| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2005;79:1650-1655
© 2005 The Society of Thoracic Surgeons
a Section of Pediatric Cardiology, Department of Pediatrics, Baylor College of Medicine, Houston, Texas
b Epidemiology Center, Department of Pediatrics, Baylor College of Medicine, Houston, Texas
c Division of Congenital Heart Surgery, Michael E. DeBakey Department of Surgery, Baylor College of Medicine, Houston, Texas
Accepted for publication November 17, 2004.
* Address reprint requests to Dr Mott, Lillie Frank Abercrombie Section of Pediatric Cardiology, Texas Children's Hospital, 6621 Fannin MC 19345-C, Houston, TX 77030 (E-mail: amott{at}bcm.tmc.edu).
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentReferences |
|---|
METHODS: Neonates (younger than 30 days) with isolated COA or COA with aortic arch hypoplasia were identified. Retrospective review was performed to identify and characterize patients who presented with cardiogenic shock, defined as impaired left ventricular (LV) or right ventricular (RV) systolic function, or both, respiratory failure requiring tracheal intubation, and metabolic acidosis.
RESULTS: Thirteen neonates presented in cardiogenic shock and underwent surgical repair. No patients required catheter or surgical reintervention for recoarctation. There were no deaths at a mean follow-up of 54 months. Group I neonates (isolated COA, n = 7) underwent end-to-end anastomosis through left thoracotomy. The mean age and pH at presentation were 9 (±1.1) days and 7.07 (±0.21), respectively. The mean preoperative and postoperative LV myocardial performance indices (MPI) were 0.81 (±0.22) and 0.37 (±0.16), respectively (p = 0.002). Group II neonates (COA with arch hypoplasia ± ventricular septal defect, n = 6) underwent aortic arch advancement and ventricular septal defect closure through median sternotomy. The mean time from diagnosis to surgery in group II was 5.5 (±1.9) days versus 2.4 (±1.5) days in group 1 (p = 0.01). The mean age and pH at presentation were 11.8 (±9.3) days and 7.02 (±0.21), respectively. The mean preoperative and postoperative LV MPI were 0.46 (±0.13) and 0.35 (±0.11), respectively (p = 0.02). The total hospital length of stay in group II patients was 18 (±6.23) days versus 11.3 (±5. 7) days in group I (p = 0.04).
CONCLUSIONS: Timely intervention with a strategy individualized to the patient anatomy can be performed with excellent outcomes in neonates with COA and cardiogenic shock. Neonates with isolated COA had worse preoperative LV MPI, which reflects more significant global left ventricular systolic dysfunction in this subgroup. The elapsed time from diagnosis to surgery was decreased in neonates with isolated COA.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
Controversies remain regarding the most appropriate timing of surgical intervention and the optimal operative technique to repair COA [6]. Direct anastomosis, commonly employed as an extended end-to-end repair, is a popular strategy for the neonate with isolated COA [2–4, 7]. The optimal surgical strategy for infants with COA, aortic arch hypoplasia, and VSD has yet to be universally accepted. A one-stage repair consisting of aortic arch reconstruction and VSD closure had been adopted with success at some centers, including our own [3, 8–10]. Alternative strategies employ a two-stage repair, consisting of aortic reconstruction with pulmonary artery band placement at the initial operation followed by a second operation at which time the VSD is closed and the pulmonary artery band is removed [11–13].
Among all patients with COA, the surgical outcome in subgroups with associated comorbidities such as prematurity and low or very low birth weight has been documented [14, 15]. Less information is available concerning the perioperative management and outcomes in the subgroup of neonates with COA presenting with cardiogenic shock. The purpose of this study is to review and discuss our institutional strategy and outcomes with this subgroup of neonates.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
We identified 97 neonates with variants of coarctation who underwent cardiac surgery, of whom 13 (13%) presented in cardiogenic shock. Cardiogenic shock was defined as having impaired right ventricular (RV) or left ventricular (LV) function, or both, respiratory failure requiring mechanical ventilation, and metabolic acidosis (pH 
7.30). Cardiac anatomic inclusion criteria included neonates with isolated COA, or COA with arch hypoplasia and the following associated defects: VSD, patient foramen ovale, atrial septal defects, and patent ductus arteriosus. Neonates with more complicated intracardiac defects and more complex forms of congenital heart disease (hypoplastic left heart syndrome, critical aortic valve stenosis, Shone syndrome) were excluded from analysis.
