| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2007;84:871-879
© 2007 The Society of Thoracic Surgeons
a Division of Pediatric Cardiovascular Surgery, C. S. Mott Children’s Hospital, Ann Arbor, Michigan
b Division of Pediatric Cardiology, C. S. Mott Children’s Hospital, Ann Arbor, Michigan
Accepted for publication April 2, 2007.
* Address correspondence to Dr Ohye, F7830 Mott/0223, 1500 East Medical Center Dr, Ann Arbor, MI 48109 (Email: ohye{at}med.umich.edu).
Presented at the Forty-third Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 29–31, 2007.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Methods: All patients undergoing an arterial switch operation at a single institution from January 1, 1999, to September 1, 2005, were reviewed.
Results: Of the 168 patients, median age was 2 days (range, 0 to 358) and weight was 3.5 kg (range, 1.9 to 11.8 kg). Eleven patients were less than 36 weeks gestational age. Forty percent had coronary patterns other than usual. Mean cardiopulmonary bypass (CPB) time was 147 ± 45 minutes, and mean cross-clamp time was 77 ± 27 minutes. At a mean follow-up of 19 ± 21 months, there were 10 (6%) hospital and 4 (3%) late deaths. Actuarial 1-month, 1-year, and 3-year survivals were 94%, 90%, and 89%, respectively. Bivariate analysis revealed weight less than 2.5 kg (p = 0.032), gestational age less than 36 weeks (p = 0.002), and CPB time greater than 150 minutes (p = 0.0075) decreased hospital survival. Intermediate-term survival was negatively impacted by weight less than 2.5 kg (p = 0.017), gestational age less than 36 weeks (p = 0.0096), CPB time greater than 150 minutes (p = 0.0050), and age at presentation greater than 4 weeks (p = 0.034). By multivariate analysis, gestational age less than 36 weeks (p = 0.0051) and CPB time greater than 150 minutes (p = 0.016) were independent risk factors for hospital mortality. Gestational age less than 36 weeks (p = 0.0096) and CPB time greater than 150 minutes (p = 0.005) were also independent predictors of intermediate-term mortality. Coronary anatomy could not be shown to affect survival, including no deaths among the 12 patients with intramural coronaries.
Conclusions: The arterial switch operation can be performed with low mortality regardless of diagnosis or coronary pattern. The premature patient and minimizing CPB time remain as challenges to optimize outcomes for the arterial switch operation.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
The purpose of this study is to review the ASO at our institution with an emphasis on patient management and outcome according to the anatomic subtype of d-TGA. In addition, this study intends to determine the current risk factors for poor outcome.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Coronary Anatomy
The coronary artery anatomy was described using the Leiden convention, whereby the coronary pattern is described based on the sinuses of origin of the various coronary artery branches. If one imagines oneself standing in the nonfacing aortic sinus looking at the pulmonary root, the right hand sinus is numbered 1 and the left hand sinus is 2. For example, in the most common arrangement, the left anterior descending and circumflex coronary arteries arise from sinus 1 and the right coronary artery from sinus 2. This pattern is described as 1LCx2R.
For purposes of analysis, the various coronary patterns were grouped initially as usual (1LCx2R) versus not usual. The patterns were then grouped utilizing a convention based upon perceived difficulty of coronary translocation, as well as outcome after ASO, as described by Pasquali and colleagues [10]. The coronary variations were divided into groups. Group 1 consisted of dual coronary artery systems without looping around either great vessel including 1LCx2R, 1LCx2LR, and 1Cx2LR. Group 2 coronary arteries had looping and two orifices including 1L2RCx, 1R2LCx, 1RCx2L, and 1L2LRCx. Groups 3 and 4 consisted of single and intramural coronary arteries, respectively.
Timing of Surgical Intervention
The congenital heart defect was generally addressed surgically within the first 14 days of life by an ASO. In case of prematurity, low birth weight, or other medical complications precluding definitive repair, surgery may have been postponed with an initial palliative procedure. In late presenting patients (>4 weeks of age), a pulmonary artery band and modified Blalock-Taussig shunt were utilized to recondition the left ventricle before operation.
