| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 1999;117:1190-1211
© 1999 Mosby, Inc.
SURGERY FOR CONGENITAL HEART DISEASE |
From the Division of Cardiovascular Surgery,a Division of Cardiology,b Hospital for Sick Children, Toronto Congenital Cardiac Centre for Adults,c University of Toronto, Toronto, Ontario, Canada, and Division of Cardiothoracic Surgery, Kosair Children's Hospital,d Department of Mathematics,e University of Louisville, Jewish Hospital Heart & Lung Institute, Louisville, Ky.
Received for publication June 23, 1998. Revisions requested Dec 10, 1998. Revisions received Feb 16, 1999. Accepted for publication Feb 26, 1999. Address for reprints: Thomas Yeh, Jr, MD, PhD, University of Louisville, Jewish Hospital Heart & Lung Institute, 201 Abraham Flexner Way, Suite 1200, Louisville, KY 40202.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussion Conclusions References |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussion Conclusions References |
|---|
1 and performance 2 of the double switch procedure in which AV and VA connections are "corrected," allowing the left ventricle and mitral valve to support the systemic circulation. A 40-year experience treating 127 patients with AV discordance by biventricular repair was reviewed for insight into the management of this lesion. |
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussion Conclusions References |
|---|
Anatomic variations
Situs solitus was present in 93% of the cohort and situs inversus in 7%. The heart was levo-positioned in 82% of the cohort, dextro-positioned in 13%, and meso-positioned in 5%. The most common VA connection was discordant in 87% of patients. Double-outlet right ventricle (6%) and VA concordance (6%) comprised equal but smaller percentages of the cohort. A single case of double-outlet left ventricle was present.
The incidence of the 3 most common lesions associated with AV discordance managed with biventricular repair varied (Fig. 1). VSD was present in 86% of the cohort, PS in 64%, and TI in 28%; however, patients had varying combinations of associated lesions. The most common of these was VSD combined with PS, accounting for 72 patients (56% of the cohort). Four other combinations accounted for smaller proportions of the cohort, namely VSD/TI (14 patients, 11%), isolated VSD (17 patients, 14%), isolated TI (13 patients, 10%), and the combination of VSD/PS/TI (7 patients, 6%). Notably, isolated PS (2 patients, 2%) and PS/TI (1 patient, 1%) formed extremely small percentages of the cohort.
|
1 In the 95 patients undergoing VSD closure with a conventional repair, the approach to the VSD varied. Thirty-nine VSDs were closed through the left ventricle, 5 through the right ventricle, 43 through the right atrium, 3 through the left atrium, and 5 through the aorta. At the time of initial open repair, 27 patients with incompetent tricuspid valves were managed initially either by replacement (with a mechanical valve [19 patients, 70%] or a bioprosthetic valve [3 patients, 11%]) or by repair (5 patients, 19%). The management of 71 patients with PS used conduits (63 patients, 89%), pulmonary valve and subvalvular resection (7 patients, 10%), or REV (réparation à l'étage ventriculaire, 1 patient, 1%). The REV was accomplished by exposing the morphologic left coronary artery in the AV groove to facilitate direct connection of the mobilized and divided main pulmonary artery on the shoulder of the heart so as to minimize anterior compression of the right ventricular outflow tract by the sternum.
In the double switch (ie, Ilbawi) procedure to correct VA discordance, 2 the left ventricular outflow was redirected to the aorta and the right ventricle was connected to the pulmonary artery either by an arterial switch (n = 1) or by a Rastelli procedure (n = 8). To correct AV discordance, an atrial switch (Mustard procedure) was added. One additional patient had AV discordance (with VSD) and was managed with a VSD repair and Mustard procedure to correct her isolated ventricular inversion.
Age at repair
Age at open biventricular repair ranged from newborn to 65 years (Fig. 2A). The median age in the entire cohort at initial repair was 8 years, but age at repair varied by diagnosis (Fig. 2B). For instance, patients undergoing biventricular repair for AV discordance who had isolated VSDs tended to undergo repair earliest at a median age of 3.5 years (range 0.1-12.8 years), whereas those with tricuspid valve regurgitation underwent valve repair or replacement at a median age of 16.6 years (range 0.6-65 years).
|
With regard to reoperation (including tricuspid valve surgery, pacemaker placement, and conduit replacement), in addition to analyzing freedom from each of these events individually, also included is an analysis of freedom from the event or freedom from death. Since death precluded patients from reoperation, omitting death as part of the censoring process for the event "reoperation" made freedom from reoperation appear better than it actually was.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussion Conclusions References |
|---|
). From date of birth, survival declined progressively over time and was 50% at age 41 years (Fig. 3B). To control for technologic advances in cardiac surgery, survival was compared between the first and second halves of the cohort, but it did not significantly vary between eras (P = .24).
|
|
|
Survival by type of repair
Survival revealed no significant differences among the 4 major diagnostic groups in the cohort when analyzed from date of initial repair (Fig. 3C
, P = .36). Although survival in patients with isolated VSDs appeared to be better, this trend was not statistically significant (P = .86). Whether or not the discordance was corrected by means of the double switch procedure also did not affect survival; however, maximum follow-up of the 9 patients undergoing a double switch procedure is limited to 8.6 years, at which time patient survival is 89%.
