Ann Thorac Surg 2008;85:1397-1402. doi:10.1016/j.athoracsur.2007.11.054
© 2008 The Society of Thoracic Surgeons
Original Articles: Pediatric Cardiac
Factors Affecting Long-Term Risk of Aortic Arch Recoarctation After the Norwood Procedure
Traci M. Ashcraft, PA-Ca,
Karen Jones, CNPa,
William L. Border, MBChBb,
Pirooz Eghtesady, MD, PhDa,
Jeffrey M. Pearl, MDa,
Phillip R. Khoury, MSb,
Peter B. Manning, MDa,*
a Division of Cardiothoracic Surgery, The Heart Center, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio
b Division of Cardiology, The Heart Center, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio
Accepted for publication November 16, 2007.
* Address correspondence to Dr Manning, Division of Cardiothoracic Surgery, Cincinnati Children’s Hospital Medical Center, 3333 Burnet Avenue, Cincinnati, OH 45229-3039 (Email: peter.manning{at}cchmc.org).
Presented at the Fifty-third Annual Meeting of the Southern Thoracic Surgical Association, Tucson, AZ, Nov 8–11, 2006.
| Dr Manning discloses that he has a financial relationship with Shelhigh, Inc.
|
 |
Abstract
|
|---|
Background: The purpose of this study was to identify factors predicting risk of aortic arch recoarctation after the Norwood procedure.
Methods: Patient records were reviewed retrospectively for consecutive patients who underwent the Norwood procedure from 1996 to 2005. Preoperative and intraoperative parameters were identified for analysis. Aortic arch recoarctation was defined by the need for catheter or surgical reintervention. Data were analyzed using survival analysis, with freedom from intervention as the outcome. Factors predicting need for reintervention were analyzed using Cox proportional hazards regression.
Results: Thirty-five recoarctations were observed in 117 patients (30%). Freedom from aortic arch reintervention at six months, one, three, and five years were 72%, 63%, 56%, and 52%, respectively. The majority of arch reinterventions occurred in the first six months (63%), involving either surgical (43%) or catheter (57%) techniques. The use of bovine pericardium showed the greatest risk for potential recoarctation (hazard ratio = 1.81 [0.90–3.64], p = 0.09). Age, gender, weight, ascending aortic diameter, ventricular morphology, primary anatomic diagnosis, and coarctation shelf resection were not found to be predictors of recoarctation.
Conclusions: Most interventions for aortic arch recoarctation after the Norwood procedure occur within the first six months of life. The type of patch material used for arch reconstruction appears to influence, most strongly, the long-term risk of aortic arch recoarctation.
The first successful palliation of hypoplastic left heart syndrome (HLHS) was originally described in 1983 by Norwood and colleagues [1]. The Norwood procedure has evolved over time and currently there are various surgical strategies to accomplish four basic underlying goals: relieve systemic outflow obstruction, provide unobstructed coronary blood flow, create a nonrestrictive atrial septal communication, and provide a regulated source of pulmonary blood flow [2]. Recoarctation of the neoaorta is widely recognized as a common complication after the Norwood procedure. In fact, the incidence has been previously reported to be between 11% and 37% [3–9] and is a significant cause of morbidity and mortality. Aortic arch obstruction can be deleterious by causing an increased afterload on the single ventricle resulting in ventricular dysfunction, tricuspid regurgitation, and an imbalance in the ratio of systemic-to-pulmonary blood flow. Factors predisposing a patient to recoarctation after the Norwood procedure have not been well-characterized. Thus, our study aims to determine which preoperative and intraoperative factors predict the development of aortic arch recoarctation after the Norwood procedure.
|
Material and Methods
|
|---|
Preoperative and intraoperative data were collected retrospectively in consecutive patients who underwent the Norwood procedure at Cincinnati Children’s Hospital Medical Center between October 1996 and March 2005. Follow-up data were collected from referring cardiologists until the study end date of October 2005. Preoperative patient factors included anatomic cardiac diagnosis, ventricular morphology, ascending aortic diameter, gender, age, and weight at the time of the Norwood procedure. Operative parameters included two factors; whether or not the coarctation shelf was resected during aortic arch reconstruction and whether a patch of pulmonary artery homograft or bovine pericardium (Shelhigh, Inc., Union, NJ) was used to reconstruct the neoaorta. This study was approved by our Institutional Review Board (IRB) at Cincinnati Children’s Hospital Medical Center on March 25, 2005. Individual consent was waived by the IRB due to the retrospective design of the study.
