| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Ann Thorac Surg 2006;82:1577-1584
© 2006 The Society of Thoracic Surgeons
a Division of Congenital Heart Surgery, Michael E. DeBakey Department of Surgery, Houston, Texas
c Department of Physical Medicine and Rehabilitation, Baylor College of Medicine, Houston, Texas
b Division of Congenital Heart Surgery, Texas Children's Hospital, Houston, Texas
Accepted for publication May 18, 2006.
* Address correspondence to Dr Morales, Division of Congenital Heart Surgery Texas Children's Hospital, 6621 Fannin St, MC-WT 19345H, Houston, TX 77030 (Email: dlmorale{at}texaschildrenshospital.org).
Presented at the Forty-second Annual Meeting of The Society of Thoracic Surgeons, Chicago IL, Jan 30–Feb 1, 2006.
| Pediatric cardiac surgery:
The Annals of Thoracic Surgery CME Program is located online at http://cme.ctsnetjournals.org. To take the CME activity related to this article, you must have either an STS member or an individual non-member subscription to the journal.
|
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
METHODS: Between July 1995 and December 2005, 60 patients underwent IAA repair using aortic arch advancement without a patch. Selective cerebral perfusion was used in 25 patients (42%). Cox proportional hazards models were used to analyze 20 variables to determine risk factors for death, arch reintervention, and left ventricular outflow tract (LVOT) reintervention.
RESULTS: Median age was 8 days (range, 2 to 271 days) and weight was 3.0 kg (range, 1.7 to 6.1 kg). IAA types were A in 18 (30%) and B in 42 (70%). Associated anomalies were multiple congenital anomalies in 30 (50%) patients, DiGeorge syndrome in 21 (35%), LVOT obstruction in 26 (43%), a single ventricle in 11 (18%), and truncus arteriosus in 6 (10%). Mean follow-up was 3.0 ± 2.6 years. Five-year freedom from aortic arch reintervention was 100%. Survival at 30 days, 1 year, and 5 years was 93%, 78%, and 76%, respectively. Since July 2000, two of 32 patients have died for an overall survival of 94%. Risk factors for death are older age, multiple congenital anomalies, DiGeorge syndrome, and bicuspid aortic valve. Selective cerebral perfusion was an independent protective variable for survival. Survival for an IAA patient with a ventricular septal defect and no complicating cardiac anomalies was 100%.
CONCLUSIONS: Aortic arch advancement without a patch can be applied to IAA patients, with the expectation of a minimal need for arch reintervention. This technique affords an excellent survival, to which selective cerebral perfusion may be a contributing factor.
Interrupted aortic arch (IAA) is a rare and severe congenital cardiac lesion that is uniformly fatal without surgical therapy. Advances in surgical technique and in anesthetic and perioperative care have resulted in continued improvements in the outcome of complete neonatal repair of this defect and its associated lesions [1–3]. The surgical repair of IAA remains challenging, however, and is associated with significant early mortality and incidence of late aortic arch obstruction or need for left ventricular outflow tract (LVOT) reintervention [4, 5].
The challenges of IAA repair were well documented by the Congenital Heart Surgeons' Society (CHSS) [1]. In particular, its recent report demonstrated that repair of the aortic arch by a method other than direct anastomosis with patch augmentation was associated with an increased risk of intervention for residual or recurrent arch obstruction. Other reports have also suggested that a direct anastomotic repair without a patch for IAA leads to arch reobstruction more often than other techniques [1, 2]. Because almost all of the IAA repairs at Texas Children's Hospital (TCH) in the past decade have been done with a direct anastomosis without using a patch, it seemed prudent to review the experience at TCH and, in particular, the rate of aortic arch reintervention.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Aortic arch advancement was guided by the principles of (1) full circumferential mobilization of the descending thoracic aorta well beyond the left bronchus, (2) aggressive resection of all ductal tissue, (3) minimization of deep hypothermic circulatory arrest (DHCA) by use of selective cerebral perfusion (SCP), and (4) anastomosis of the descending aorta to the posterior aspect of the distal ascending aorta with extension into an arch vessel, at times to ensure a wide anastomosis. The result is illustrated in Figure 1.