Reoperation
The causes of late death are enumerated in Table II. Death at reoperation (usually from myocardial failure) accounted for 36% of late deaths. Sudden unexplained death, documented arrhythmia, and progressive myocardial failure comprised the remainder. Forty-four patients required 1 or more reoperations. The risk of death at reoperation increased exponentially with each procedure (Table I).
|
).
|
|
Tricuspid valve failure
Tricuspid valve failure was an important cause of morbidity and mortality, both at initial repair and during subsequent follow-up. Forty-two of 127 patients (34%) have required tricuspid valve repair or replacement to date. Twenty-seven of these required tricuspid valve replacement or repair at initial repair. Of these 27, 4 have required a subsequent operation for tricuspid valve failure at a mean age of 10 years (median 14.2 years) after the initial repair.
Of the 98 patients who did not require tricuspid valve surgery at initial repair, 14 (14%) have required subsequent tricuspid valve surgery by a mean age of 19.9 years (median 17.4 years). Although some of these patients had Ebsteinoid anomaly of the tricuspid valve, the exact tricuspid valve disease is unknown in our cohort for several reasons. The defect in the tricuspid valve was often subtle; some patients underwent surgery before the advent of echocardiography, and the left-sided AV valve was difficult to visualize at surgery.
Fifty percent of patients with AV discordance required tricuspid valve surgery by age 35 years (Fig. 5). Tricuspid valve failure requiring surgery increased with age. When we compared patients who underwent tricuspid valve surgery at initial repair versus those who required tricuspid valve surgery after their initial repair, no difference in actuarial survival was noted.
|
|
3 one might expect that pacemaker placement would confer a survival advantage; however, patients who required pacemakers had no better (or worse) survival than those who did not. In our early experience with the double switch procedure, 5 of 9 required perioperative pacemaker placement for complete AV block.
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussion Conclusions References |
|---|
3 Our outcomes are probably explained by the theories that have been advanced to explain the natural history of discordance, that is, (1) failure of the morphologic right ventricle in the systemic position, (2) failure of the morphologic tricuspid valve in the systemic position, and (3) failure of the conduction system. Indeed, these problems are interrelated and each has been postulated to result in the other two. Which was primarily responsible was impossible to discern retrospectively.
Controversies in the pathophysiology of AV discordance
In some patients, repair of VSD with or without left ventricular outflow tract obstruction may have led to progressive systemic AV valve regurgitation. 4 Because function of the morphologic right ventricle was often depressed, discerning whether valve regurgitation reflects a failing right ventricle and annular dilatation or whether severe AV valve regurgitation has negatively affected ventricular function was problematic. In some older patients with AV and VA discordance "in isolation," right ventricular function is seemingly preserved. A number of reports of adults aged 60 to 80 years with AV discordance suggest that the morphologic right ventricle may function successfully as the systemic ventricle for a normal life span in patients with no major associated lesions. 5-9 Radionuclide assessment of ventricular function supports this as well. 10 In contrast, data from Peterson and associates 11 suggested that patients with AV discordance have an abnormal exercise response in that they failed to increase their systemic ejection fraction with exercise. Regardless of the ongoing debate, multiple centers have reported disappointing 10-year survival using conventional management (Table VII). 1218 In our series, survival 20 years after the initial repair was only 48%.
|
) and mortality (Table VII). The atrial switch procedures (Mustard, Senning) 19 and procedures designed to correct VA discordance (Jatene arterial switch, Rastelli procedure) 20-22 each have their own shortcomings.
In analyzing outcomes (Table VI), surgical correction of discordant VA connections resulted in superior preservation of systemic (left) ventricular function and systemic (mitral) valve competence. Furthermore, avoidance of atrial switch procedures protected atrial conduction. Although survival was not improved in simple transposition of the great arteries (Table VI), in the treatment of complex transposition (typically associated with VSD), correcting VA connections improved 10-year survival to 77% (versus 63% obtained using the Mustard procedure). 23 Teleologically, this improvement supports the theory that correcting discordant VA connections relieves the morphologic right ventricle and the tricuspid valve of the need to support systemic hemodynamic loads and is more important in patients who have a major associated lesion than in those with isolated complete transposition.
|
Unfortunately, patients with AV discordance almost always have associated lesions; therefore, AV discordance is conceptually similar to complex transposition of the great arteries (transposition and VSD). The survival of patients with AV discordance managed conventionally is remarkably similar to that of complex VA discordance managed with an atrial switch operation (therefore without correcting VA connections, Table VI). 23 Inferentially, the residual pathophysiologic problem in both VA discordance after atrial switch and AV-VA discordance managed conventionally without correcting discordant connections is that the morphologic right ventricle and tricuspid valve are exposed to systemic hemodynamic loads.