During this time period, our technique for aortic arch reconstruction evolved. Early in our experience the technique included opening the aortic arch and descending aorta across the coarctation followed by reconstruction of the neoaorta with a pulmonary homograft patch. During 1999, we began using a technique of resection of the coarctation shelf and "interdigitating" reanastomosis of the distal arch to the posterior descending aorta in selected cases. Due to periodic difficulty in obtaining pulmonary artery homograft patches, we began using Shelhigh bovine pericardium in the year 2000 as an alternative patch material to reconstruct the neoaorta. Shelhigh bovine pericardium was approved by the Food and Drug Administration for use in vascular surgery in 1997. The harvested bovine pericardium was detoxified by the pioneered No-React Detoxification Process of the company. This process allowed the material to be biocompatible, yet non-absorbable and durable. Furthermore, investigational data in children and young adults using similar Shelhigh products exhibited no evidence of degeneration or calcification (10). The "off the shelf" availability made the product attractive and useful as an alternative patch material for aortic arch reconstruction.
Descriptive data of both preoperative and intraoperative patient factors are shown for the overall cohort in Tables 1 and 2.
We defined recoarctation as the need for catheter or surgical reintervention. The need for reintervention was at the discretion of the cardiac surgeon, the cardiologist, and the cardiac interventionalist, based upon the patient’s hemodynamic status during follow-up visits and catheterization measurements. At our center we routinely performed catheterizations on all single-ventricle patients prior to their Glenn procedure. While no absolute pressure gradient was adopted as a specific indication for intervention, due to the recognized significant adverse impact of this problem on the long-term function of the single-ventricle physiology, we followed an aggressive strategy for managing restenosis. The decision to perform surgical versus catheter-based intervention depended on several variables including the site and configuration of recoarctation (proximal/midarch or near branch vessels favored surgical treatment), timing of next operation, and physician discretion.
Data are reported as frequencies, medians with ranges, or means with standard deviations as appropriate. The Fisher exact tests for categoric variables or analysis of variance t tests for continuous variables were used for intergroup comparisons. Cox proportional hazard regression models were used to predict the need for reintervention. Hazard ratio, p value, and 95% confidence intervals are reported, respectively. Data were analyzed using survival analysis, with freedom from intervention as the outcome. All statistical analyses were performed with SAS version 9.1 software (SAS Institute, Inc, Cary, NC).
|
Results
|
|---|
One hundred seventeen patients underwent the Norwood procedure at our center between October 1996 and March 2005. Thirty-five patients (30%) underwent reintervention for recoarctation, 15 (43%) surgical, and 20 (57%) catheter-based. One patient was lost to follow-up, and thus, censored from these results. Table 3
demonstrates the incidence of recoarctation for each preoperative and operative factor independently. Comparison of patients experiencing recoarctation to those without recoarctation showed no patient factors statistically correlating with the incidence of restenosis. Table 4
displays an independent risk analysis performed using a Cox proportional hazard regression model. This model examined the risk for recoarctation for each variable independently. Although not statistically significant, bovine pericardium (hazard ratio = 1.81[0.90–3.64], p = 0.09) showed an increased risk for potential recoarctation. Age, gender, weight, ascending aortic diameter, ventricular morphology, primary anatomic diagnosis, and coarctation shelf resection were not found to be significant predictors (Table 4).
View this table:
[in this window]
[in a new window]
|
Table 4 Univariate Cox Proportional Hazard Ratio Model of Preoperative and Intraoperative Patient Factors With Relative Risk of Recoarctation After the Norwood Procedure
|
|
Survival analysis allowed us to censor those patients who were either lost to follow-up or died within the time period of the study. Figure 1
shows the cohort’s overall freedom from recoarctation after the Norwood procedure. The majority of aortic arch reinterventions occurred during the first six months after the Norwood procedure (n = 22, 63%). Overall freedom of recoarctation for the total cohort (n = 117) at six months, one year, three years, and five years was 72%, 63%, 56%, and 52%, respectively. Figure 2
displays the freedom from recoarctation dependent on the patch material used during the Norwood procedure; either pulmonary homograft or bovine pericardium. Although not statistically significant (p = 0.09), the use of bovine pericardium trended toward an increased risk of recoarctation over time compared with the use of pulmonary artery homograft.
