| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2007;84:1331-1337
© 2007 The Society of Thoracic Surgeons
a The Cardiac Centre, Hospital for Sick Children and the University of Toronto, Toronto, Ontario, Canada
b King Faisal Heart Institute at King Faisal Specialist Hospital and Research Centre, Riyadh, Saudi Arabia
Accepted for publication May 1, 2007.
* Address correspondence to Dr Alsoufi, King Faisal Heart Institute (MBC 16), King Faisal Specialist Hospital and Research Centre, PO Box 3354, Riyadh 11211, Saudi Arabia (Email: balsoufi{at}hotmail.com).
Presented at the Forty-third Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 29–31, 2007.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Methods: Between 1990 and 2006, 141 neonates with CoA and VSDs underwent operation using two management strategies. In group A (n = 89), initial simple CoA repair was done through posterolateral thoracotomy, plus concomitant pulmonary artery banding (n = 54), followed by VSD closure. In group B (n = 52), both defects were repaired simultaneously through a sternotomy.
Results: Overall 10-year survival was 90.8%, with no difference between groups. The 5-year freedom from arch reoperation was 93.5%, with no difference between groups. The 10-year freedom from reoperation for subaortic obstruction was 95% for group A and 75% for group B (p = 0.016). In group A, 41 patients required secondary VSD closure at a median interval of 48 days after CoA repair. Freedom from reoperation at 1 month and 5 years was 78.5% and 45.8% in group A versus 97.8% for both in group B. Preoperative predictors for requirement for later VSD closure in group A were VSD type other than muscular (p = 0.0009) and larger VSD identified by higher VSD diameter/aortic valve annulus ratio (p < 0.0001).
Conclusions: Results of both treatment strategies are good. Neonates with larger VSDs, especially outlet, malalignment, and perimembranous types, are likely to require VSD closure. Although midline sternotomy and combined treatment strategy may be necessary in neonates with proximal arch hypoplasia, initial coarctation repair alone is valid option at the possible expense of additional operation.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
In the current study, we compared the surgical outcomes between those two treatment approaches in neonates undergoing operation for CoA and associated VSDs at a single institution.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Ventricular Septal Defect Morphology
We used the unified VSD nomenclature system for VSD classification [12]. A single VSD was present in 81% of patients, and 19% had multiple VSDs. The most common VSD type was perimembranous (57%), followed by muscular (35%), outlet (6%), and inlet (3%). Of the perimembranous VSDs, 9% extended into the outlet septum, 6% into the inlet septum, and 4% into both outlet and inlet septum.
Operative Technique
In patients undergoing initial CoA repair alone (n = 89), this was performed through a left posterolateral thoracotomy. After dissection and heparinization, vascular control was obtained proximally and distally, the ductus arteriosus was ligated, the coarctation area was completely resected, and the aortic reconstruction was performed with a subclavian artery flap or an extended end-to-end anastomosis.
When PAB was used (n = 54), the pericardium was incised, the pulmonary artery was then isolated, and a 3-mm-wide Teflon tape (DuPont, Wilmington, DE) was applied below the bifurcation. Trusler’s formula was used to assess the circumference of the band [13]. The band was secured to the adventitia of the pulmonary artery at multiple sites to prevent migration.
Second stage closures of VSDs (n = 41) were performed through a midline sternotomy with standard aortic and bicaval cannulation techniques with cold antegrade blood cardioplegia for myocardial protection. Debanding was performed with pulmonary artery angioplasty. The VSD was closed transatrially by using a Dacron (DuPont) patch.
Combined total repairs (n = 52) were performed through a midline sternotomy. Standard aortic and bicaval cannulation techniques were applied with cold antegrade blood cardioplegia for myocardial protection. DHCA was used for arch reconstruction, although in recent years we have used selective antegrade cerebral perfusion in patients requiring prolonged DHCA for arch reconstruction. The coarctation area was completely excised and the aortic continuity reestablished. The VSD closure was performed transatrially during the rewarming period. Transesophageal echocardiography was routinely used postoperatively to assess the adequacy of the repair.