The double switch procedure: Panacea or Pandora's box?
In 1988, to address this concern, de Leval and associates 1 first theorized the double switch procedure. Ilbawi and colleagues, 2 in 1990, were the first to report successful cases (n = 2). The results of all published series in which this approach was used 2,24-28 are summarized in Table VII and compared with other published series using conventional correction of associated lesions. 12-18 Operative mortality was similar to that of conventional repair, but long-term follow-up was insufficient to conclude an overall superiority for this procedure. Our data for conventional repair (74% survival at 10 years and 48% survival 20 years after repair) suggest that differences in survival may not become evident until 10 years of follow-up have elapsed.
Because AV discordance is uncommon, published series are relatively small. Small numbers of patients combine with institutional biases (including our own) to make objective demonstrations of superiority (by randomized controlled trial) extremely unlikely. The double switch operation combines 2 operations: an atrial switch procedure (Mustard or Senning) and a procedure to correct VA discordance (either a Jatene arterial switch in the absence of pulmonary stenosis or a Rastelli-like procedure in the presence of VSD and PS). After a double switch procedure, Imai and coworkers 25 reported a decrease in right ventricular end-diastolic volume from 122% to 78% with unloading of the right ventricle and no significant change in left ventricular end-diastolic volume before (134% of normal) and after (127% of normal) the procedure. The early development of 3 cases of left ventricular outflow tract obstruction in our series after Mustard/Rastelli repair was disconcerting and argues for careful and ongoing follow-up of patients managed by this algorithm. The Rastelli procedure is most easily accomplished when the VSD is anterior, muscular, subaortic, and also free of conduction tissue. The more typical inlet perimembranous VSD presents more of a challenge. Furthermore, in our hands, the incidence of acquired complete AV block was high. Others have avoided AV block (Table VII), and we believe that sinus rhythm can be preserved by avoiding technical errors in tunnel construction. Paradoxically, the high incidence of block in corrected transposition may be protective against the supraventricular tachycardias incurred after atrial switch procedures. Finally, in cases in which a pulmonary conduit was used, and possibly even in cases in which an REV procedure was employed, 29 the morbidity arising from conduit stenosis will be formidable.
Results with conventional management of AV discordance were disappointing. Elimination of pulmonary conduits by using an REV procedure, advancements in pulmonary conduit preservation and longevity, and elucidation of differences between surgical techniques (or patient substrate) allowing more successful avoidance of complete AV block may diminish risks of the double switch approach.
Anecdotally, the single atrial switch/arterial switch in our series had an excellent result. Patients amenable to an arterial switch as part of a double switch repair will not have PS and are unlikely to have a restrictive VSD. It is probable that an atrial switch/Rastelli combination (used for the diagnosis combination of VSD and PS) is associated with significantly higher morbidity and mortality than the atrial switch/arterial switch combination (typically used for the diagnosis of VSD or tricuspid regurgitation). The subgroup eligible for the atrial switch/arterial switch combination are ideal double switch candidates and will avoid the problems of the Rastelli procedure (intraventricular tunnel stenoses and pulmonary conduit failure), as well as the problems of AV/VA discordance (tricuspid valve incompetence and right ventricular dysfunction).
For completeness, 2 alternatives to conventional management or double switch procedures should also be considered. Many patients with AV discordance and PS remain free of symptoms for years. Some benefit from Helen Taussig's philosophy of placing a systemic–pulmonary artery shunt as long-term definitive palliation. This approach was not examined in our cohort; however, these data serve as the baseline for future comparisons. A second alternative would be converting these patients to a univentricular circulation, avoiding the morbidity and mortality associated with iatrogenic AV block and pulmonary conduit failure. Twenty-year survival after Fontan's procedure for patients with a single ventricle at our institution was 50% to 60%, but a univentricular conversion in a patient with 2 adequate ventricles may not provide superior survival or quality of life.
