| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2007;84:142-146
© 2007 The Society of Thoracic Surgeons
Department of Cardiac Surgery, The Diana Princess of Wales, Birmingham Children‘s Hospital, Birmingham, United Kingdom
Accepted for publication February 5, 2007.
* Address correspondence to Dr Barron, Department of Cardiac Surgery, Diana Princess of Wales, Birmingham Children‘s Hospital NHS Trust, Steel house Lane, Birmingham B4 6NH, United Kingdom (Email: david.barron{at}bch.nhs.uk).
| 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 MethodsResultsCommentAcknowledgmentsReferences |
|---|
Methods: A review of 16 patients with aortic atresia or complex left ventricular outflow tract obstruction who underwent BVR between 1990 and 2005; a comparison with outcomes for the Norwood I procedure over the same period.
Results: Early mortality was 19% (3 patients) with no deaths in the last 12 years (13 patients). Twelve patients had associated aortic interruption (56%) or coarctation (19%). Anatomic subtype was not a risk for early death. Late age at operation was the only risk factor identified for early death (p = 0.01). Median follow-up was 32 (range, 4 to 190) months. Actuarial survival at one and five years was 60% and 53%, respectively. This compares with an early mortality of 29% (p < 0.01) and actuarial survival of 58% and 50% in the Norwood group. Freedom from reintervention was 68% and 20% at one and five years, respectively. One patient required balloon dilatation of recurrent coarctation, all others were balloon dilatation (n = 2) or surgical (n = 4) conduit replacement. All survivors are currently in New York Heart Association class I.
Conclusions: Biventricular repair of aortic atresia and VSD can be achieved with results that compare well with univentricular palliation. Despite the need for conduit change, the long-term benefit of a BVR would support this technique. Delay in performing the initial repair may increase mortality.
| This article has been selected for the open discussion forum on the CTSNet Web Site: http://www.ctsnet.org/sections/newsandviews/discussions/index.html
|
Aortic atresia is not always associated with hypoplasia of the left heart. The term of hypoplastic left heart syndrome (HLHS) encompasses a spectrum of lesions affecting the left heart with main features of hypoplasia of the left ventricle and atresia or stenosis of inlet and outlet of the ventricle. However, 4% to 7% of patients with aortic atresia-hypoplasia have a coexisting ventricular septal defect (VSD) which may allow for the development of a normal sized mitral apparatus and left ventricle [1–4]. These patients may not have HLHS and may be amenable to biventricular repair (Fig 1). Although aortic atresia with a VSD has previously often been managed by conversion to a univentricular physiology [5], a biventricular repair (BVR) has been attempted in stages or a single stage using allografts or prosthetic conduits. The first description of a single stage repair of aortic atresia with a normal left ventricle in a neonate was published by Yasui and colleagues [6] in 1987 and by Austin and colleagues [7] in 1989. Various technical improvements have been suggested since these initial reports.
|
We have reviewed a 15-year experience with this "Damus-Rastelli" technique to define the anatomic subtypes and assess the clinical outcomes in comparison with a standard univentricular approach.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
The diagnosis of aortic atresia or hypoplasia with VSD was based on two-dimensional echocardiography. The threshold values for BVR were good left ventricular function with no evidence of fibroelastosis, mitral valve annulus z-score greater than –2, ratio of left to right ventricular long axis greater than 0.8, volume of the left ventricle greater than 20 mL/m2 and unrestrictive VSD. We defined an unrestrictive VSD as one with no Doppler gradient between left and right ventricle and a clear pathway by which the VSD could be committed through to the pulmonary artery. The comparative group was composed of 336 consecutive Norwood procedures performed at this institution for HLHS over the same time period.
Operative Technique
The technique adopted was single-stage repair involving a combination of Damus-Kaye-Stansel and Rastelli techniques (see Fig 2). All operations were performed using deep hypothermic cardiopulmonary bypass with periods of circulatory arrest for repair of arch (tailored to individual anatomy). Antegrade cerebral perfusion, introduced in September 2000, was used during arch reconstruction where possible. The VSD was closed through the ventriculotomy using a bovine pericardial patch.
|
Delayed sternal closure was employed in all patients. Inotropic support comprised of dobutamine (10 ug · kg–2 · min–2) or milrinone (0.3 to 0.7 ug · kg–2 · min–2) plus epinephrine (0 to 0.3 ug · kg–2 · min–2) as required.