Follow-Up
Late outcomes were determined from recent office visits at the Hospital for Sick Children or from written correspondence with patients’ community cardiologists. Freedom from recurrent coarctation was determined by reviewing serial echocardiograms. Mean follow-up was 6.2 ± 5.2 years for group A and 3.2 ± 3.6 years for group B (p = 0.0004).
Statistical Analysis
All the data were analyzed with SAS 9 software (SAS Institute, Inc, Cary, NC). Data were presented as frequency, median with range, or mean ± standard deviation, as appropriate, with the number of nonmissing values indicated. Unrelated two-group comparisons were done with unpaired, two-tailed t tests for continuous variables and the 
2 or Fisher exact test for categoric data. Significant factors on univariate analysis were entered into a logistic regression model to determine significant factors on multivariable analysis. Estimates for long-term survival or freedom from morbid events were made by the Kaplan-Meier method. Differences between survival curves were evaluated with the log-rank statistic.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
For group A and B patients, respectively, the mean age was 10.9 ± 6.9 versus 14.7 ± 8.2 days (p = 0.0042), the mean weight was 3.2 ± 0.6 versus 3.1 ± 0.6 kg (p = 0.44), and the mean body surface area was 0.22 ± 0.04 versus 0.21 ± 0.04 (p = 0.42). The mean VSD diameter in group A and B patients was 0.43 ± 0.19 cm versus 0.59 ± 0.23 cm (p < 0.0001), the mean transverse aortic arch diameter was 0.36 ± 0.09 versus 0.34 ± 0.08 cm (p = 0.10), and the mean aortic annulus was 0.58 ± 0.1 versus 0.55 ± 0.09 cm (p = 0.06), respectively. Comparison between the two patient groups is listed in Table 1.
|
Factors Associated With Later Ventricular Septal Defect Closure
In group A, 41 patients (46%) required secondary VSD closure at a median interval of 48 days (range, 9 days to 10.4 years) from the initial CoA repair. One patient in group B required reoperation and closure of a significant recurrent VSD. Freedom from secondary VSD operation in group A at 1 month and 5 years was 78.5% ± 4% and 45.8% ± 6% versus 97.8% ± 2% and 97.8% ± 2% in group B (p < 0.0001; Fig 1).
|
|
|
During the follow-up period, 9 patients, 6 in group A and 3 in group B, underwent reoperation and revision of arch reconstruction for recurrent arch obstruction. Eight of the reoperations were performed within the first year of initial coarctation repair. The 5-year freedom from arch reoperation was 92.2% ± 3% for group A and 95.7% ± 3% for group B (p = 0.95; Fig 4).
|
|
Recurrent coarctation was noted in 15 patients. This was defined as an echocardiographic peak gradient exceeding 20 mm Hg. Freedom from recurrent coarctation was 86% ± 5% for group A and 74% ± 12% for group B (p = 0.23).
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Predictors of Spontaneous Ventricular Septal Defect Closure
Surgeons advocating initial CoA repair alone note that many VSDs close spontaneously, and thus cardiopulmonary bypass can be avoided in a significant proportion of patients [1–3, 10, 11]. In the current study, nearly half the patients in the two-stage group enjoyed this benefit. Nevertheless, many neonates continued to have persistent evidence of congestive heart failure and failure to thrive after isolated CoA repair alone and have required reoperation during the same hospital stay, with the additive morbidity, cost, and length of stay [1–3, 10, 11].
Factors considered when evaluating management strategies for neonates with VSD and CoA should include the magnitude of the shunt through the VSD, the likelihood that symptoms of congestive heart failure after CoA repair alone will persist, and whether the VSD is of a type that is likely to close spontaneously. In our series, significant multivariable predictors for the need of a later VSD closure in patients undergoing initial CoA repair were VSD type other than muscular and large VSD size. Muscular VSDs were more likely to close spontaneously than other types such as inlet and perimembranous.
The number of VSDs present in our series was not found to be a predictor for the need of a secondary VSD operation after the initial CoA repair alone. This finding may be related to the fact that most of those multiple VSDs were muscular and small in nature and therefore were more likely to close spontaneously. In few cases, however, the cumulative shunt was significant enough to warrant intervention. Although a combination of CoA and PAB may be proper in those patients, it is our current policy to avoid PAB and completely close the VSDs in those patients with significant cumulative size of defects [14].
