| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Nationwide Children's Hospital, Columbus, Ohio
Accepted for publication June 25, 2009.
* Address correspondence to Dr Galantowicz, Nationwide Children's Hospital, 700 Children's Dr, Columbus, OH 43205 (Email: mark.galantowicz{at}nationwidechildrens.org).
Presented at the Forty-fifth Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Jan 26–28, 2009.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Methods: From July 2002 to March 2008 66 consecutive hybrid procedures for hypoplastic left heart syndrome were performed. Patients requiring RAAO intervention based on cardiology–surgery consensus were defined as group 1 (n = 16), whereas all other hypoplastic left heart syndrome patients formed group 2 (n = 50).
Results: At birth there were no differences between groups in terms of demographics or cardiac function. Group 1 had more patients with aortic atresia (94% versus 58%; p = 0.01), and 69% of patients had initial echocardiographic comments regarding incipient RAAO versus 26% in group 2 (p = 0.007). The type of ductal stent, balloon versus self-expandable, did not influence the subsequent development of RAAO. Before RAAO intervention (mean age, 74 days), group 1 patients had significantly more tricuspid regurgitation. The main treatment for RAAO in group 1 was coronary stent insertion, with 3 patients having a reverse central shunt. At a mean follow-up of 611 days, group 1 had reduced survival interstage (56.3% versus 88%; p = 0.005) and overall (43.7% versus 70%; p = 0.03).
Conclusions: Clinically important RAAO occurred in 24% of the hypoplastic left heart syndrome patients in this series. If RAAO is detected at birth or early interstage, a Norwood operation is now favored. Palliative interventional catheterization remains very important mid and late interstage for continuing the hybrid strategy toward comprehensive stage 2.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
A legitimate concern is that until stage 2 reconstruction hybrid patients have coronary and cerebral perfusion by means of retrograde flow through the aortic arch, with the possibility of subsequent obstruction at this site. In patients with aortic atresia this flow is entirely dependent on retrograde perfusion. Retrograde aortic arch obstruction (RAAO) can occur in utero, postnatally before surgical treatment begins, or afterward in acute or more insidious fashion at the time of PDA stenting or during subsequent follow-up [8, 9]. We report results with the hybrid approach in a series of HLHS patients with an emphasis on the development of this particular complication. In defining RAAO here we have taken a pragmatic approach. As discussed below, a comparison is made between patients who required a specific intervention for RAAO and all other HLHS patients in the series. For the benefit of capturing the multiple facets of this entity we have included here all HLHS patients treated on the hybrid pathway at our center, including the initial experience previously described as "the learning curve" [5]. It was through this experience that important insights have been made influencing our current treatment algorithm.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Technique
Our hybrid strategy and results were recently described in detail [6, 7]. Briefly, the neonate undergoes an initial hybrid stage 1 palliation consisting of bilateral pulmonary artery bands and a PDA stent. The operation is done through median sternotomy off cardiopulmonary bypass using a 3.0- or a 3.5-mm Gore-Tex (W.L. Gore & Assoc, Flagstaff, AZ) tube graft for creating the bands depending on whether the patient's weight was less than or greater than 2.5 kg, respectively. After bilateral pulmonary banding a sheath is inserted in the pulmonary artery and a stent is deployed aiming to cover the PDA completely. In our initial experience we used predominantly balloon-expandable stents, but in more recent years newly designed self-expandable stents were favored. Types of PDA stent used are as follows: balloon expandable, Genesis Blue, Genesis XD, premounted Genesis (Cordis/Johnson & Johnson, Miami, FL), and Formula 418 (Cook Inc, Bloomington, IN); and self-expandable, Precise, SMART (Cordis/Johnson & Johnson), Protégé (ev3, Plymouth, MN), and Zilver (Cook, Inc). In our initial experience we found that a balloon atrial septostomy at the time of the hybrid stage 1 is fraught with technical and hemodynamic difficulties [5, 7]. Currently this is performed in more stable conditions when the neonate is ready for discharge, at about 2 weeks after undergoing the hybrid procedure. The comprehensive stage 2 operation consisted of: removal of the PDA stent and the pulmonary bands, reconstruction of the aortic arch and pulmonary arteries (if needed), reimplantation of the diminutive ascending aorta into the pulmonary artery, atrial septectomy, and a bidirectional cavopulmonary anastomosis. The PDA stent was removed in all cases, the right pulmonary band site was incorporated in the Glenn shunt, and a left pulmonary patch augmentation was performed in approximately half of the cases.
