|
|
||||||||
Ann Thorac Surg 1998;66:842-848
© 1998 The Society of Thoracic Surgeons
a Cardiac Surgical Unit, Royal Childrens Hospital, Melbourne, Australia
b Department of Cardiology, Royal Childrens Hospital, Melbourne, Australia
Address reprint requests to Dr Karl, Cardiac Surgical Unit, Royal Childrens Hospital, Flemington Rd, Melbourne, Australia 3052
Presented at the Thirty-fourth Annual Meeting of The Society of Thoracic Surgeons, New Orleans, LA, Jan 2628, 1998.
| Abstract |
|---|
|
|
|---|
Methods. In this study we analyze long-term outcome for 26 such transposition patients (6.3% of our ASO cohort). Median age and weight were 69 days (7 to 3,631 days) and 4.5 kg (2.6 to 34 kg). Pulmonary valve abnormalities included bicuspid valve (n = 4) and dysplastic valve (n = 5). The LVOT abnormalities (n = 17) included accessory atrioventricular valve/endocardial cushion tissue, fibromuscular ring, anomalous muscle bands, and septal malalignment. Patients with dynamic LVOT obstruction were excluded. The median preoperative left ventricular to pulmonary artery peak systolic pressure gradient was 30 mm (0 to 93 mm), or 50 mm (16 to 93 mm) if patients with isolated valve abnormalities are excluded. The ASO was performed according to our standard technique with or without LVOT resection or pulmonary valvotomy as required.
Results. There were two perioperative deaths (7.7%; 95% confidence interval, 0.9% to 25%), and no late deaths during 1,934 patient-months of follow-up time. Actuarial freedom from reoperation for neoaortic valve or LVOT problems is 87% (± 7) at 130 months, representing two reoperations. One was performed for neoaortic insufficiency plus LVOT obstruction, and the other for isolated LVOT obstruction. One patient currently has significant neoaortic insufficiency, and median gradient at last follow-up is 0 mm Hg (range, 0 to 35 mm Hg).
Conclusions. The ASO can be performed in selected patients with transposition of the great arteries and with LVOT abnormalities with early and late survival and functional status similar to that of matched patients with normal pulmonary valves and LVOT (p > 0.05), but with a greater hazard for reoperation (p < 0.05). Selection for ASO should be based on anatomic criteria rather than left ventricular to pulmonary artery gradient alone, to avoid assigning these patients with transposition of the great arteries to treatment strategies less satisfactory than ASO.Transposition, arterial switch
| Introduction |
|---|
|
|
|---|
|
|
| Patients and methods |
|---|
|
|
|---|
Twenty-six of 413 children (6.3%) undergoing ASO for TGA in our institution had anatomic abnormalities of the LVOT (including pulmonary valve or subvalvar region, or both). Twenty of the 26 patients had a VSD. Median age and weight at the time of operation were 69 days (7 to 3,631 days) and 4.5 kg (2.6 to 34 kg). Twenty of the 26 patients had evidence of a significant left ventricle to pulmonary artery gradient on preoperative studies, and the remainder had isolated pulmonary valve abnormalities that did not generate a gradient (Table 2 ). The causes of LVOT obstruction in our own patients included accessory tissue attached to an atrioventricular valve, other types of accessory endocardial cushion tissue, subpulmonary fibromuscular tissue (discrete or tunnel), anomalous muscle bars, septal malalignment, and valvar pulmonary stenosis (see Table 2). Nonobstructive and obstructive abnormalities observed either in pulmonary valves (in isolation or with other lesions) included bicuspid valve (n = 5) and dysplastic, thickened, or obviously asymmetric leaflets (n = 7). The aortic Z scores for annular size in isolated pulmonary valve abnormalities ranged from -3.16 to +3.79 (median, 1.4). The median peak systolic gradient from left ventricle to pulmonary artery was 30 mm Hg (range, 7 to 93 mm Hg) (Fig 2 ; Table 2), or 50 mm Hg (range, 16 to 93 mm Hg) if patients with isolated pulmonary valve abnormalities were excluded (p = 0.001).Coronary artery patterns other than 1LCx, 2R were noted in 15 of 26 patients, a proportion similar to other patients undergoing ASO in our unit (p = 0.09). In the presence of LVOT abnormalities, ASO was performed when we believed that an unobstructed (or minimally obstructed) left ventricle to neoaortic connection could be established by resection, without injury to surrounding structures, and that pulmonary valve competence could be maintained. The final judgment was made intraoperatively. Other surgical options chosen during this time period to deal with the LVOT obstruction included Rastelli, REV (réparation à létage ventriculaire), or Senning techniques (n = 28).