As shown in Figures 2 and 3, larger-size defects are less likely to close spontaneously, are more likely to cause persistent symptoms, and therefore, are more likely to require secondary closure after the initial CoA repair. When VSD diameter was less than 50% of aortic valve diameter, less than 20% of VSDs required surgical closure; conversely, more than 60% of VSDs equal to or larger than the aortic valve diameter required secondary VSD closure (Fig 2). The relationship between the size of the VSD and the need for an operation was more evident for VSD types other than muscular VSDs (Fig 3). The predicted freedom from later closure for VSDs equal to or larger than the aortic valve annulus was less than 20% for those types of VSDS compared with more than 70% for muscular VSDs. Therefore, a VSD with a size larger than half of the aortic valve diameter—especially a perimembranous, inlet, outlet or mal-alignment VSD—is likely to require eventual surgical closure, and this consideration should be included in choosing between one-stage and two-stage management strategies.
Arch Hypoplasia and Recurrent Obstruction
Proximal arch hypoplasia is commonly encountered in neonates with CoA. Surgeons use different criteria for the definition of proximal arch hypoplasia, including using the Z value, comparing with the ascending aorta, and using different formulas according to the patient’s size or weight. The prevalence of arch hypoplasia in our series using these criteria was 44% to 75%.
Total repair through a median sternotomy allows more proximal extension of the incision into the ascending aorta and the construction of a larger anastomosis and with augmentation of the entire aortic arch with a patch when necessary.
In this retrospective review, the mean transverse arch size was smaller in patients who underwent combined total repair than in those who underwent initial CoA repair alone. The presence and extent of proximal arch obstruction should be taken into consideration during the initial assessment of patients for potential one-stage repair. The use of this selection bias resulted in similar freedom from recurrent coarctation between groups A and B despite the presence of greater arch hypoplasia in group B, suggesting that arch hypoplasia was neutralized as a risk factor for recurrent arch obstruction through triage of patients with more proximal arch hypoplasia into the combined management strategy.
Subaortic Stenosis
Patients in the current series who underwent combined repair were significantly more likely to require reoperation for subaortic membrane obstruction during the follow-up period. This finding may be because those who were selected to undergo combined repair had a smaller aortic valve annulus and smaller arch diameter preoperatively. Consequently, the size of the subaortic area and the adequacy of the aortic valve annulus should also be taken into consideration when the initial operation is planned. Resection of a discrete fibrous membrane was possible in all but 1 patient (who had a second recurrence and required a third operation with a modified Konno procedure).
The combined treatment approach offers total correction and eliminates second-stage VSD closure and the need for an initial left thoracotomy, followed by a later sternotomy. DHCA is needed during arch reconstruction, however. Concerns have been raised about the adverse neurodevelopmental effects of DHCA in the neonatal period [15–18], but there has not been a direct comparison of neurologic injury or long-term neurologic development between patients undergoing combined repair with the use of cardiopulmonary bypass and DHCA and patients undergoing initial CoA repair alone.
Nevertheless, it seems prudent to avoid DHCA when possible, and it can be argued that an approach using initial CoA repair through a thoracotomy is most appropriate in patients without arch hypoplasia in whom a VSD is likely to close spontaneously. Benefits of this approach would include the avoidance of cardiopulmonary bypass, transfusion of blood products, and potential neurologic injury. In patients undergoing initial CoA repair and in whom a PAB is required, the relative benefit of this approach is diminished due to the requirement that the band be removed—even if the VSD closes—thus necessitating a second operation with possible cardiopulmonary bypass if pulmonary angioplasty is needed.