Interstage cardiac catheterization in this series of 66 patients was performed under general anesthesia in selected cases. The indication was diagnostic (eg, evaluation of physiology or associated anomalies such as pulmonary vein stenosis) or therapeutic (eg, treatment of RAAO, PDA restenosis, or atrial septal restenosis). The diagnosis of RAAO was confirmed by pressure gradient measurement in the retrograde aortic arch using the Radi PressureWire (Radi Medical Systems, Inc, Uppsala, Sweden), and when appropriate, RAAO was treated by deploying a balloon-expandable coronary stent from the descending aorta (Fig 1). Angiographic improvement was verified after stenting, as well as by a drop in the directly measured RAAO gradient.
|
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
|
|
|
|
As detailed above a total of 4 patients in groups 1 and 2 had postoperative deaths after the hybrid procedure (6%). Figure 3 shows actuarial survival for the two groups, the figures to date being 56.3% versus 88% interstage (p = 0.005) and 43.7% versus 70% overall at a mean follow-up of 611 days (p = 0.03).
|
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
In the absence of a clear definition for what RAAO should represent, we adopted a practical approach in defining it as a combination of echocardiographic or clinical indicators suggestive enough to warrant further targeted intervention. The treatment allocation reflects clinical consensus but is not immune from bias. However, this classification allowed some clinical insight into an elusive and probably progressive entity, which cannot be fully described by a snapshot characteristic. Group 1 is not homogeneous in terms of treatments applied and reflects our changes in management strategy based on serial accrual of patients.
An important lesson is that RAAO can exist as a congenital anatomic variant in a subset of HLHS patients potentially manifesting itself in utero or postnatally before any surgical intervention. Patient 3 in group 1 experienced intermittent bradycardia in utero leading to an early delivery. After birth there were intermittent periods of instability thought to be secondary to infection delaying a hybrid stage 1. Immediately on deploying the PDA stent, cardiac ischemia and collapse necessitated extracorporeal membrane oxygenation support. Within hours this support was stopped because of brain death. Autopsy clearly demonstrated congenital RAAO (Fig 4). Subsequently our group learned how to evaluate this area by echocardiography, thereby ruling RAAO in or out before any surgical intervention (Fig 5). Using this screening we have seen that approximately 10% of patients with HLHS have RAAO at birth. We believe this is a contraindication to a hybrid stage 1 and refer those patients for a traditional Norwood procedure. An interesting aside may be that this subgroup of patients reflects some of the pretreatment variability reported in HLHS by neurologic imaging and neurocognitive studies. In terms of patient size, it may be intuitive that RAAO is more prevalent in small patients, but we have not seen any relationship, which encourages us to continue treating low-weight patients on the hybrid pathway.
|
|
Another lesson learned was that an early comprehensive stage 2, such as in patients 1 and 2, is not a good solution for RAAO. Although Jaquiss and colleagues [12] elegantly demonstrated how a cavopulmonary anastomosis can work before 4 months of age after a Norwood operation, we do not think that there is sufficient experience to recommend it in the current context. The comprehensive stage 2 is a bigger procedure than a stage 2 in the traditional pathway with success dependent on optimal cardiac and pulmonary function, which is not the case in a patient with RAAO.
Coronary stent insertion, on the other hand, has proved to be a more durable and predictable solution to RAAO. A few words of caution about this modality are also warranted. If resorted to in early interstage, stenting may trigger an intimal hyperplasia response that is self-perpetuating. In patients 8, 12, and 13 further interventions for RAAO were indeed required in the form of a central shunt or stent angioplasty with only 1 survivor. Nonsurvivors had stent implantation at a younger age but statistical significance and a cause and effect relationship are absent for now. In this experience, however, interventional catheterization was invaluable in mid and late interstage to transition patients toward comprehensive stage 2. Regardless of the timing of RAAO intervention, the importance of interstage surveillance cannot be overemphasized [1, 5]. Unmonitored sudden death was also present in group 2 and remains an occurrence in all contemporary series.