Surgical Technique
The surgical technique for the ASO can be summarized as follows. Standard median sternotomy is performed and a patch of pericardium is harvested and placed in dilute glutaraldehyde solution. The operation is generally conducted with cardiopulmonary bypass (CPB) through standard aortic and bicaval cannulae at temperatures between 20°C and 32°C depending on surgeon preference. If the superior vena cava diameter does not permit cannulation, single atrial cannulation is utilized along with deep hypothermic cardiac arrest. In addition, hypothermic arrest is also occasionally employed at the surgeon’s discretion when difficulties in operation are encountered or for aortic arch reconstruction. Cold blood cardioplegia is utilized approximately every 20 minutes during the cross-clamp period. Delivery is initially in an antegrade fashion followed by retrograde cardioplegia after division of the ascending aorta.
The aorta is transected at its midpoint (Fig 1A), and the coronary arteries are excised from their respective sinuses of Valsalva. The main pulmonary artery is divided at its bifurcation (Fig 1A), and the anterior commissure of pulmonic valve is marked with a suture. Next, the Lecompte maneuver is performed (Fig 1B). In contrast to the more common technique of performing the coronary translocation with the neoaortic root open distally, the technique employed at our institution involves first completing the neoaortic anastomosis. At this point, the aortic clamp can be removed, allowing the neoaorta to fill and facilitating the precise placement of the coronary arteries.
|
In the presence of d-TGA/VSD, the VSD is closed with a patch of polytetrafluoroethylene. In the case of DORV, the VSD is enlarged if necessary and baffled to the neoaorta with polytetrafluoroethylene. Occasionally, in the case of a noncommitted VSD requiring a complex intraventricular tunnel, it is useful to form the patch from a piece of stretch polytetrafluoroethylene tube graft.
Statistical Analysis
Data were collected using Microsoft Excel software (Microsoft, Redmond, Washington). Normal and nonnormal data are expressed as means (±SD) and medians (range), respectively. The SPSS software (SPSS, Chicago, Illinois) and JMP Statistical Discovery Software (SAS Institute, Cary, North Carolina) programs were utilized to evaluate data. Dichotomous variables were analyzed with Fisher’s exact test and continuous variables were subject to Student’s t test. Wilcoxon rank-sum testing was utilized for nonnormally distributed data. Multivariate logistical models were constructed to evaluate potential independent influence of the variables on survival. Kaplan-Meier curves were used for freedom from reoperation and survival analysis. The nonparametric Mann-Whitney U test was utilized to compare variables according to type of operation. A p value of less than 0.05 was considered to be statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
The median age at presentation was 2 days (range, 0 to 358). The mean weight at operation for the entire group was 3.5 kg (± 1.1). Of the 168 patients to undergo an ASO, 15 (9%) had a weight of below 2.5 kg at the time of operation and 11 (7%) were born at a gestation age of less than 36 weeks. Bivariate analysis revealed weight at operation as a continuous variable (p = 0.010), weight less than 2.5 kg (p = 0.032), and gestation age less than 36 weeks (p = 0.020) adversely affected hospital survival. For intermediate-term survival, bivariate analysis demonstrated weight less than 2.5 kg (p = 0.017), age at presentation greater than 28 days (0.034), and gestation age less than 36 weeks (p = 0.096) to be significant risk factors. By multivariate modeling, gestation age less than 36 weeks was found to be an independent predictor of hospital mortality (p = 0.0051) with an odds ratio of 12.0 (95% confidence interval: 2.1 to 66.7). Similarly, gestation age less than 36 weeks was an independent risk factor for intermediate-term mortality (p = 0.0096) with an odds ratio of 11.4 (95% confidence interval: 1.89 to 66.7).