|
Follow-up was at the discretion of the referring cardiologist. The decision to reintervene on the aortic arch was multifactorial and determined from diagnostic and clinical findings. An instantaneous peak velocity in the aortic arch of more than 2 m/s indicated a potential arch restenosis. If, however, there was no diastolic run-off in the descending aorta by echocardiography, no extremity blood pressure gradient, and no symptoms, then follow-up on these patients was continued. If echocardiography of the aortic arch raised a question about its anatomy or blood flow, then magnetic resonance imaging was done, followed by a catheterization if necessary for diagnostic or therapeutic purposes.
Statistical Analysis
Nonparametric, binomial, and descriptive statistics were computed when appropriate. Data are described as ratios, medians with ranges, and means with standard deviations. The Fisher exact test and the 
2 test were used to analyze binary variables. Kaplan-Meier survival analyses plotted survival curves from the follow-up data. For these analyses, children were censored at the time of death or were withdrawn alive at the point of last contact. Survival time was defined as age in days with date of surgery taken to be time zero.
Multivariable analyses were performed using Cox proportional hazard models. Twenty variables entered the models independently and not stepwise. They were specified and tested for significance as risk factors for mortality, aortic arch reintervention, and LVOT reintervention. Continuous variables tested were age at operation, weight, length of DHCA, and cross-clamp time. Discontinuous variables were SCP, DHCA, presence of multiple congenital anomalies, either defined syndromes (ie, DiGeorge) or a constellation of multiple congenital defects with no specific diagnosis as assessed by the genetics department, DiGeorge syndrome, type A IAA, type B IAA, ventricular septal defect, bicuspid aortic valve, anomalous subclavian artery, LVOT obstruction, truncus arteriosus, aortopulmonary window, single ventricle, subaortic resection, episode of preoperative cardiogenic shock (hemodynamic instability requiring resuscitation), and emergent status (going to operating room within 24 hours of presentation). All analyses were conducted with SPSS 13.0 (SPSS, Inc, Chicago, IL).
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
DHCA was the primary bypass strategy in 57% (33/58) of the patients, with an average time of 43.9 ± 23.5 minutes. SCP was first used in January 2001 and since then has been the primary bypass strategy in 81% (25/31) of the patients, with an average time of 56 ± 23 minutes. Average DHCA in these patients was 11.8 ± 15.8 minutes. Mean cardiopulmonary bypass time overall was 184 ± 60 minutes. Mean cross-clamp time overall was 108 ± 47 minutes. Two patients with IAA type A underwent repair via left thoracotomy without cardiopulmonary bypass.
Concomitant procedures included ventricular septal defect closure in 40 patients (67%), truncus repair in 6 (10%), subaortic resection in 18 (30%), pulmonary artery band in 7 (12%), Ross-Konno in 2 (3%), DKS (Damus-Kaye-Stansel) in 4 (7%), repair of Taussig-Bing heart in 2 (3%), aortopulmonary window repair in 3 (5%), and modified Norwood (no patch) in 3 (5%). LVOT obstruction was addressed at the time of primary repair in 43% (26/60) of patients. There was no postoperative heart block. One-stage IAA repair was done in 45 (92%) of the 49 two-ventricle patients. Figure 2 illustrates the management and outcomes of all 60 patients.
|
|
There were no reinterventions (ie, catheter or operative) on the 60 aortic arch repairs using the AAA technique, with a maximal follow-up of 9.5 years. Also, there were no reinterventions for bronchial compression. Five patients have had an instantaneous peak velocity of more than 2 m/s by echo across the aortic arch and only 2 with more than 2.5 m/s. However, none have had an upper-to-lower BP gradient exceeding 5 mm Hg, symptoms, or diastolic run-off by echo in the descending aorta. Therefore, none have been put forth for catheterization or magnetic resonance imaging. Actuarial freedom from reintervention on the aortic arch is 100% at 5 years.