All relevant perioperative and operative data were collected and recorded on a Microsoft Excel spreadsheet (Microsoft Corporation, Redmond, Washington). The immediate preoperative and postoperative echocardiograms were obtained and reviewed by one cardiologist (B.W.E.) who was blinded to patient outcomes. Echocardiographic data included LV and RV shortening fraction (SF), aortic valve morphology and function, aortic arch dimensions, LV shortening fraction, and Doppler quantification of global ventricular function as assessed by the LV and RV MPI. Recurrent coarctation was defined as an echocardiographic Doppler peak velocity measurement across the region of coarctation repair greater than 2.0 m/s and right arm to leg systolic blood pressure gradient greater than 20 mm Hg.
Patients were divided into two groups based on operative strategy as was dictated by patient anatomy: group I had isolated COA and underwent end-to-end repair through thoracotomy, whereas group II patients were judged to have severe aortic arch hypoplasia that required aortic arch advancement and VSD closure through median sternotomy.
All data are presented as the mean ± standard deviation, and when relevant, additional ranges are given. Inferential statistics were performed using the SAS statistical package (SAS Institute, Cary, North Carolina) for analysis. Categorical variables were compared with the 
2 test. Analyses using continuous variables were conducted with the Student t test or the Wilcoxon rank sum test when there were concerns about the data not following a normal distribution. Statistical significance was defined as p less than or equal to 0.05.
Operative Technique
Upon induction, all patients undergo routine preoperative transesophageal echocardiography and placement of cerebral oxygenation monitors in the form of near-infrared cerebral cortical spectroscopy (NIRS [INVOS 5100; Somantics Corporation, Troy, Michigan]) and middle cerebral artery transcranial Doppler monitoring (EME Companion; Nicolet Biomedical, Madison, Wisconsin).
Group I neonates underwent a muscle-sparing left posterolateral thoracotomy through the third or fourth intercostal space with the proximal aortic cross-clamp applied between the left common carotid and left subclavian artery in each case. Coarctectomy, complete resection of all ductal tissue, and extended end-to-end anastomosis was performed using a clamp and sew technique.
As significant aortic arch hypoplasia accompanied aortic coarctation in group II neonates, complete relief of obstruction was achieved through median sternotomy, establishment of cardiopulmonary bypass (CPB) and aortic arch advancement. Single-stage repair of any additional defects in the atrial or ventricular septum is routine in this group, as is the placement of a peritoneal dialysis catheter.
Our use of aortic arch advancement has been previously described [9] and consists of an end-descending aorta to side-ascending aortic anastomosis, thereby eliminating any potential obstruction at the level of the aortic arch. Phenoxybenzamine is routinely utilized for patients lessw than 10 kg undergoing CPB.
While patients previously required deep hypothermic circulatory arrest (DHCA) for aortic arch advancement, this has been avoided in the most recent patients by the use of regional low-flow cerebral perfusion (RLFP). The RLFP is typically provided through a 3.5 mm Gore-Tex shunt anastomosed to the innominate artery at the initiation of the operation. We begin full-flow CPB at a calculated baseline of 150 mL · kg–1 · min–1 and, after snare placement on the proximal brachiocephalic vessels, initiate RLFP by reducing pump flow to 50% of baseline. We make further adjustments such that baseline cerebral blood flow velocity as measured by transcranial Doppler and cerebral oximetrics as measured by NIRS are optimally maintained. We have previously studied and described the precise use of these monitors to guide RLFP [16–18].
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
Each patient required tracheal intubation and placement of central venous catheters. A prostaglandin E1 infusion was instituted in each patient. Sodium bicarbonate was used to treat metabolic acidosis. The following vasoactive medication infusions were instituted: dopamine (n = 4), dobutamine (n = 3), or dopamine/dobutamine (n = 6). No patient required cardiopulmonary resuscitation or underwent cardiac catheterization before operation.