The coronary artery pattern distribution, according to the patient diagnosis, can be seen in Table 1. For the entire group, the coronary artery patterns were usual (1LCx2R) in 100 patients (60%) and abnormal in 68 patients (40%), including 29 (17%) with circumflex from the right coronary artery (1L2CxR) and 12 (7%) with intramural patterns. By group, as defined above, there were 104 patients in group 1, 43 in group 2, 9 in group 3, and 12 in group 4. Mortality by usual versus not usual coronary artery pattern (p = 0.527) and by group (p = 0.514) could not be correlated to hospital survival. Notably, there were no deaths among the 12 patients with intramural coronary arteries.
|
|
Modified Blalock-Taussig shunts were performed in 4 patients (2%), for reconditioning in 3 patients and as an initial plan for a single ventricle repair at another institution in the remaining patient. Of the 168 patients with d-TGA, 7 (4%) required a pulmonary artery band, including 2 patients who underwent palliation at an outside institution before transfer, 2 patients who had a pulmonary artery band concurrently with a modified Blalock-Taussig shunts for left ventricular reconditioning, and 1 patient who was palliated for necrotizing enterocolitis. Two patients with d-TGA/DORV underwent repair for coarctation of the aorta before the ASO. In addition, 1 patient (1%) underwent division of the pulmonary artery with Damus-Kaye-Stansel anastomosis performed at an outside institution as result of an initial plan for a single ventricle repair.
Definitive Corrective Procedure
The median age at the time of repair for the 168 patients undergoing the ASO was 7 days (range, 2 to 937). The age at operation for d-TGA/DORV patients was statistically higher than the age at operation for both d-TGA/IVS (p < 0.0005) and d-TGA/VSD (p < 0.0005) patients (Table 3). Fourteen patients (8%) underwent the ASO outside of the neonatal period, most commonly because of late presentation (n = 6). Other causes included poor clinical condition (gestation age less than 36 weeks = 3, necrotizing enterocolitis = 1, bilateral strokes = 1) and planned delayed definitive operation from outside institutions (n = 3). Age at operation as a continuous variable (p = 0.361) and ASO greater than 30days (p = 0.169) could not be shown to be associated with hospital mortality.
|
|
Hospital Stay
The median length of hospital and intensive care unit stay for all patients undergoing an ASO was 15 days (5 to 83) and 6 days (2 to 50), respectively. Both the length of hospital (p = 0.045) and intensive care unit stay (p = 0.005), were statistically higher for patients with d-TGA/DORV compared with those with d-TGA/IVS (Table 3). Significant morbidities are summarized in Table 5. From the 168 patients who presented with d-TGA, there were 10 hospital mortalities (5.95%). Causes of hospital mortality are listed in Table 6. A difference in hospital mortality based upon underlying diagnosis could not be proven (p = 0.47).
|
|
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Our experience consisted of a relatively high percentage of d-TGA patients with a VSD (51%), whether d-TGA/VSD or d-TGA/DORV. In previous studies, the proportion of patients with a VSD is reported to be approximately 30% to 40% [3, 6–7]. Our higher than average percentage of patients with a VSD may be explained by three reasons. The first is that our center is a tertiary care institution and thus receives a higher proportion of complex cases. These complex cases are more frequently d-TGA/VSD and d-TGA/DORV patients, as these subtypes are more commonly associated with associated cardiac anomalies. Similarly, we are a tertiary fetal echocardiography center. A larger proportion of patients with a congenital cardiac malformation associated with a VSD are detected by routine fetal screening ultrasonography, compared with those with no findings on a routine four-chamber view, such as d-TGA/IVS. Thus, patients referred to our fetal echocardiography program with positive findings on a routine screening ultrasonogram also tend to be biased toward d-TGA/VSD. Lastly, our high percentage of patients with a VSD can also be explained as a result of our experienced echocardiographers noting small VSDs that otherwise may be missed. Although 63 patients were diagnosed with d-TGA/VSD, 12 (19%) of these patients had hemodynamically insignificant VSDs that did not require closure.
The presence of a VSD or DORV in association with d-TGA has been previously shown to increase mortality [3, 7, 9–10]. The current data were not able to show a relationship between diagnosis and hospital or intermediate-term survival. The absence of this relationship in our experience may be the result of patient numbers and low mortality, or the somewhat higher detection rate for hemodynamically insignificant VSDs. The lack of correlation may also be attributed to improvements in management of ASO patients over time. These advances are also evidenced by more recent studies, which tend to show that the adverse effect of the presence of VSD or DORV on survival is decreasing over time [1, 5].