The freedom from reintervention on the LVOT at 30 days, 1 year, 3 years, and 5 years is 100%, 77%, 74%, and 66%, respectively (Fig 4). There were 12 reinterventions on the LVOT overall, with 83% (10/12) occurring in the first postoperative year. Half (6/12) of the patients who required LVOT reintervention had LVOT obstruction surgically addressed at their initial IAA repair (Table 1). The hospital survival for LVOT reintervention was 92% (11/12), with no late deaths. The risk of LVOT reintervention increased with the age (p = 0.021) at the initial surgery. The diagnoses of LVOT obstruction (p = 0.047) and anomalous subclavian artery (p = 0.040) were risk factors for LVOT reintervention.
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
The overall survival for the entire cohort in the current study compares favorably with the recent large single and multiinstitutional series that have reported 1-year survival of 35% to 74% and 5-year survivals of 34% to 61% for all patients with IAA [1, 2, 7]. Era-related improvement was a common feature in all the recent series, as it was in the TCH experience, where overall survival in the second half of the series was 94% regardless of diagnoses, which was significantly greater than survival in the first half of the series [1, 3].
The present series differed from much of the literature in that truncus arteriosus, LVOT obstruction, single ventricle, and other complicating cardiac diagnoses were not risk factors for death [5, 8, 9]. Others have reported younger age at surgery to be a risk factor for death, but in contrast, this series found mortality to increase with age as a continuous variable [7]. At TCH, most IAA patients undergo repair in the first few days of life. Patients with delay in diagnosis and transfer to TCH, or who have an episode of shock or other noncardiac issues (ie, necrotizing enterocolitis or aspiration), are older at initial repair, allowing for stabilization and recovery of end-organ function. Therefore, the older the patient at time of operation, the more likely it is that complicating issues will negatively impact mortality and late morbidity (ie, reintervention rate).
Sixty-two percent of all the deaths were in patients that had DiGeorge syndrome or multiple congenital anomalies. Not highlighted in most series, the impact of chromosomal abnormalities manifesting as metabolic or gastrointestinal issues can make an already challenging postoperative course dire. Bicuspid aortic valve was also a risk factor for death and in fact accounted for 50% (3/6) of the late deaths. Two of these patients died suddenly at home, and the other died after late LVOT reintervention. The well-accepted fact that bicuspid aortic valves are more likely to become stenotic, even though adequate during the neonatal period, may lead to increased late morbidity and thus mortality in the IAA cohort.
Interestingly, SCP had a protective influence on overall mortality. It allowed for a 73% reduction in the average DHCA time, from 44 to 12 minutes. This bypass strategy may help not only with cerebral protection but may also help protect the viscera through nutritive flow [6, 10]. SCP allows the surgeon to perform complex procedures while not under the constraints of DHCA time. This may help explain why the different complex diagnoses, despite getting a one-stage repair, were not risk factors for death. However, one must consider that because SCP was used almost exclusively in the latter part of this series, its effect is confounded by era-related improvement.
Many surgical techniques have been described for the correction of IAA, with most series reporting a variety of techniques within one institutional experience. Because cohorts in these series usually contain fewer than 70 patients, interpretation of any one surgical technique is difficult. To overcome this problem, the CHSS was able to collect data from more than 30 institutions to analyze and report on many issues that are difficult to answer in single institutional series. In regards to best surgical technique to lower incidence of aortic arch reobstruction, the study concluded that repair of interrupted aortic arch by a method other than direct anastomosis with patch augmentation is a risk factor for reintervention on the aortic arch [1].
The TCH experience is distinct in that it has used one technique (direct anastomosis without patch augmentation) for more than a decade in 97% of IAA repairs. The analysis of this experience has demonstrated no need for catheter or surgical reintervention on the aortic arch for obstruction, which would imply that the technique of aortic arch advancement is not a risk factor for aortic arch reobstruction or reintervention. However, this technique requires aggressive ductal resection and circumferential mobilization of the descending thoracic aorta well beyond the left bronchus. This is the preferred technique at TCH, but clearly, other institutions apply other techniques with comparable results. The current experience simply concludes that aortic arch advancement can be applied in patients with IAA, with the expectation of minimal need for reintervention for aortic arch obstruction.