The overall mean age and weight at presentation were 10.3 days (range, 3 to 29) and 3.3 kg (range, 2.5 to 4.8). The mean elapsed time from diagnosis to surgery was 3.8 days (range, 1 to 8) and 1.3 days in the latter 4 patients in the series. Preoperative resuscitation was successfully performed in all patients, and no patient required emergency rescue operation.
Group I
There were 7 neonates (5 male and 2 female) with isolated COA (Table 1). The mean age at presentation was 9.0 (±1.1) days (range, 7 to 10) and the mean weight was 3.8 kg (±0.6 kg). The mean arterial pH at diagnosis was 7.07 (±0.21; range, 6.82 to 7.30). The mean age at surgery was 11.6 (±1.9) days (range, 9 to 14), and the mean elapsed time from diagnosis to surgery was 2.4 (±1.5) days (range, 0.5 to 4).
|
Analysis of echocardiograms (Table 2) revealed a tricommissural aortic valve in 1 patient (14%) and bicommissural in 6 patients (86%). One neonate had a very small VSD that did not require surgical intervention. At the time of diagnosis, the ductus was small, yet patent in 3 of 7 patients. After initiation of prostaglandin E1 and immediately before surgery, the ductus was patent in 4 of 7 patients.
|
The average cross-clamp time was 21.2 (±3.1) minutes (range, 18 to 25). A pericardial effusion was noted in 1 patient and was drained at the time of surgery. Invasive blood pressure monitoring in the upper and lower extremity compartments revealed no residual gradient in any patient at the conclusion of the operation.
The mean duration of postoperative mechanical ventilation and cardiovascular intensive care unit lengths of stay were 62.3 (±59.3) hours (range, 24.0 to 192.0) and 97.9 (±65.0) hours (range, 44 to 240), respectively. The mean duration of postoperative hospital length of stay was 9.1 (±5.5) days (range, 4 to 20). No patient has met criteria for recurrent arch obstruction. There were no early or late deaths, and no patient required intervention at a mean of 62 months postoperatively.
Group II
There were 6 neonates (3 male and 3 female) diagnosed with COA with aortic arch hypoplasia (Table 1). Five of six neonates had an associated VSD. The mean age at presentation was 11.8 (±9.3) days (range, 3 to 29), and the mean weight was 3.2 kg (±0.6 kg; range, 2.5 to 4.3). The mean arterial pH at diagnosis was 7.02 (±0.21; range, 6.82 to 7.30). The mean age at surgery was 16.8 (±9.7) days (range, 9 to 34) and the mean elapsed time from diagnosis to surgery was 5.5 (±1.9) days (range, 3 to 8).
The mean serum creatinine was 1.02 (±0.74; range, 0.3 to 2.4) at diagnosis, 0.72 (±0.50; range, 0.3 to 1.6) immediately before surgery, and 0.45 (±0.19; range, 0.3 to 0.8) at hospital discharge. No patient had clinical evidence of seizure activity before or after surgery.
Analysis of echocardiograms (Table 2) revealed a trabecular VSD in 4 patients (83%) and perimembranous VSD in 1 patient (17%). The aortic valve was bicommissural in 5 patients (83%) and tricommissural in 1 patient (17%). At the time of diagnosis, after institution of prostaglandin E1, and immediately before surgery, the ductus was small, yet patent, in 4 of 6 patients.
The mean preoperative and immediate postoperative LV SF was 29.6% (±4.3%; range, 26% to 36%) and 41.9% (±3.0%; range, 38% to 43%), respectively. The mean preoperative and postoperative LV MPI were 0.46 (±0.13) and 0.35 (±0.11), respectively. However, the mean preoperative and postoperative RV MPI were 0.30 (±0.09) and 0.23 (±0.06), respectively (p = 0.02).
The mean CPB time was 145.8 (±27.7) minutes (range, 110 to 179). Four patients underwent DHCA for 36.8 (±28.5) minutes (range, 18 to 61) while DHCA was avoided in the 2 most recent patients by the use of RLFP. As in group I neonates, invasive blood pressure monitoring in the upper and lower extremity compartments revealed no residual gradient in any patient at the conclusion of the operation.