|
|
| Results |
|---|
|
|
|---|
Total follow-up time is 1,934 patient-months, and there have been no late deaths. Overall actuarial survival probability was 92% ± 5% at 130 months. For patients who had a significant left ventricle to pulmonary artery gradient detected preoperatively, survival was 94.7% ± 5% at 130 months (p = 0.87). Late reoperations were required for neoaortic insufficiency in 1 patient, whose initial procedure had been ASO, VSD closure, and resection of accessory tricuspid tissue. The pulmonary valve was judged normal at the time, but more than 9 years later, progressive neoaortic dilation and aortic insufficiency developed (Fig 3 ), as well as a discrete fibrous membrane in the subvalvar LVOT. He underwent aortic valve replacement, aortic root roof reconstruction, and subvalvar resection. A second patient had reoperation to repair supravalvar pulmonary stenosis, and a third to relieve progressive LVOT obstruction due to septal malalignment. All patients survived reoperation. Actuarial freedom from reoperation for all patients was 72.3% ± 9% at 130 months. For patients who had significant left ventricle to pulmonary arterial gradient at first operation, this probability was 73.3% ± 10% at 130 months (p = 0.72). Also, if one considers only reoperations for relief of LVOT obstruction or pulmonary valve revision, the freedom from reoperation is 87.2% ± 7% at 130 months. All survivors are currently in New York Heart Association class I, and median gradient at latest follow-up (mean interval, 81 months) was 0 mm Hg (range, 0 to 35 mm Hg) (Fig 2). The only patient who had a pulmonary valvotomy currently has moderate neoaortic insufficiency but no gradient, and will need reoperation in the future. No other patient currently has more than mild neoaortic insufficiency (although as mentioned above, 1 child has had an aortic valve replacement). No patients with isolated pulmonary valve abnormalities have required reoperation for pulmonary valve or LVOT problems.
|
| Comment |
|---|
|
|
|---|
The presence of LVOT obstruction has important implications for the timing and technique of TGA repair. If significant obstruction is present at birth, survival after balloon septostomy may be favorably affected for the first 3 to 6 months in uncorrected patients. Our patients with LVOT obstruction have undergone ASO at a somewhat older age than those without LVOT obstruction. However, most types of obstruction tend to increase in severity with time, ultimately adversely affecting survival due to reduction in pulmonary blood flow and poor intracardiac mixing [5, 6].
Preoperative diagnosis can be made accurately with echocardiography in the majority of patients, but also with cardiac catheterization and magnetic resonance imaging studies [7]. The anatomic details of the obstruction may be more important than the gradient per se, which could be overestimated due to the high pulmonary blood flow typical of TGA (especially if a VSD is present). It is important to establish these anatomic details in the beating heart, as they may be obscured intraoperatively after administration of cardioplegia.
Pulmonary valve in transposition of the great arteries
The pulmonary valve in many TGA hearts may have unequal cusp sizes, leading to eccentric closure [8]. The bicuspid valve is perhaps a rare and extreme example, and was found in only 1.2% of our ASO patients. Cusp abnormalities may have significance for postoperative valve function, and may contribute to neoaortic insufficiency when the valve functions at systemic pressure.
Pulmonary valve stenosis as an isolated entity is rare in TGA, and tends to occur more frequently with subvalvar stenosis of various types. As in concordant hearts, this association usually constitutes a more complex form of obstruction, and the valve tends to be normal in patients with only a subvalvar membrane. Limited data are available regarding the results of ASO in patients with bicuspid or otherwise abnormal pulmonary valves [9]. In our own experience the function of neoaortic bicuspid valves appears identical to that of trileaflet valves after ASO, although the number of patients in the former group is very small.