The theoretic advantage of correcting discordant AV connections must outweigh the complications of the component procedures comprising a double switch. Inasmuch as early mortality using the double switch procedure was no higher than that obtained with conventional management of AV discordance (Table VII), and since 20-year survival of conventionally managed patients who did not undergo correction of VA discordance was only 48% (Table VI), continued application of the double switch procedure seems reasonable in the expectation that correcting VA discordance will provide a survival advantage as it did in complex transposition. The patients in whom an arterial switch can be used are ideal candidates for the double switch. Perhaps a randomized controlled trial would most rapidly identify any advantage (or danger) of using this approach.
|
Conclusions |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussion Conclusions References |
|---|
Twenty years after initial repair, only 48% of the cohort was surviving (average patient age 42 years). Within 16 years, 57% of patients required reoperation and 35% required pacemaker placement. Within 12 years of conduit placement, 50% of patients required conduit replacement. By age 40, 47% of patients had required tricuspid valve surgery. Early results with the double switch operation were encouraging, with equivalent early mortality to conventional management; however, its potential advantages will require long-term follow-up.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussion Conclusions References |
|---|
This article has been cited by other articles:
|
|
 |
|
  T. Shin'oka, H. Kurosawa, Y. Imai, M. Aoki, M. Ishiyama, T. Sakamoto, S. Miyamoto, K. Hobo, and Y. Ichihara Outcomes of definitive surgical repair for congenitally corrected transposition of the great arteries or double outlet right ventricle with discordant atrioventricular connections: Risk analyses in 189 patients J. Thorac. Cardiovasc. Surg., May 1, 2007; 133(5): 1318 - 1328. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Bautista-Hernandez, G. R. Marx, K. Gauvreau, J. E. Mayer Jr, F. Cecchin, and P. J. del Nido Determinants of Left Ventricular Dysfunction After Anatomic Repair of Congenitally Corrected Transposition of the Great Arteries Ann. Thorac. Surg., December 1, 2006; 82(6): 2059 - 2066. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. A. Alghamdi, B. W. McCrindle, and G. S. Van Arsdell Physiologic Versus Anatomic Repair of Congenitally Corrected Transposition of the Great Arteries: Meta-Analysis of Individual Patient Data Ann. Thorac. Surg., April 1, 2006; 81(4): 1529 - 1535. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. A. Russell, K. Rouine-Rapp, G. Stratmann, and W. C. Miller-Hance Congenital heart disease in the adult: a review with internet-accessible transesophageal echocardiographic images. Anesth. Analg., March 1, 2006; 102(3): 694 - 723. [Full Text] [PDF]
|
|
|
|
|
|
M. Koh, T. Yagihara, H. Uemura, K. Kagisaki, I. Hagino, T. Ishizaka, and S. Kitamura Intermediate Results of the Double-Switch Operations for Atrioventricular Discordance Ann. Thorac. Surg., February 1, 2006; 81(2): 671 - 677. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Koh, T. Yagihara, H. Uemura, K. Kagisaki, and S. Kitamura Functional biventricular repair using left ventricle-pulmonary artery conduit in patients with discordant atrioventricular connections and pulmonary outflow tract obstruction--does conduit obstruction maintain tricuspid valve function? Eur. J. Cardiothorac. Surg., October 1, 2004; 26(4): 767 - 772. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Heggie, N. Poirer, W. G. Williams, and J. Karski Anesthetic Considerations for Adult Cardiac Surgery Patients with Congenital Heart Disease Seminars in Cardiothoracic and Vascular Anesthesia, June 1, 2003; 7(2): 141 - 152. [Abstract] [PDF]
|
|
|
|
|
|
S. M. Langley, D. S. Winlaw, O. Stumper, R. Dhillon, J. V. de Giovanni, J. G. Wright, P. Miller, B. Sethia, D. J. Barron, and W. J. Brawn 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. Cardiovasc. Surg., June 1, 2003; 125(6): 1229 - 1241. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. M. Beauchesne, C. A. Warnes, H. M. Connolly, N. M. Ammash, A. J. Tajik, and G. K. Danielson Outcome of the unoperated adult who presents with congenitallycorrected transposition of the great arteries J. Am. Coll. Cardiol., July 17, 2002; 40(2): 285 - 290. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Dodge-Khatami, I. I. Tulevski, G. B.W.E. Bennink, J. F. Hitchcock, B. A.J.M. de Mol, E. E. van der Wall, and B. J.M. Mulder Comparable systemic ventricular function in healthy adults and patients with unoperated congenitally corrected transposition using MRI dobutamine stress testing Ann. Thorac. Surg., June 1, 2002; 73(6): 1759 - 1764. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Jahangiri, A. N. Redington, M. J. Elliott, J. Stark, V. T. Tsang, and M. R. de Leval A case for anatomic correction in atrioventricular discordance? Effects of surgery on tricuspid valve function J. Thorac. Cardiovasc. Surg., June 1, 2001; 121(6): 1040 - 1045. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