Our study revealed a relationship between prematurity (gestation age less than 36 weeks) and both hospital and intermediate-term mortality, whereas weight less than 2.5 kg alone could not be shown to be a significant risk factor. Although the vast majority of premature neonates will also be of low birth weight, they are also at risk for other short- and long-term risks associated with preterm birth, such as necrotizing enterocolitis and acute and chronic respiratory insufficiency. This association of prematurity and poor outcome for congenital heart disease, and for the ASO in particular, has been found in other studies [12–15]. As with the lack of significance for the presence of VSD, limited patient numbers and low mortality rates may have impacted the ability to detect differences.
Cardiopulmonary bypass time greater than 150 minutes was an independent predictor of both hospital and intermediate-term mortality. Cardiopulmonary bypass as a technique has well-described, time-dependent deleterious effects including a whole-body inflammatory response, metabolic changes, and fluid and electrolyte imbalance [16]. In addition to these negative influences, prolonged CPB may also be a surrogate for other factors, such as complexity of anatomy and operation. The inherent injury due to CPB, as well as the associated patient factors, may contribute to less favorable outcomes after ASO, as has been seen with other congenital heart lesions [17].
In past studies, it has been found that patients with certain abnormal coronary patterns have a significantly increased mortality risk [3, 6, 9–10, 18]. In particular, Pasquali and associates [10] reported that patients with any variant coronary pattern other than the usual had an almost twofold increased risk of hospital mortality. They found that the highest risk patterns were a single coronary ostium and intramural coronary artery pattern with a threefold and greater than sixfold increased mortality risk, respectively [10]. In our study, however, the presence of usual versus not usual coronary artery pattern (p = 0.527) and coronary artery pattern by group (p = 0.514) could not be correlated to hospital survival.
While the limited patient population in this report may have contributed to this lack of association, several of these previous studies correlating coronary artery pattern to hospital survival were conducted with similar numbers of patients. Although statistically the data can only be strictly interpreted as having insufficient numbers to demonstrate a correlation, we can infer that the risk has at least been decreased compared with previous studies, if not eliminated. Our technique for coronary translocation, which entails anastomosing the main pulmonary artery to the ascending aorta before translocating the coronary buttons, may contribute to this improvement. This method allows the neoaortic root to be distended to a natural state by removing the aortic cross clamp, thereby facilitating optimal coronary artery placement without tension or twisting. Brown and colleagues [6] found that when they switched to this technique, coronary artery anatomy was eliminated as a significant risk factor for mortality. While some surgeons have even gone as far as to suggest that certain coronary patterns are relative contraindications to an ASO, we strongly disagree with this position. Despite a population with a high incidence of these high-risk coronary patterns, the survivals were excellent, including no deaths among 9 patients with a single coronary ostium and 12 patients with intramural coronary arteries (3 bilateral intramural).
Our group demonstrated overall excellent outcomes, comparable to other contemporary studies [1, 4–6, 15]. The current hospital and intermediate-term survivals for the ASO can be attributed to improvements in preoperative, intraoperative, and postoperative care of these patients. We would infer that the technique for coronary transfer utilized at out institution may play an important role in maximizing outcomes for both usual and complex coronary patterns. The premature neonate and minimizing cardiopulmonary time remain as challenges in achieving optimal results for the ASO.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
DR MUHAMMAD A. MUMTAZ (Cleveland, OH): Zuhab, that’s an outstanding job.
MR QAMAR: Thank you.
DR MUMTAZ: Really, it is. A very wonderful presentation. I enjoyed it. I read the abstract before I came today, and had a few things that I’d like to understand. If you could elaborate a little bit more on the mortality and the cause of death, I’d like to know what’s the mortality in the group that was premature? I’d like you to elaborate on the fact that the cardiopulmonary bypass time of greater than 150 minutes, was that related to difficult coronaries? Your end conclusion, it’s not. Then what is it related to? And the low birth weight, again, is low birth weight in your group related to mortality if you take prematurity out?
MR QAMAR: I’d like to refer these questions to Dr Ohye.
DR RICHARD G. OHYE (Ann Arbor, MI): If I could take the liberty to maybe address some of the questions. Having just graduated from college, Zuhab is maybe not best equipped to handle some of the questions.