The mean duration of postoperative mechanical ventilation and intensive care unit length of stay was 65.2 (±26.8) hours (range, 36 to 96) and 155.2 (±48.6) hours (range, 96 to 211), respectively. The mean duration of postoperative hospital length of stay was 12.5 (±4.9) days (range, 8 to 21). No patient has met criteria for recurrent aortic arch obstruction. However, 1 patient has a documented upper to lower extremity blood pressure gradient of 18 mm Hg and peak velocity in the descending aorta of 2.1 m/s. There were no early or late deaths, and no patient has required surgical intervention for recurrent aortic coarctation at a mean of 44 months postoperatively.
Group I Versus Group II
In group I versus group II (Table 1), there were no significant differences in the age at diagnosis, weight, pH, or serum creatinine at diagnosis. The elapsed time from diagnosis to surgery was, however, longer for group II (5.5 days) versus group 1 (2.3 days; p = 0.01).
Analysis of postoperative resource utilization (Table 3) demonstrated no difference in duration of mechanical ventilation between groups, whereas group II patients had a longer cardiovascular intensive care unit stay 97.9 (±65.1) hours versus 155.2 (±48.6) hours (p = 0.04). Although there was no difference in postoperative hospital length of stay, group II patients had an overall longer hospital stay secondary to longer preoperative resuscitation 18.0 (±6.3) days versus 11.3 (±5. 7) in group I (p = 0.04).
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
Even in the current era, there remain significant controversies with regard to timing and the optimal surgical strategy. To prevent early recoarctation, the ideal operative procedure must successfully address transverse arch hypoplasia (if present), resection of all ductal tissue, and prevention of residual circumferential scarring at the aortic anastomotic site [20]. It is important to consider that regardless of surgical technique, younger age at operation, and the presence of aortic arch hypoplasia remain risk factors for recoarctation [2, 4].
End-to-end anastomosis or a modification thereof is a common surgical strategy for the neonate with isolated COA [2–4, 7, 8]. Acceptable results have been reported after modifications of subclavian flap angioplasty [7, 13, 21]. Younoszai and associates [22] have reported an impressive 3% recoarctation rate while utilizing an end descending to side transverse arch anastomosis through left thoracotomy in neonates and infants with COA and arch hypoplasia. Although this technique is notable for not requiring CPB and DHCA, it does not fully eliminate the possibility of recurrent coarctation at the level of the aortic arch. We have reported that in 18 patients requiring operative intervention for recurrent coarctation, all had some element of residual aortic arch hypoplasia, and we therefore continue to advocate aortic arch advancement as the optimal surgical strategy for neonatal COA with severe arch hypoplasia [23].
A one-stage surgical strategy of aortic reconstruction and VSD closure can be successfully employed for neonates with COA/aortic arch hypoplasia and VSD [3, 9, 10]. As an alternative surgical strategy, a two-stage repair strategy with repair of the COA/pulmonary artery band placement as an initial intervention and followed by VSD closure [11–13].
As a primary focus, we report the surgical outcome of 13 neonates with COA who presented with cardiogenic shock. Our results are encouraging and compare favorably to contemporary reported outcomes. The operative strategy at our institution is not uniform in that the operation is tailored to the demands of the patient's anatomy; end-to-end repair for isolated COA and a one-stage repair with aortic arch advancement and VSD closure for aortic arch hypoplasia and VSD. Perioperative care of these neonates is demanding, requiring vigorous resuscitation and simultaneous complete cardiac evaluation and assessment of end-organ function.
We defined adequate resuscitation of cardiogenic shock at out center is achieved when there is hemodynamic stability and resolution of metabolic acidosis. In this series, 10 of 13 neonates were initially evaluated at outside institutions where arterial pH was more consistently monitored than were serum lactate levels. As such, tracking and using a serum lactate level as a marker of adequate resuscitation was not possible in this series. There is also a concerted effort to proceed with surgery in a timely manner in these patients, although our experience demonstrated that complete resuscitation was possible with no patient requiring an emergent "rescue" operation.
A secondary focus of this report is the discussion of two emerging strategies—preoperative and postoperative echocardiographic myocardial performance index and use of intraoperative noninvasive neuromonitoring and neuroprotective strategies in this unique subset of patients.