The most challenging aspect of these patients' care is often the decision-making about whether LVOT obstruction needs to be addressed, in what manner, and optimal timing. Several studies have tried to set standards for the adequacy of the LVOT by using preoperative echocardiography results such as aortic annulus size or cross-sectional area of the LVOT, or both [4, 11, 12]. Unfortunately, application of these variables has not been consistently beneficial in this decision-making, probably because of the heterogeneity in the morphology of LVOT obstruction [4, 13].
The TCH approach to IAA with LVOT obstruction is an individualized strategy emphasizing perioperative transesophageal echocardiography and intraoperative inspection of the various components of the LVOT. First, the size of the LVOT is assessed with a 5-mm Hegar dilator. If the Hegar passes through the aortic annulus into the systemic ventricle, then the LVOT is probably adequate, at least for the neonatal period. However, the morphology and complexity of the LVOT is also examined, and even if the LVOT is adequate by Hegar dilator, the LVOT's morphology may require intervention, usually by subaortic muscle resection via the VSD. This technique is applied to avoid neonatal aortic root replacement in hope that adequate aortic valve growth can avert or delay aortic root replacement. Thus far, 72% (13/18) of the cohort that had subaortic resections has avoided aortic root replacement. Creation of a Damus-Kaye-Stansel or Norwood type of connection to address LVOT obstruction with reversal to a biventricular repair (Fig 2) is technically challenging but a viable management strategy [14, 15].
Anomalous take off of the subclavian artery, usually associated with type B IAA, was found to be a risk factor for LVOT obstruction, as it has been in other series [4]. Patients with anomalous subclavian arteries have even less LVOT antegrade flow than in other type B IAA patients because the flow is only supplying the right and left carotid arteries. This, in addition to the decreased flow through the LVOT secondary to intracardiac shunting and ductal patency may further increase the severity of LVOT underdevelopment. Also, this decreased flow preoperatively does not allow for accurate preoperative assessment of the gradient across the LVOT, which can be unmasked after repair and lead to reintervention.
The preoperative diagnosis of LVOT obstruction was a risk factor for reintervention, as might be expected. The attempt to repair or palliate LVOT obstruction with subaortic resection, valvuloplasty, or by the tolerance of mild obstruction to avoid infant aortic root replacement is not always successful and will at times require LVOT reintervention. However, this strategy has allowed 15 (79%) of 19 patients with the preoperative diagnosis of LVOT obstruction, who did not undergo aortic root replacement at initial IAA repair, to avoid aortic root replacement thus far in the current series.
Mortality after interrupted aortic arch repair has continued to improve, and the negative influence of complex cardiac diagnoses such as truncus arteriosus, single ventricle, and LVOT obstruction on mortality has diminished. Reintervention on the LVOT continues to dominate the late morbidity of this patient cohort, whereas reintervention for postoperative aortic arch obstruction has been minimized by the application of such techniques as aortic arch advancement without a patch.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
And also, could you comment, your reintervention rate was also lower than our study, could you tell us what your current indications are for LVOT reintervention?
DR MORALES: Age was analyzed with an adjustment for era. Also, the mean age of the patients in the earlier era was not different than those in the later era. I think the reason that age is a risk factor is because most of the patients go to surgery without delay, unless there is some complicating diagnosis, such as NEC or aspiration. Therefore, older patients probably did worse because they went to surgery later secondary to complications.
And in terms of left ventricular outflow tract obstruction, I think we look at each of the patients individually. We do not use any specific preoperative echo criteria such as minimal aortic annulus or LVOT cross-sectional area. We do use a preoperative TEE to gauge the morphology, or the complexity, of the left ventricular outflow tract. Intraoperatively, we use Hegar dilators to measure the aortic annulus.
If the annulus is over 5 mm and we did not think the subaortic area was complex, then we usually do not intervene on the LVOT. But even if the aortic annulus is 5 mm, if we think there is a complex subaortic area, we will perform a subaortic resection. This explains our high incidence of subaortic resection, consisting of nearly a third of our patients.
DR CARLO F. MARCELLETTI (Palermo, Italy): I would like you to expand a little bit more on the complete strategy in hearts with one-ventricle physiology associated with interrupted arch. Do you do banding, which would be very risky in such hearts, which normally have subaortic obstruction, or do you do a shunt, or what?