Actuarial survival and freedom from reintervention based were estimated by the Kaplan-Meier method, using SPSS for Windows (version 11; SPSS Inc, Chicago, IL) statistical package. Variables were expressed as mean ± SD or median with range, and comparative univariate analyses were performed with the 
2 test, the 2-sided Fisher exact test, or binomial logistic regression, as appropriate. A probability value p < 0.05 was taken to represent a statistically significant difference between groups.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
There was complete atresia in five patients (31%) and hypoplasia in 11 patients (69%). In aortic hypoplasia, the mean diameter of the aortic annulus was 2.7 (±0.5) mm and the ascending aortic diameter was a mean of 4.1 (±1.5) mm. All patients had an aortic annulus z-score of less than –2. The mean mitral valve annulus was 8.9 (±1.9) mm.
Associated lesions were common and are summarized in Table 1, the most common being either aortic interruption (n = 9, 57%) or coarctation (n = 3, 19%). Only four patients (25%) had a normal aortic arch.
|
The 30-day mortality after the BVR was 19% (n = 3 patients). There have been no deaths in the last 12 years or 13 cases. All patients died of low cardiac output and pulmonary hypertension in the first few days after surgery. Extracorporeal membrane oxygenation. support was not used. The early mortality compared favorably with the overall mortality for a Norwood I procedure of 29% at this institution over the same period [8]. Although not statistically different, the survival benefit by era became significant over the past 10 years (p < 0.01) when compared with the survival from the Norwood procedure.
Complications included sepsis in three cases (two chest infections and one urinary tract infection) and necrotizing enterocolitis in one case, which was treated conservatively. The three patients that died had evidence of low output state and required peritoneal dialysis. The median intensive care and hospital stays were 11 days (range, 0 to 39 days) and 16 days (range, 0 to 39 days), respectively. No patient required a pacemaker, although one of the early deaths had remained in heart block at time of death on day 1. There were no documented adverse neurological events, although the three early deaths had remained paralyzed and sedated, so clinical neurological assessment had not been possible.
On univariate analysis only older age at operation (>6 weeks) was significant in predicting early mortality (p < 0.001). Preoperative inotrope requirement and ventricular function came close to reaching significance, both with p values of 0.06. Of note, anatomic subtype, presence of aortic interruption or coarctation, and smaller size or age were not significant risk factors. The need to enlarge the VSD was also not significant (see Table 2).
|
Freedom from reintervention after BVR was 68% (±6) at one year and 20% (±9) at five years. Three patients needed balloon dilatation; two of right ventricle to pulmonary artery conduit at five and at six months after initial repair, and one of the aortic arch at five months. There were five reoperations: four for replacement of RV-PA conduit at 17, 20, 40, and 116 months and one patient had a enlargement of LVOT at 34 months after the initial repair. Despite the anteriorly placed conduit, there were no incidences of the heart being injured at sternal reentry. The freedom from reintervention was not related to the type of conduit used at the initial operation. There was no mortality associated with reoperation-reintervention and homograft conduits were used for all rereplacements.
Actuarial survival for the entire group of 16 patients was 75 (±3)%, 68 (±3)%, 60 (±5)%, and 53 (±7)% at three months, six months, one year, and five years, respectively (Fig 3). This was comparable with the survival for Norwood procedure at the institution which was 66 (±2)%, 61 (±2)%, 58 (±3)%, and 50 (±3)% for the same period of time [8]. All the survivors are in New York Heart Association class 1.
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Prior to the past 20 years, these cases were thought not to be suitable for BVR. Norwood and Stellin published the first case report of BVR in 1981 [9], involving a two-stage repair and the placement of a conduit from the left ventricle apex to the descending aorta.
The technique described in this paper was first postulated by Freedom and colleagues [10] and then peformed by Yasui and colleagues in 1987 [6] with the first two successful cases of BVR, one of whom had transposition. Boston Children’s Hospital published a series of four patients two years later [7]. The technique involved the establishment of main pulmonary artery to aorta communication either directly or with a homograft and baffling of the VSD to this neoaorta. Right ventricle to distal pulmonary artery continuity was then provided with a homograft conduit.