We performed multivariate modeling on many potential risk factors, including cardiopulmonary bypass time and coronary arrangement. We did not specifically look at the relationship between coronary anatomy and cardiopulmonary bypass time. However, bypass time did come out to be an independent predictor of hospital and intermediate term death, whereas coronary pattern did not.
What were your other questions?
DR MUMTAZ: Prematurity and low birth weight.
DR OHYE: In what respect? I’m sorry.
DR MUMTAZ: In terms of mortality. What was the mortality and the level of prematurity?
DR OHYE: By multivariate analysis, gestational age less than 36 weeks was an independent risk factor for both hospital and intermediate term mortality. Although on bivariate analysis weight less than 2.5 kg adversely affected both hospital and intermediate term survival, low birth weight did not reach significance on subsequent multivariate modeling.
DR CHRISTOPHER A. CALDARONE (Toronto, ON): Rick, are you comfortable with the conclusion that bypass time is inherently responsible for poor outcome, or are you more comfortable with the idea that they’re just simply a number of subsets in your population that happen to be associated with long bypass time and poor outcome?
DR OHYE: Well, I think the answer to that is I don’t know, frankly. We looked at a whole laundry list of factors that we thought might be related to survival. These factors included many that one might consider to be a surrogate for complexity of operation, such as coronary anatomy, diagnosis, presence of associated anomalies, the requirement for arch reconstruction, etcetera. When you factor all these considerations in, long cardiopulmonary bypass time still comes out to be an independent risk factor.
Now, that doesn’t mean, of course, that we plugged everything into the model that it could be a surrogate for, but we tried to think of everything that could be represented within that bundle of cardiopulmonary bypass time. In general, if you look at many other studies, cardiopulmonary bypass time seems to repeatedly come out as a risk factor for hospital mortality.
How we explain the fact that conditional survival once you get out of the hospital still continued to be affected by cardiopulmonary bypass is much more difficult.
With this in mind, you asked me how comfortable I am that it’s cardiopulmonary bypass time. Probably not all that comfortable, I think that it is probably a surrogate for other factors that we’re just not able to determine. But we did plug many, many demographic, anatomic, etcetera, variables into the equation, and the data would say that it is indeed an independent risk factor.
DR CALDARONE: And presumably you had a few patients who were late diagnoses. I noticed you had a few who were banded and retrained. What’s your standard policy for patients who arrive late with a diagnosis of transposition?
DR OHYE: These late presenting patients were a relatively small population. Therefore, we couldn’t really do any rigorous analysis from a statistical standpoint. From a policy standpoint, generally we would reserve banding only for those early presenting patients in poor clinical condition or those patients who present after about 4 weeks. After that time we would consider performing a rapid two-stage repair.
DR RAFAEL R. GUERRERO (Leeds, UK): Congratulations to Mr Qamar on his graduation and his excellent presentation. A question to Dr Ohye. Discussions about outcome after the arterial switch operation have been around for quite a while. However, I agree that it is very appropriate to update and review the risk factors and outcome in the current era. A few months ago, Tom Karl published the impact of preoperative atrial septostomy in the neurologic outcome of this group of patients. Do you have any experience in your institution with this approach and if yes, has the neurologic outcome changed or improved?
DR OHYE: We didn’t specifically look at their neurologic outcome. A large percentage of our patients did require either septostomy or open septectomy, but we didn’t attempt to relate it to neurologic outcome.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. Mace, F. Vanhuyse, J.-M. Jellimann, D. Youssef, A. Moulin-Zinsch, J.-P. Lethor, and F. Marcon Arterial Switch Operation With a Single Coronary Artery: The Autograft Concept. Ann. Thorac. Surg., June 1, 2009; 87(6): 1967 - 1968. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. E. Neufeld, B. G. Clark, C. M.T. Robertson, D. M. Moddemann, I. A. Dinu, A. R. Joffe, R. S. Sauve, D. E. Creighton, L. Zwaigenbaum, D. B. Ross, et al. Five-year neurocognitive and health outcomes after the neonatal arterial switch operation J. Thorac. Cardiovasc. Surg., December 1, 2008; 136(6): 1413 - 1421. [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 |