DR MORALES: We do not tend to band, unless of course they have unrestricted pulmonary blood flow.
DR MARCELLETTI: So what do you do, you repair the arch?
DR MORALES: We repair the arch and assess the amount of pulmonary blood flow, which may often require a band and rarely, a BT shunt.
DR MARCELLETTI: But normally patients with aortic arch interruption do not have restricted pulmonary blood flow. How do you control pulmonary blood flow in these cases?
DR MORALES: Well, with 11 single ventricles, we did PA bands on 6 of these patients. However, we did not restrict blood flow in the other 5 patients, all of whom did well.
DR CHRISTOPHER KNOTT-CRAIG (Oklahoma City, OK): I have two questions. The first question is: Do you think the fact that you don't have any type C interruptions positively influenced the low risk of reintervention on the left ventricular outflow tract in your patients? The second question is: How do you deal with patients with efferent right subclavian arteries from the descending aorta in patients with interrupted aortic arch?
DR MORALES: I'll answer the second question first. We usually do not divide anomalous subclavian arteries during interrupted arch repair. With circumferential aggressive dissection of the descending aorta and the aortic arch, we have not found this to be necessary.
And in terms of the first question, I can't really comment since we did not have any patients with type C anatomy. However, I will say that type C has not been consistently found to be a risk factor in the literature for LVOT reintervention.
DR ALI DODGE-KHATAMI (Zurich, Switzerland): With this direct anastomosis, did you have any left bronchus stenosis? And if not, what did you do to avoid it? And if you did, what did you do to treat that?
DR MORALES: We have had no reinterventions for a stenosis of the left bronchus. I do not mean to state that there has been no left bronchial compression in our series, but it has been mild enough not to require intervention. And in terms of how we try to avoid bronchial compression, we again emphasize that aggressive mobilization of the descending aorta well past the left bronchus is essential, at least for our technique.
DR BOHDAN MARUSZEWSKI (Warsaw, Poland): Your deep hypothermia circulatory arrest time was about 40 minutes. Could you elaborate a little bit about the myocardial ischemia and if you ever had to also isolately perfuse the myocardium of these patients.
DR MORALES: We do not perfuse the myocardium in these patients during selective cerebral perfusion or circulatory arrest. However, during circulatory arrest, we often will reperfuse the patient for 5 minutes to break up the arrest time. So even though the total circulatory arrest time could be 40 minutes, that most likely represents two 20-minute periods, and that's how we avoid long uninterrupted periods of circulatory arrest. Also in recent years, the length of our circulatory arrest and reperfusion periods are guided by near infrared spectroscopy.
DR JOSEPH M. FORBESS (Dallas, TX): Along those lines, I just wanted to ask you an additional question, something that you haven't touched on. The circulatory arrest time for a typical interrupted aortic arch, to do the arch, would be, say, 15 to 20 minutes, in that range?
DR MORALES: Yes.
DR FORBESS: Am I to understand, is the VSD done on circulatory arrest?
DR MORALES: No, it is not.
DR FORBESS: Talk to me a little bit about, or tell the audience, a little bit about, how the VSD is approached at Texas Children's.
DR MORALES: Well, we bicavally cannulate almost all patients with VSDs, to avoid circulatory arrest. However, this practice was less consistent in the beginning of our series.
DR FORBESS: So if you do that, the selective cerebral perfusion is to save the 15 to 20 minutes of circulatory arrest?
DR MORALES: Yes it is. Although, on average most agree that over 35 minutes of circulatory arrest is bad, I do not think anyone feels that circulatory arrest is good for the patient. Thus, if I can avoid it easily, I will. Also, if we're doing a more complex procedure, such as some type of modified Norwood, selective cerebral perfusion will save the patient from a significant amount of circulatory time.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  W. K. Jhang, J.-J. Park, D.-M. Seo, H. W. Goo, and M. Gwak Perioperative evaluation of airways in patients with arch obstruction and intracardiac defects. Ann. Thorac. Surg., May 1, 2008; 85(5): 1753 - 1758. [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 |