What causes clinically important RAAO in the first place and how is it best diagnosed? Our current series can offer some hints, but more targeted research is needed to answer these questions. First of all groups 1 and 2 had few baseline differences in terms of demographics, anatomy, and ventricular function. Group 1 had a significant clustering of cases with aortic atresia (94%), a situation entirely reliant on retrograde arch flow. It is interesting to note that significantly more patients in group 1 had some echocardiographic suggestions of risk for subsequent RAAO (Table 2). The hint was an echocardiographic gradient or a subjective comment such as "narrowing of isthmus" or "juxtaductal shelf." This shows that early indicators of RAAO may be present, and we have embarked on studying these further so as to be able to have a better triage of patients on the surgical pathways.
In terms of PDA stents our choices parallel those of the group in Giessen, Germany [3]. We prefer the new design of the self-expandable stent and will only use a balloon-expandable device if the PDA is stenosed and requires dilatation. The type of stent was not associated with RAAO in this analysis, but to what extent the origin of the retrograde aortic arch was covered initially by the stent is being examined in a separate study. In our opinion it is not advisable to have too short a stent at the junction of the PDA with the aorta because this can lead to recoarctation and possible problems with retrograde flow. Crossing the retrograde orifice with an open-cell–design stent has not been problematic in our experience and allows further catheterization targeted at RAAO. From a surgical point of view it was possible to remove the stent entirely in every case at comprehensive stage 2. Compared with another group with extensive experience in the hybrid approach it is not clear why we are seeing more RAAO. It may be partly related to patient profile; Akinturk and coworkers [3] reported fewer cases of aortic atresia and more patients eventually amenable to a biventricular repair.
This study has a number of limitations. First, RAAO was not defined at the outset, and group 1 does not have a "before and after" comparison of treatment effects in this retrospective design. Defining the groups on the basis of nonalgorithmic treatment allocation is another source of bias, but on the other hand it reflects the clinical reality of treating these patients. The report is mainly descriptive in presenting this problem and the steps we took to deal with it. There can be only speculative discussions at this stage about the anatomy of clinically important RAAO and its true physiologic impact. Autopsy data were very limited because of a lack of consent in all patients. With only 16 RAAO index cases, it is also impossible to have more cause and effect insight because multivariate analysis remains out of bounds. Finally, it is of course too early to evaluate neurologic outcomes in hybrid patients with and without RAAO. In the interim we have taken the view that satisfactory cardiac function is a surrogate of equally satisfactory cerebral function between stages 1 and 2. This assumption is in keeping with a detailed autopsy study showing that the aortic diameter distal to the origin of the innominate artery is significantly bigger than the aortic dimension just proximal to this landmark [9]. Nonetheless, this study demonstrates satisfactory results with the hybrid approach for HLHS in a series of consecutive patients and is a detailed clinical report of RAAO. Our group is currently researching early imaging indicators of RAAO as well as the influence that a hybrid pathway has on cerebral perfusion and function. In a separate analysis we have shown that echocardiographic variables are predictive of the clinical course interstage [13].
In conclusion, a high index of suspicion is required at all stages for detecting and treating this complication. It has become useful for us to think of three separate time frames for the development of RAAO that may be managed differently. First, RAAO may be diagnosed echocardiographically at birth or preoperatively. In this case we would view it as a contraindication to a hybrid approach, recommending instead a Norwood operation. Second, for rapidly progressive early interstage RAAO one should consider either a conversion to a Norwood procedure or plan the retrograde stenting such that the area can be restented if in-stent restenosis develops. Finally, if RAAO is detected in mid or late interstage, the transcatheter placement of a retrograde stent can alleviate the stenosis, allowing the patient to transition to a successful comprehensive stage 2 on the hybrid pathway.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Doctor Stoica and colleagues have made a comprehensive report of outcomes using a hybrid strategy for palliation of hypoplastic left heart syndrome, and, along with the Geissen group, these authors are to be congratulated for pioneering an innovative strategy which challenges the more traditional Norwood strategy in this difficult group of patients. At the Hospital for Sick Children in Toronto, we have developed a hybrid program as well based on the experience of these groups and crafted our program based on their lead. This is a tough group of patients, and registry data typically is consistent with 2- to 3-year survival in the 60 to 80% range. It is against this backdrop that the hybrid approach has been advocated as an alternative to the Norwood strategy and represents an important innovation in our field.