Although BVR is now well-established, controversy remains as to whether this should be performed in a single stage as a neonate or in two stages with an initial, palliative procedure. Each technique has inherent strengths and weaknesses. An initial Norwood procedure as part of a staged repair carries a shorter ischemic time and avoids the placement of a RV-PA valved conduit in these small neonates (if a classical Norwood technique is used then the need for a ventriculotomy is also avoided). However, the postoperative course may be more labile because of the shunt-dependent pulmonary blood flow and the patient has a volume loaded, cyanotic circulation. Nevertheless, the procedure is attractive as it does not preclude from conversion to a BVR at any stage in the future.
The technical aspects of BVR present a considerable challenge, including the placement of the RV-PA conduit in a neonate. The course for the conduit is unfavorable because of the relatively bulky Damus-type arch reconstruction limiting the available space and the relatively posteriorly positioned pulmonary arteries. An alternative approach would involve translocating the branch pulmonary arteries anterior to the neoaorta by a Lecompte maneuver, which may facilitate the placement of a conduit [11]. Alternatively, conduit insertion may be avoided entirely and a direct anastomosis with or without a monocusp valve may be performed [12, 13].
Ohye and colleagues [14] compared the single versus two-stage approach but could not demonstrate any superiority with either technique. The actuarial survival at five years for the entire group was 78% (16 of 20 patients) and after primary BVR was 73% (8 of 11 patients) compared with 89% (8 of 9 patients) for staged repair with initial Norwood followed by BVR at a later date. The conclusion was that both methods were acceptable but that refinements in their operative techniques had led them to favor the single-stage approach.
The findings from the current study would support those of Ohye and colleagues, and our experience has led us to aim for primary BVR where possible. There were no specific anatomic features that carried a risk for repair and the presence of aortic interruption or coarctation was not an additional risk factor for poor outcome. The finding that older age at operation (>6 weeks) was a risk factor for early death appeared to be related to low cardiac output state and ventricular failure in the first few postoperative days. The delay before surgery was due to a late diagnosis reflecting that these patients were relatively well in the first few weeks of life. Certainly, this would suggest that these patients would be at risk of pulmonary hypertension and may well have had a period of volume loading preoperatively. Ventricular function, however, was no different preoperatively to the younger patients. The finding is in keeping with older age being a risk factor for Norwood palliation of HLHS where a similar pathophysiology has been suggested [15].
Conduit reintervention is the most common late complication from neonatal BVR, and the issues are the same as those with other procedures requiring small conduits such as Truncus repair. As with most other studies, we prefer to use a valved conduit in view of the superior hemodynamics in the early postoperative period and accept the need for conduit replacement as a necessary evil. Certainly the use of direct RV-PA anastomosis or monocusp patches have tended to delay the need for RVOT reintervention [16, 17] but the concern would be that the resulting pulmonary regurgitation may be poorly tolerated, particularly if there is a concern over underlying pulmonary hypertension.
There is no doubt that the improving outcomes in the Norwood operation may generate increasing enthusiasm for a two-stage approach. Despite a preference for BVR in aortic atresia-VSD there is still little evidence that a staged approach carries any greater risk than single-stage BVR. Thus, with increasing confidence with the Norwood procedure there is still a valid argument that this approach will allow for a larger conduit to be placed at the time of the definitive repair. The rarity of the condition makes it difficult to produce a large series without there being a long time period over which improvements in surgical and intensive care management may have influenced outcomes. The outcomes of the Norwood approach versus BVR are sufficiently similar that a randomized prospective study may be the only way to establish whether one strategy is genuinely superior.
The Damus-Rastelli technique can be applied to aortic atresia-VSD with clinical outcomes that compare well with a standard univentricular approach. Anatomic subtype was not a significant risk for early death. Late age at first operation was the only risk factor identified for early death and may be avoided by earlier diagnosis and surgery.
Despite a high reintervention rate, mainly for conduit change, the long-term benefit of a biventricular circulation would support this technique. Delay in performing the initial repair may increase mortality.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