Subvalvar left ventricular outflow tract in transposition of the great arteries
Although most hearts with TGA have some degree of pulmonary-to-mitral continuity, a small proportion have a subpulmonary muscular infundibulum, usually associated with a VSD. The infundibulum may be obstructive, often in conjunction with a small pulmonary valve ring. When a VSD is present, the outlet septum may be displaced toward the left in relation to the infundibular septum, thus reducing the caliber of the LVOT. Isolated subvalvar membranes are also seen in TGA, and appear quite similar to those seen in concordant hearts, in which the corresponding semilunar valve is usually normal (Fig 4 ). It appears that in TGA, as in concordant hearts, resection or enucleation usually restores a normal caliber LVOT, with good long-term outcome.
|
|
|
|
Operation for transposition of the great arteries + left ventricular outflow tract obstruction
A number of surgical alternatives are available for TGA with LVOT obstruction, with and without VSD (Fig 1). The presence of LVOT obstruction in TGA has led some groups to use a Mustard or Senning repair, with or without a specific procedure to relieve the LVOT obstruction [2, 3, 18]. Daicoff and colleagues [19] reported in 1969 direct LVOT obstruction resection combined with Mustard operation. In the same year, Rastelli and colleagues [6, 20] reported the use of left ventricle to pulmonary artery conduits combined with intraventricular baffle for repair of TGA with unresectable LVOT obstruction and VSD. This operation, along with the more recent REV, remains a good option for certain anatomic situations in which LVOT obstruction cannot be directly relieved, such as fibromuscular tunnel and hypoplastic pulmonary annulus. All patients operated on with the Rastelli or REV technique had complex multilevel LVOT obstruction, with a hypoplastic, unusable pulmonary valve.
Initial results with atrial-level repairs in the presence of LVOT obstruction have been satisfactory, but patients are still subject to late problems associated with a discordant atrioventricular and ventriculoarterial connection, as well as to progression of the LVOT gradient. We would consider ASO (with or without resection in the LVOT) to be a better option in technically suitable patients. The advantages of this strategy, as compared with others listed, include restoration of concordant ventriculoarterial connection, minimal prosthetic material load, and avoidance of extracardiac conduit placement. Although our own approach has certainly not been free of morbidity and mortality, we believe that it would compare favorably with other procedures used in a similar patient cohort. The difficulty for the surgeon lies in deciding when resection will be likely to result in an unobstructed left ventricle to neoaortic connection in the long term, without inducing neoaortic insufficiency or other problems. This is a critical decision in neonates and infants who may not tolerate a residual gradient after ASO, and who may also be poor candidates for other types of repairs (eg, REV and Rastelli). On the other hand, a significant degree of obstruction may allow one to defer the ASO, providing hemoglobin saturation remains adequate.
Wernovsky and associates [1] suggested that in various anatomic forms of LVOT obstruction, resection and ASO could lead to sustained relief in most patients, excluding those with posterior deviation of the infundibular septum or those with a combination of causes that included septal deviation. As in our own series, only 1 of 10 patients with an abnormal pulmonary valve required valvotomy. Only 2 of our patients to date have had significant progression of subvalvar LVOT obstruction after ASO. One had posterior deviation of the infundibular septum, and the other had accessory tricuspid tissue and later development of a subvalvar discrete membrane. Both patients have had successful reoperation, as noted above.
Our results for ASO in the presence of LVOT abnormalities would suggest that operative and long-term mortality risk are not significantly higher than that of other ASO patients. Our mortality in the same era for ASO in patients with TGA with VSD or intact septum (without LVOT abnormalities) was 3.8% (95% confidence interval, 1% to 9.9%) and 0.9% (95% confidence interval, 0% to 3%), respectively. This was not significantly different from patients with LVOT abnormalities adjusted for intracardiac features (p = 1.00). The probability of reoperation is higher if all reoperations are considered, but probably not if we consider only procedures for revision of the LVOT or neoaortic valve (p > 0.05). As in concordant hearts, valvotomy is likely to increase the risk of late insufficiency and need for reoperation. Relief of LVOT obstruction appears to be sustained at medium-term follow-up. Nonobstructive valve abnormalities generally have not tended to progress to cause valve dysfunction in the medium term, but longer follow-up time will be necessary to resolve this issue. Hemodynamically insignificant neoaortic insufficiency is detectable in many patients after ASO [21]. This has also been the case in our patients with LVOT abnormalities. Our own unpublished data suggest that the neoaortic annulus after ASO in most patients follows the growth pattern for a normal pulmonary valve, rather than exhibiting a trend toward ectasia. It is possible that distortion of the annulus during coronary translocation, or structural leaflet abnormalities are the important factors in the long term. Whether or not a structurally abnormal valve will function at systemic pressure in the very long term remains unknown, but results to date look favorable for most patients.