Echocardiographic assessment of ventricular function has historically been based on models that assume a specific geometric shape. Accurate quantitative assessment of ventricular function in neonates and children is often challenging owing to abnormal ventricular geometry present in congenital heart disease. Routine echocardiographic assessment of LV systolic function has relied upon radial measures of LV performance such as ejection fraction or shortening fraction. Assessment of RV function or LV function in children with abnormal ventricular geometry often relies on a qualitative visual assessment of systolic performance.
The MPI is a simple, reproducible, noninvasive echocardiographic measure of combined systolic and diastolic ventricular function. The MPI is a ratio between the sum of the isovolumic contraction time (ICT) and isovolumic relaxation time (IRT) divided by ventricular ejection time. The quantitation of ventricular function by the MPI is unique in that it is a ratio of Doppler-derived time intervals and is therefore not limited by geometric shape of the ventricle. With worsening systolic dysfunction, isovolumic contraction time lengthens and ejection time shortens, producing an increased MPI. Similarly, with worsening diastolic dysfunction, isovolumic relaxation time lengthens leading to an increased MPI.
This index has been shown to correlate well with other invasive and noninvasive measures of LV function in both children and adults [24]. More recent studies have demonstrated the ability of the MPI to predict clinical outcome in patients with dilated cardiomyopathy, pulmonary hypertension, congestive heart failure, and acute myocardial infarction [24]. The impact of loading conditions on the MPI has also been recently reported in pediatric and adult studies [24].
Preoperatively, this index was significantly increased in neonates with intact ventricular septum, consistent with significant global systolic and diastolic LV dysfunction. These patients had significantly higher LV MPI values compared with those neonates with an associated VSD. Both groups demonstrated normalization of the LV MPI in the immediate postoperative period, despite smaller changes in LV SF. This most likely reflects improvement in both LV systolic and diastolic function with relief of significant LV afterload. Of note, neonates in both groups also had evidence of RV dysfunction (as quantified by RV MPI) with improved postoperative function. This is likely due to improved postoperative septal geometry or decreased left ventricular end-diastolic and left atrial pressure with decreased pulmonary artery pressure and RV afterload.
Neonates with severe COA are at risk for preoperative, intraoperative, and postoperative neurologic injury and such injury (while difficult to quantitate) can impact short-term and long-term functional outcome. Our center is an avid proponent of routine intraoperative physiologic cerebral monitoring (NIRS and transcranial Doppler monitoring) for patients undergoing open-heart surgery [16–18]. Although no definitive evidence of neurologic protection has yet been demonstrated, we believe this monitoring may be especially useful for guiding RLFP during aortic arch reconstruction by preventing overcirculation or undercirculation.
Based on mounting evidence from Wypij and colleagues [25] regarding the deleterious effects of DHCA on the developing nervous system, we have adopted RLFP during neonatal aortic arch reconstruction to limit exposure to DHCA. We and Pigula and colleagues have demonstrated that RLFP provides consistent cerebral circulatory support and that this support is bilateral, despite being applied to the inominate artery [16–18, 26].
Pigula and colleagues [27] have, furthermore, demonstrated that RLFP provides a nutrient somatic perfusion (as measured by NIRS) and it is unknown if this provides a clinical benefit during aortic arch reconstruction. Further investigation is necessary to define the role of RLFP in ameliorating the long-term derogatory effects known to occur with prolonged DHCA time.
Conclusion
Neonates with COA who present with profound cardiogenic shock present a unique medical and surgical challenge. A surgical strategy that is tailored to the demands of the individual patient's cardiac anatomy has contributed to excellent outcomes. Patients with COA with intact ventricular septum had worse perioperative global LV function (as quantified by the MPI) and had shorter elapsed time from diagnosis to surgery compared with neonates with COA/VSD. Postoperative MPI detected immediately improved ventricular function in all patients, and the improved function continued at long-term follow-up.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Tabbutt, S. C. Nicolson, T. E. Dominguez, W. Wells, C. L. Backer, J. S. Tweddell, P. Bokesch, and M. Schreiner Perioperative course in 118 infants and children undergoing coarctation repair via a thoracotomy: A prospective, multicenter experience. J. Thorac. Cardiovasc. Surg., November 1, 2008; 136(5): 1229 - 1236. [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 |
