional intracardiac repair with tricuspid valve replacement. This patient remains well, with good biventricular and AV valvular function after 5 years. An additional 6 patients have had a combined arterial and atrial switch procedure as primary correction. Follow-up on these patients is, as yet, too brief to allow comparison with the main cohort. Combined arterial and atrial switch should be most appropriate for AV and VA discordance with unrestricted pulmonary flow, not only because the cardiac anatomy is usually suitable but also because such patients appear to be at increased risk of development of RV dysfunction after conventional intracardiac repair. Unfortunately, only 25% of all patients have no pulmonary outflow obstruction in our experience, and hence the majority of patients are not suitable for this form of repair. It is also unclear whether combined atrial switch and Rastelli-type repair is feasible for the majority of patients with AV and VA discordance associated with pulmonary outflow obstruction.
An alternative procedure for discordant AV and VA connection with pulmonary obstruction, especially atresia, and favorable pulmonary circulation is a Fontan operation. This may well be the first choice when additional complex factors such as unbalanced ventricular sizes, straddling AV valve, or abnormal venous return coexist. Another four patients with AV and VA discordance have had Fontan repair with no mortality during the same period. The long-term follow-up on these patients has not been included because of the small numbers.
In conclusion, the intermediate-term outcome of conventional intracardiac repair for patients with AV and VA discordance is acceptable in terms of survival and functional status. However, there is concern about progressive systemic RV dysfunction with development of significant TR after the operation. Preoperative unrestricted pulmonary flow and postoperative development of TR may be risk factors for progressive RV dysfunction after conventional intracardiac repair. Combined arterial and atrial switch procedure or, in complex cases that are unlikely to be amenable to biventricular repair, a Fontan-type repair may produce better long-term results, including better ventricular and AV valvular function, in such patients.
Selected Abbreviations and Acronyms
|NYHA||=||New York Heart Association|
Dr Sano was supported in part by The Uehara Memorial Foundation, Tokyo, Japan.
Presented in part at the 67th Scientific Sessions of the American Heart Association, Dallas, Tex, November 14-17, 1994, and published in abstract form (Circulation. 1994;90[suppl II]:II-97).
- Copyright © 1995 by American Heart Association
Kirklin JW, Barratt-Boyes BG. Congenitally corrected transposition of the great arteries. In: Barratt-Boyes BG, Kirklin JW, eds. Cardiac Surgery. New York, NY: Churchill Livingstone; 1993;2:1511-1533.
Fox LS, Kirklin JW, Pacifico AD, Waldo AL, Bargeron LM. Intracardiac repair of cardiac malformations with atrioventricular discordance. Circulation. 1976;54:123-127.
Westerman GR, Lang P, Castaneda AR, Norwood WI. Corrected transposition and repair of associated intracardiac defects. Circulation. 1982;66(suppl I):I-197-I-202.
Di Donato RM, Wernovsky G, Jonas RA, Mayer JE, Keane JF, Castaneda AR. Corrected transposition in situs inversus: biventricular repair of associated cardiac anomalies. Circulation. 1991;84(suppl III):III-193-III-199.
Parrish MD, Graham TP Jr, Bender HW, Jones JP, Patton J, Partain CL. Radionuclide angiographic evaluation of right and left ventricular function during exercise after repair of transposition of the great arteries: comparison with normal subjects and patients with congenitally corrected transposition. Circulation. 1983;67:178-183.
Grossman W. Profiles in valvular disease. In: Grossman W, Baim DS, eds. Cardiac Catheterization, Angiography and Intervention. Philadelphia, Pa: Lea & Febiger; 1991:557-581.
Ilbawi MN, DeLeon SY, Backer CL, Duffy CE, Muster AJ, Zales VR, Paul MH, Idriss FS. An alternative approach to the surgical management of physiologically corrected transposition with ventricular septal defect and pulmonary stenosis or atresia. J Thorac Cardiovasc Surg. 1990;100:410-415.
Di Donato R, Troconis C, Marino B, Carotti A, Iorio F, Eugenio R, Marcelletti C. Combined Mustard and Rastelli operation: an alternative approach for repair of associated anomalies in congenitally corrected transposition in situs inversus (I.D.D.). J Thorac Cardiovasc Surg. 1992;104:1246-1248.
Metcalfe J, Somerville J. Surgical repair of lesions associated with corrected transposition: late results. Br Heart J. 1983;50:476-482.
Wilkinson JL, Smith A, Lincoln C, Anderson RH. The conducting tissues in congenitally corrected transposition with situs inversus. Br Heart J. 1978;40:41-48.
Sano T, Ogawa M, Taniguchi K, Matsuda H, Nakajima T, Arisawa J, Shimazaki Y, Nakano S, Kawashima Y. Assessment of ventricular contractile state and function in patients with univentricular heart. Circulation. 1989;79:1247-1256.
Graham TP Jr, Franklin RCG, Wyse RKH, Gooch V, Deanfield JE. Left ventricular wall stress and contractile function in childhood: normal value and comparison of Fontan repair versus palliation only in patients with tricuspid atresia. Circulation. 1986;74(suppl I):I-61-I-69.