View larger version (8K):
[in this window]
[in a new window]
|
Fig 1. Overall freedom from recoarctation after the Norwood procedure. (Patients at risk noted at selected time intervals.)
|
|
View larger version (12K):
[in this window]
[in a new window]
|
Fig 2. Neoaorta patch material: freedom from recoarctation after the Norwood procedure. (Patients at risk noted at selected time intervals.)
|
|
|
Comment
|
|---|
Aortic arch recoarctation after the Norwood procedure is common [11–14]. The Norwood procedure itself remains technically challenging, with various options available for aortic arch reconstruction. Differing surgical techniques have evolved over the years to reduce the incidence of recoarctation. It is well-recognized that even mild anatomic obstruction of the neoaorta can have significant deleterious impact. The increased afterload placed on the single pumping ventricle can lead to severe ventricular dysfunction, lead to tricuspid regurgitation, and drive more flow through a Blalock-Taussig shunt, causing an unfavorable systemic-to-pulmonary blood flow ratio. Previous investigators examined 122 necroscopic heart specimens from patients who died after the Norwood procedure, and recoarctation was judged to be the fourth most common cause of death, occurring in 17 patients (14%) [12]. Others have suggested that 7% to 8% of all deaths after the Norwood procedure can be attributed to aortic arch recoarctation [12, 13].
A number of theories have been proposed to explain the underlying cause of aortic arch recoarctation after the Norwood procedure. Reconstructed neoaorta may display dramatic caliber changes through the arch, which may increase the chance of kinks or folds in the patched neoaorta. Furthermore, Machii and Becker [15] concluded that any remaining ductal tissue extending into the aorta increased the incidence of recoarctation after the Norwood procedure. Thus, complete coarctation excision is recommended by some to avoid future aortic obstruction. Burkhart and colleagues [16] describe an interdigitating technique used during the aortic arch reconstruction that splits any potential residual ductal tissue in both an anterior and posterior position. Their interdigitating technique showed a significant decrease in postoperative aortic arch obstruction.
The availability of cryopreserved homograft tissue has, at times, been unreliable. Shelhigh bovine pericardium is treated with a proprietary technique claimed to eliminate the risk of calcification, and the patches have been manufactured in a curved, tubular configuration to facilitate the aortic arch reconstruction performed with the Norwood procedure. The "off the shelf" availability made the product attractive and useful as an alternative patch material for aortic arch reconstruction. In addition, bovine pericardium is nonimmunogenic in contrast to homograft tissue, which has been demonstrated to induce elevation of panel-reactive antibody levels that may persist for years after implantation [17]. This may be an important consideration in a population which carries a higher than average risk for future transplant consideration. Although not statistically significant, bovine pericardium showed an increased risk for recoarctation in our study. From experience with other Shelhigh vascular products, specifically the valved pulmonary artery conduit (manufactured in part from pericardium), we and others have noted a significant multinucleated giant cell reaction leading to stenosis, despite a lack of calcification [18, 19].
In summary, most interventions for aortic arch recoarctation after the Norwood procedure occur within the first six months of life. There was a trend toward higher restenosis risk associated with the type of patch material, which did not reach statistical significance. Age, gender, weight, ascending aortic diameter, ventricular morphology, primary anatomic diagnosis, and coarctation shelf resection were not found to be significant predictors.
|
Discussion
|
|---|
DR THOMAS L. SPRAY (Philadelphia, PA): This is a very interesting study, mainly because it brings up the continued problem of aortic arch obstruction after a repair that should have a low incidence of this type of problem. Obviously there are many factors that effect reintervention on the aortic arch. One factor is that when you look for arch obstruction. Typically patients are catheterized prior to their Glenn procedures, so this is an obvious time when one might anticipate finding a coarctation that is either clinically irrelevant or mild. I support the concept that you have outlined that many of these coarctations may be a shelf for some minor abnormality of the arch that is identified in the catheterization laboratory, and then interventional balloon dilation for a modest gradient is done.