In conclusion, the arterial switch operation can be performed in selected patients with TGA and LVOT abnormalities with early and late survival similar to that of other patients in our own ASO cohort. The reoperation probability may be higher, however, in patients who have anatomic LVOT abnormalities at the time of ASO (p < 0.05). The preoperative left ventricle to pulmonary artery gradient is not the best predictor of resectability. Coronary arteries, atrioventricular valves, and conduction tissue are all susceptible to injury during resection using the transpulmonary artery and transatrial approach. Most patients can expect good relief of LVOT obstruction and good neoaortic valve function in the medium term. Selection criteria for ASO in the presence of LVOT abnormalities should be based primarily on anatomic considerations, rather than gradient alone, the best candidates being those with isolated membranes, accessory atrioventricular valve leaflet tissue, and nonobstructive pulmonary valve abnormalities. Our findings support those of Yacoub, Uemura, and Wernovsky and their colleagues [1, 4, 9] in earlier reports, and suggest that LVOT abnormalities are not necessarily a contraindication to safe ASO.
| Acknowledgments |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
S. M. Emani, R. Beroukhim, D. Zurakowski, F. A. Pigula, J. E. Mayer, P. J. del Nido, T. Geva, and E. A. Bacha Outcomes After Anatomic Repair for D-Transposition of the Great Arteries With Left Ventricular Outflow Tract Obstruction Circulation, September 15, 2009; 120(11_suppl_1): S53 - S58. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Leobon, E. Belli, M. Ly, C. Kortas, E. Le Bret, A. Sigal-Cinqualbre, R. Roussin, and A. Serraf Left ventricular outflow tract obstruction after arterial switch operation Eur. J. Cardiothorac. Surg., November 1, 2008; 34(5): 1046 - 1050. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. M. Khan, A. B. Sallehuddin, Z. R. Al-Bulbul, and Z. Y. Al-Halees Neoaortic Bicuspid Valve in Arterial Switch Operation: Mid-Term Follow-Up Ann. Thorac. Surg., January 1, 2008; 85(1): 179 - 184. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Hazekamp Invited commentary Ann. Thorac. Surg., January 1, 2008; 85(1): 184 - 185. [Full Text] [PDF] |
||||
![]() |
M. Hazekamp, F. Portela, and M. Bartelings The optimal procedure for the great arteries and left ventricular outflow tract obstruction.: An anatomical study Eur. J. Cardiothorac. Surg., May 1, 2007; 31(5): 879 - 887. [Abstract] [Full Text] [PDF] |
||||
![]() |
V. Bautista-Hernandez, G. R. Marx, E. A. Bacha, and P. J. del Nido Aortic Root Translocation Plus Arterial Switch for Transposition of the Great Arteries With Left Ventricular Outflow Tract Obstruction: Intermediate-Term Results J. Am. Coll. Cardiol., January 30, 2007; 49(4): 485 - 490. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Sharma, S. K. Choudhary, A. Bhan, R. P. Kumar, R. Juneja, S. S. Kothari, A. Saxena, and P. Venugopal Late outcome after arterial switch operation for complete transposition of great arteries with left ventricular outflow tract obstruction Ann. Thorac. Surg., December 1, 2002; 74(6): 1986 - 1991. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Kreutzer, J. De Vive, G. Oppido, J. Kreutzer, K. Gauvreau, M. Freed, J. E. Mayer Jr, R. Jonas, and P. J. del Nido Twenty-five-year experience with Rastelli repair for transposition of the great arteries J. Thorac. Cardiovasc. Surg., August 1, 2000; 120(2): 211 - 223. [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 |