Conclusions
Several factors must be considered when evaluating management strategies for neonates with CoA and an associated VSD. Those include the likelihood that the VSD will close spontaneously; whether the VSD size is large enough to result in continued symptoms of congestive heart failure after CoA repair, and whether proximal arch hypoplasia or stenosis of the aortic valve and subaortic area is present. The present data suggest that a combined repair approach is best suited for neonates with proximal arch hypoplasia, whereas initial CoA repair is best suited for neonates in whom the VSD is likely to close, a band is not required, and arch hypoplasia is not present. Significant overlap exists, however, and many patients fall in the "gray zone." In those patients, our data suggest that both approaches are associated with good operative results and almost similar time-related morbidity.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
DR ALSOUFI: I agree with you. In our series, neonates who received single-stage surgery had smaller transverse aortic arch diameter, more proximal arch hypoplasia, and smaller left ventricular outflow tract and aortic valve annulus diameter compared to those who received the two-stage approach. That may explain our finding that single-stage patients required more reinterventions for sub-aortic left ventricular outflow tract obstruction on follow-up compared to those in the staged group.
DR GERHARD ZIEMER (Tuebingen, Germany): Congratulations on your results. We have similar results in a smaller group. Over the last 10 years, we looked up over 45 patients. I am missing some data, though. First of all, obviously your two-stage candidates were not ideal two-stage candidates because your conclusion says an ideal two-stage candidate is a patient who amongst other things needs no band. But most of your two-stage candidates needed bands.
Our approach is different. We either do one-stage repair or coarctation only. And if a coarctation-only patient requires a VSD closure either early or later after coarctation surgery, he doesn’t get a band, then he gets the VSD closed.
You have only 41 patients who required a secondary VSD closure, but 54 had been banded. This means obviously some patients needed debanding only, which we definitely call an avoidable operation. As clearly some nonbanded patients required secondary VSD closure also, making up for a certain part of your 41 patients, I expect even more avoidable operations in your group as the number of debandings must be around 20 or 25 even, or almost 50% of banded patients. So this supports our preference, not to band at all. This is what Dr Castaneda taught me already more than 20 years ago, any ways.
DR ALSOUFI: Thank you. This series goes back to 1990. In 1994, a CHSS study showed that in neonates with coarctation and VSD, the best nonadjusted survival was in those patients who received initial coarctation repair in conjunction with pulmonary artery banding. That explains why there were numerous patients who received coarctation repair and banding early in our series. We very rarely utilize pulmonary artery bands for patients with VSD in our current practice.
Many patients who had received an initial coarctation repair and pulmonary artery banding, and in whom the VSD closed spontaneously, did require reoperation to remove the PA band and occasionally perform pulmonary angioplasty, which makes your point that by banding we are subjecting those patients to an unnecessary second operation very valid.
DR TAE-GOOK JUN (Seoul, Korea): Thank you for the nice presentation. My question is what technique do you use in your single-stage operation, that is end-to-end anastomosis or end-to-side anastomosis for reconstruction of arch? What technique do you use?
DR ALSOUFI: In the single stage?
DR JUN: Yes.
DR ALSOUFI: We do end-to-side anastomosis of the descending aorta to the aortic arch after complete resection of all ductal tissue. Very occasionally, in patients with diffusely hypoplastic aortic arch, patch augmentation was necessary.
DR JUN: Well, in my opinion, as you presented, the transverse arch is small, but the incidence of recoarctation is higher I think. In our group, the incidence of recoarctation is less than 5%. I already use end-to-side anastomosis. I use already on anastomosis the descending aorta to the proximal ascending aorta.
DR ALSOUFI: So are you suggesting to add the patch?
DR JUN: No, no. I mean, how often do you use the end-to-side anastomosis?
DR ALSOUFI: The majority of the patients required end-to-side, with only a few receiving patch augmentation for diffusely hypoplastic arch.
DR JUN: Oh, really. I see.
DR CARL L. BACKER (Chicago, IL): That was a very nice study with a large number of patients. I think it is very interesting that we are still comparing these two treatment strategies! This is one of those debates that we will still be engaged in 10 years from now.
I would draw everyone’s attention to Hal Walters’ poster, which is on page 303 of the program booklet. He reaches the same conclusion that there was no difference in outcomes between the staged approach and the single-stage repair.