Another concern is that there may be late recoarctation, but unless you have a consistent recatheterization at various time periods or you have consistent follow-up looking for gradients, you may underestimate the actual incidence of late recoarctation. While your reintervention curve tends to drop off after the first six months, as you might expect, there may be even a higher rate of recoarctation than you have identified. Regardless, you still have shown that recoarctation is a problem despite operative techniques that have developed significantly over the past 20 years. In our own experience, which in the same timeframe of your report would represent approximately 550 Norwood stage I procedures, I think that our actual recoarctation rate is not dissimilar to yours and to some other reported series. The interesting and unusual thing about your series compared to most others is the high incidence of surgical reintervention. In the last 12 years that I have been in Philadelphia there have been only two patients out of approximately 700 stage I procedures that have required surgical reintervention for coarctation. Balloon dilation appears to be very effective for treatment of recoarctation in these patients. Why did you feel surgical coarctation intervention was necessary in a significant portion of your patients and what did you find at the time of surgical intervention? Was there something unusual about the patch; was it not extended far enough distally; was it the patch material itself? What were the characteristics of the Shelhigh patch material that might make recurrent coarctation more likely?
There have been some recent reports suggesting that with coarctation resection there is a very low incidence of recurrent obstruction and I think there is a trend in most centers to do coarctation resection much more aggressively at the first stage operation. Based on your experience are you going to alter your techniques to try to prevent recoarctation in the future?
How did you define coarctectomy? Why did you do it in some patients and not others? I think most of us recognize that with the wide variability in anatomy some patients have a very distinct, very tight coarctation area with a lot of associated ductal tissue and others have more of an infolding of the aortic wall without a true coarctation shelf. I wonder if you looked into anatomic features as part of your analysis. Were those patients that had true coarctation with ductal tissue that wasn’t resected more likely to develop coarctation? I think this is an interesting report that emphasizes the fact that we really aren’t doing as well as we need to do with the first stage Norwood procedure.
DR MANNING: Thank you for your comments. I will address the last issue first, and it is one of the limitations of a retrospective study, to some extent, because what we did and how we did it and why we did it was an evolutionary thing over time. Early in the series the coarct resections that we did were predominantly in the patients that had an obvious coarct shelf and that segment of aorta was very narrow, and it was a relatively limited excision and direct reanastomosis without what some people have termed an interdigitating reconstruction. I think as time has gone on we have had a lower threshold for doing that, partly because we have become more familiar with it and partly because, as you pointed out, other authors have described that as a good maneuver and something that they have found has decreased the incidence of restenosis. We have employed it more often in recent years. I think we have employed the technique of trying to interdigitate the arch down onto the descending aorta a little bit more, but it was impossible to really tease out the influence on restenosis of the different ways and reasons why we used coarct resection. I was sort of surprised. I was hoping that we would find that it would make a difference. I believe that it is a good thing to do and I still employ it.
I will hit your second issue in terms of what we found and why the high rate of surgical intervention, because that is what really prompted us to embark upon the study. We were surprised, compared to most series, how often surgical reintervention was needed, and we were suspicious it was the pericardial patch. The analysis did not show the patch material as a risk for surgical reintervention, but what we observed tended to be more of a dense fibrosis. Microscopically when we had biopsy specimens there was a multinucleated giant cell reaction associated with the pericardial patch when that had been used. It wasn’t typically always distally either. We had some unusual narrowings that were even more in the more proximal portion of the aorta, which is part of the reason why it was done surgically rather than attacked in the cath lab.
DR JOHN H. CALHOON (San Antonio, TX): Dr Manning, I enjoyed your presentation. Could you just comment on the pericardial patch that you used? I have used that only once, had terrible difficulty with getting it to come down to the native aorta, and keep it from pulling through tissue. I found it to not handle very well or work very well. Maybe you could comment on that. It seems like you didn’t have trouble with it because you used it more than once, but I wouldn’t.
DR MANNING: Honestly, the handling of the pericardial patch really was not too much of a problem. Some of the times in these reconstructions I have been bothered by the homograft tissue sometimes being a little bit flimsy. This stuff was very flexible and plenty sturdy. We didn’t seem to have a lot of bleeding problems with the reconstruction. So in our hands it was a fine material to work with, but unfortunately I think we are concerned that it was a good part responsible for some of our unusual restenoses.
|
References
|
|---|
- Norwood WI, Lang P, Hansen DD. Physiologic repair of aortic atresia-hypoplastic left heart syndrome N Engl JMed 1983;308:23-26.[Medline]
- Tweddell JS. The Norwood procedure with an innominate artery-to-pulmonary artery shunt Operat Tech Thorac Cardiovasc Surg 2005;10:123-140.