It is said that a great deal of medical innovation undergoes three phases. First is a "boom" phase in which the innovation appears as a panacea which neatly solves a clinical problem and is often associated with rapid adoption by many centers. After the boom phase comes the disillusionment phase when problems surface, which suddenly render the innovation somewhat flawed and perhaps not ideal for all patients, and, finally, a niche phase is established. This transition to a niche phase is aided by, typically, refinements in technique and improvements in patient selection.
I would submit that the Columbus group is striving to enter the niche phase where the patient selection and refinements in technique will establish the role of the hybrid as an alternative to the Norwood strategy. Now is the time we must entertain debate and wrangle over selection criteria and refinements in technique. Within this context I have the following questions.
The authors have identified retrograde flow through the isthmus as the Achilles' heel of the hybrid strategy. Their current approach relies on rescue therapy using catheter-based interventions after the problem is diagnosed. This commonly occurs after detectible decreases in myocardial performance. Twelve patients were treated with this catheter-based strategy and 6 either died or had diminished myocardial function requiring transplantation. My basic question is whether the authors have considered that their treatment of this problem may simply be ineffective. The data suggests that once the problem has occurred, catheter-based interventions do not effectively prevent sudden death or myocardial failure. Could the authors provide data to demonstrate the efficacy of these interventions at the time of intervention in terms of augmentation of isthmus dimensions and gradient reduction?
Second, what are the characteristics of patients in whom in-stent restenosis has developed? Could failure to achieve effective restoration of isthmus patency have contributed to the increased mortality in this group?
Again, I compliment the authors on their pioneering efforts and look forward to their responses. Thank you.
DR STOICA: I wonder if I could please have my last slide back. Dr Caldarone, first of all, thank you very much for your comment and questions. We do value very much the contributions you are making in this new field.
Before answering your first question, I would like to put the numbers in perspective, if I may. This complication is only seen in a quarter of the patients, and of those at risk, for example, with aortic atresia, only half of the patients with aortic atresia were in group 1 and twice as many were in group 2. We expect that these numbers at risk and the prevalence of this complication will go down after the lessons we learned here and when our diagnosis methods are continuously refined.
I would argue that our treatment is also proactive and, to a large extent, effective. If we look at our current philosophy in subset 1 (early, preintervention diagnosis) before treatment is embarked on, the Norwood strategy will be effective and will solve the problem of unnecessary attrition. Subset 3 (diagnosis late interstage) can also be treated effectively we think by catheter-based intervention.
I wouldn't say that this is done late, because RV (right ventricular) function is still preserved. I will come back to the data. There is the group in the middle (subset 2, early interstage or recurrent RAAO [retrograde aortic arch obstruction]), however, for which we don't have a good solution as yet. It is a gray area, it is an evolving area, and we think that a Norwood conversion is the best way forward until this problem is sorted out.
In-stent stenosis does occur. We see it at about 6 to 8 weeks. But if the patient is old enough, for example 3 months on average, 2 months later they can have a safe comprehensive stage 2 with a good result. We rely heavily on interstage monitoring, and the diagnosis is made while RV function is preserved.
In terms of using a reverse BT (Blalock-Taussig) shunt, this complication does not occur in everybody, so we don't think that everybody needs to be prophylactically treated for that. There is a strong potential for coronary steal in this new physiology that we don't understand very well, and in fact, 2 of our 3 patients developed this and died, but all of them had coronary steal.
To conclude, I will say that this cohort reflects our early experience, the lessons learned from adopting a hybrid approach and, more recently, for retrograde arch obstruction.
Thank you very much.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Pizarro, J. Kolcz, C. D. Derby, D. Klenk, J. M. Baffa, and W. A. Radtke Hard Choices for High-Risk Patients With Critical Left Ventricular Outflow Obstruction: Contemporary Comparison of Hybrid Versus Surgical Strategy World Journal for Pediatric and Congenital Heart Surgery, July 1, 2010; 1(2): 187 - 193. [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 |