The question that I have relates to your statistical analysis. Did you consider the different time periods when the procedures were performed in your analysis that found no difference in outcomes? What I would suggest is that it is likely that the two-stage group is clustered in the early part of your cohort, which would have been the early 1990s, as compared to the later cohort. I think we would all agree that there have been significant improvements in surgical techniques for cardiopulmonary bypass, modified ultrafiltration, et cetera, in the last quartile of that study. In point of fact, if you used the staged approach now, in 2007, you might have better results than your single-stage approach if those staged patients are indeed clustered in the early part of your series. That is my first question.
I have two other questions. First, did you look at the incidence of recurrent laryngeal nerve injury? When we compared our results with coarctation repair from the front versus the side, we had a higher incidence of recurrent laryngeal nerve injury through the sternotomy approach.
My final question relates to the avoidance of cardiopulmonary bypass in the neonatal group to protect the brain. This relates to some of the data that Bill Gaynor has presented about periventricular leukomalacia. Should we be considering that factor as to whether or not we want to put this newborn on bypass or wait until the brain is a little bit more mature? And if we have equivalent outcomes vis-a-vis the anatomic result and mortality, should we be considering this and avoiding bypass in the neonatal group?
DR ALSOUFI: Thank you, Dr Backer. We did not look at the incidence of recurrent laryngeal nerve injury, so unfortunately I don’t have this data.
I agree with you that based on many recent reports, including Dr Gaynor’s, there are concerns that the neonatal brain is more vulnerable to ischemic injury. That may suggest that avoiding deep hypothermic circulatory arrest and cardiopulmonary bypass during the neonatal period may be beneficial. I think this question can only be answered with a randomized, prospective review when a thorough neurological assessment is performed periodically following surgery. I don’t think it can be answered by any retrospective review.
As for your first question, you are correct: the two groups are not exactly contemporaneous. Many of the patients who received two-stage procedure plus/minus pulmonary artery banding were done earlier in our series and more patients received single-stage total correction later on in the series. It is obviously correct that outcomes following cardiac surgery have improved significantly in the last decade due to advances in perfusion, surgical technique, and postoperative care. Nonetheless, I believe that the maximum improvement was achieved in those patients requiring prolonged cardiopulmonary bypass and deep hypothermic circulatory arrest. Therefore, a more contemporary review of surgical outcomes may prove to be to the benefit of the single-stage group.
DR HENRY L. WALTERS III (Detroit, MI): I want to congratulate you on a really superb study. This is going to be the very best analysis of predictors for spontaneous VSD closure in coarctation-VSD in the literature.
Another way to look at it, though, to take the opposite tack to what Dr Backer was saying, is that when one uses a single-stage approach in these patients, they reach normal anatomy and physiology at a statistically significantly earlier age than when the two-stage approach is used. At least, that is the case when we analyzed our data. And we know from Bill Gaynor’s work at CHOP that there are prenatal, postnatal, preoperative, and postoperative factors that are probably even more important than intraoperative variables in determining eventual neurodevelopmental outcome. This fact may, in itself, constitute a good reason to use the single-stage approach to achieve a complete repair as early as possible.
DR ALSOUFI: Absolutely. As you mentioned, the neonatal brain is vulnerable to ischemic and hypoperfusion injury not only during surgery but also in the postoperative period. Therefore, a procedure that will provide the optimal postoperative hemodynamics and stability is desired. Again, the question of neurologic development can only be answered with a randomized, prospective review when a thorough neurological assessment is performed periodically following surgery and cannot be answered by any retrospective review.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  O. S. Goksel and E. Tireli Surgical Strategy in the Treatment of Neonates With Aortic Coarctation and Associated Ventricular Septal Defects Ann. Thorac. Surg., July 1, 2008; 86(1): 352 - 352. [Full Text] [PDF]
|
|
|
|
|
|
B. Alsoufi and C. A. Caldarone Reply Ann. Thorac. Surg., July 1, 2008; 86(1): 352 - 353. [Full Text] [PDF]
|
|
|
|
 |
|
 B. Alsoufi, S. Cai, W. G. Williams, J. G. Coles, C. A. Caldarone, A. M. Redington, and G. S. Van Arsdell Improved results with single-stage total correction of Taussig-Bing anomaly Eur. J. Cardiothorac. Surg., February 1, 2008; 33(2): 244 - 250. [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 |