- Chang AC, Farrell PE, Murdison KA, et al. Hypoplastic left heart syndrome: hemodynamic and angiographic assessment after initial reconstructive surgery and relevance to modified Fontan procedure J Am Coll Cardiol 1991;17:1143-1149.[Abstract]
- Chessa M, Dindar A, Vettukattil JJ, et al. Balloon angioplasty in infants with aortic obstruction after the modified stage I Norwood procedure Am Heart J 2000;140:227-231.[Medline]
- Zellers TM. Balloon angioplasty for recurrent coarctation of the aorta in patients following staged palliation for hypoplastic left heart syndrome Am J Cardiol 1999;84:231-233.[Medline]
- Ishino K, Stümper O, de Giovanni JJV, et al. The modified Norwood procedure for hypoplastic left heart syndrome: early to intermediate results of 120 patients with particular reference to aortic arch repair J Thorac Cardiovasc Surg 1999;117:920-930.[Abstract/Free Full Text]
- Meliones JN, Snider R, Bove EL, Rosenthal A, Rosen DA. Longitudinal results after first-stage palliation for hypoplastic left heart syndrome Circulation 1990;82(suppl 5):IV151-IV156.[Medline]
- Murdison KA, Baffa JM, Farrell Jr PE, et al. Hypoplastic left heart syndrome. Outcome after initial reconstruction and before modified Fontan procedure. Circulation 1990;82(suppl 5):IV199-IV207.[Medline]
- Tworetzky W, McElhinney DB, Burch GH, Teitel DF, Moore P. Balloon arterioplasty of recurrent coarctation after the modified Norwood procedure in infants Catheter Cardiovasc Intervent 2000;50:54-58.[Medline]
- Marianeschi SM, Iacona GM, Seddio F, et al. Shelhigh No-React porcine pulmonic valve conduit: a new alternative to the homograft Ann Thorac Surg 2001;71:619-623.[Abstract/Free Full Text]
- Helton JG, Aglira BA, Chin AJ, Murphy JD, Pigott JD, Norwood WI. Analysis of potential anatomic or physiologic determinants of outcome of palliative surgery for hypoplastic left heart syndrome Circulation 1986;74(3 pt 2):I70-I76.[Medline]
- Bartram U, Grunenfelder J, van Praagh R. Causes of death after the modified Norwood procedure: a study of 122 postmortem cases Ann Thorac Surg 1997;64:1795-1802.[Abstract/Free Full Text]
- Jonas RF, Lang P, Hansen D, Hickey P, Castaneda AR. First-stage palliation of hypoplastic left heart syndrome J Thorac Cardiovasc Surg 1986;92:6-13.[Abstract]
- Weinberg PM, Chin AJ, Murphy, Pigott JD, Norwood WI. Postmortem echocardiography and tomographic anatomy of hypoplastic left heart syndrome after palliative surgery Am J Cardiol 1986;58:1228-1232.[Medline]
- Machii M, Becker AE. Nature of coarctation in hypoplastic left heart syndrome Ann Thorac Surg 1995;59:1491-1494.[Abstract/Free Full Text]
- Burkhart HM, Ashburn DA, Konstantinov IE, et al. Interdigitating arch reconstruction eliminates recurrent coarctation after the Norwood procedure J Thorac Cardiovasc Surg 2005;130:61-65.[Abstract/Free Full Text]
- Hooper DK, Hawkins JA, Fuller TC, Profaizer T, Shaddy RE. Panel-reactive antibodies late after allograft implantation in children Ann Thorac Surg 2005;79:641-644.[Abstract/Free Full Text]
- Pearl JM, Cooper DS, Bove KE, Manning PB. Early failure of the Shelhigh pulmonary valve conduit in infants Ann Thorac Surg 2002;74:542-548.[Abstract/Free Full Text]
- Schreiber C, Sassen S, Kostolny M, et al. Early graft failure of small-sized porcine valved conduits in reconstruction of the right ventricular outflow tract Ann Thorac Surg 2006;82:179-185.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
T. P. Graham Jr
The Year in Congenital Heart Disease
J. Am. Coll. Cardiol.,
October 28, 2008;
52(18):
1492 - 1499.
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
