| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Eur J Cardiothorac Surg 2003;24:28-36
© 2003 Elsevier Science NL
a Cardiothoracic Unit, Great Ormond Street Hospital for Children NHS Trust, Great Ormond Street, London WC1N 3JH, UK
b Institute of Child Health, London, UK
Received 29 October 2002; received in revised form 12 March 2003; accepted 17 March 2003.
* Corresponding author. Tel.: +44-20-7405-9200; fax: +44-20-7831-4931
e-mail: tsangv{at}gosh.nhs.uk
 |
Abstract |
|---|
 Top Abstract 1. Introduction2. Methods3. Results4. DiscussionAppendix A. Conference...References |
|---|
Key Words: Total anomalous pulmonary venous connection • Pulmonary venous obstruction
|
1. Introduction |
|---|
|
TopAbstract1. Introduction2. Methods3. Results4. DiscussionAppendix A. Conference...References |
|---|
Data from the literature suggest that PVO may predominate after correction of infracardiac obstructed TAPVC and cardiac TAPVC [2]. However, none of the anatomic arrangements appear to be spared by the disease [3]. The anatomic and pathologic correlates may vary from fibrosis and neointimal proliferation at the anastomotic site, to segmental or diffused intimal hyperplasia within individual pulmonary veins (PVs) remote from the anastomosis [1,4]. Accordingly, the clinical spectrum of the disease ranges from pure anastomotic obstruction to isolated narrowing of one or more PVs, although it is not unusual for these patterns to be found in combination. In patients treated for cardiac TAPVC with connection to the coronary sinus (CS), there seems to be a specific risk for the development of PVO due to the accumulation of neointimal tissue at the site of intracardiac repair, irrespective of the patch material used.
According to the limited published data, the occurrence of PVO following correction of TAPVC portends a poor prognosis [5]. Obstruction that involves individual PVs and manifests itself early after primary repair seems to be more common and especially difficult to treat, in contrast to that presenting late and involving the anastomotic site in isolation [3,4]. However, while the course and pathologic features of the disease have been well documented, there is still uncertainty as to what factors precipitate these lesions, particularly intimal hyperplasia of PVs away from the anastomosis. Also, the substantial failure rate of re-interventions for PVO remains incompletely understood, and the risk factors for failure have not been clarified.
The aim of this study was to review our experience of reoperations for PVO occurring after correction of TAPVC in infancy, in an attempt to expand the knowledge on the disease and identify risk factors for adverse outcome.
|
2. Methods |
|---|
|
TopAbstract1. Introduction2. Methods3. Results4. DiscussionAppendix A. Conference...References |
|---|
2.1. Patients' clinical characteristics
All patients had their primary repair within the first 2 months of life. Clinical data are summarised in Table 1. Primary repair was carried out by using a combination of hypothermic cardiopulmonary bypass and circulatory arrest. Supra- and infracardiac variants were repaired by connecting the retrocardiac pulmonary venous confluence to the back of the left atrium, after the PV confluence was widely incised. The incision was not routinely extended into the PV ostia. A running 7-0 polypropylene suture was used for anastomosis. An ascending vein was ligated in all but two patients.
|
2.2. Diagnosis of PVO after TAPVC repair
PVO was diagnosed by transthoracic Doppler echocardiography in all patients, and confirmed by transoesophageal echocardiography (TOE) in three patients. Assessment of pulmonary venous flow relied on the identification of individual PVs, on the measurement of maximal pulmonary venous flow velocity (m/s), and on the presence or absence of dynamic flow curve variations in response to cardiac and respiratory cycles. Right ventricular (RV) pressure was estimated from the maximal flow velocity of the tricuspid valve regurgitant jet, when present. In order to further delineate the extent of the obstruction and the anatomy of individual PVs, nine patients underwent cardiac catheterisation (8/20; 40%).
Patients who presented with PVO immediately after TAPVC repair (i.e. pulmonary venous flow velocity of 1.7 m/s or greater, continuous flow not returning to baseline), requiring re-intervention within the first month of primary correction, were excluded from this series. In the others, PVO was defined as ‘anastomotic’ when it involved solely the anastomosis previously constructed between the pulmonary venous confluent and LA, and ‘ostial’ when it involved individual PVs and/or their ostia (i.e. the junction between PVs and LA). Additionally, PVO was categorised as new onset when it appeared after an entirely satisfactory primary repair (i.e. echocardiography after correction of TAPVC showing non-turbulent pulmonary venous flow at less than 1.2 m/s, biphasic) or as ‘residual’ when it developed on the ground of subtle echocardiographic changes detected early after TAPVC repair (pulmonary venous flow velocity of 1.2–1.6 m/s, biphasic or nearly biphasic). These patients, initially managed expectantly, subsequently progressed to developing severe PVO for which surgical treatment became necessary.
2.3. Surgical treatment of PVO
As the location and extent of PVO varied widely among patients, a variety of surgical techniques were employed to relieve the obstruction, alone or in combination (Table 2). Generally, reoperations were performed under hypothermic cardiopulmonary bypass (CPB); brief periods of HCA were used infrequently as needed. Enlargement of the anastomosis and PVs using native left atrial (LA) tissue (nine times) and LA appendage (five times) were preferred, whenever possible. Patch enlargement with Gore-tex was used in several patients as an alternative to a repair with native atrial tissue, based on the belief that this material would induce less scarring and neointimal growth at the repair site, as compared to other artificial materials. In two patients (#2 and #6), we recently carried out a ‘sutureless’ in situ pericardial repair to relieve obstruction of the right-sided PVs, as described by Lacour-Gayet et al. [6]. In one patient (#8), we used a modification to this technique, which consisted of performing a ‘sutureless’ repair with a free pericardial patch, as illustrated in Fig. 1 . After the obstructed PVs were widely incised, a free patch of autologous pericardium was sewn over the incision, but well away from its edges, so as to reconstruct the pulmonary venous pathway (Fig. 1). Placing the suture line away from the endothelium and avoiding traumatic manipulations of the pulmonary venous wall could reduce the risk of subsequent re-stenosis. In another patient (# 20) presented with obstruction of the CS–LA junction following repair of cardiac TAPVC to CS, the autologous pericardial patch originally used for the repair was found to be severely thickened and obstructive. After the CS was cut further back into the left atrium, a wide patch of Gore-tex was used to complete the repair and reconstruct the atrial septum.
|
|
The presence of residual obstruction after re-intervention was categorised as mild or moderate, based on arbitrary echocardiographic parameters. Mild residual obstruction entailed the presence of mild flow acceleration between 1.2 and 1.7 m/s with preservation of cardiac/respiratory variations (i.e. biphasic flow curve returning to baseline). Conversely, moderate residual obstruction involved the presence of flow velocities of 1.8 m/s or greater, with loss of dynamic variations (i.e. continuous flow), or complete occlusion of any of the PVs. Echocardiographic findings of residual obstruction were interpreted in the context of other clinical manifestations such as congestive heart failure (CHF), chest radiograph (CXR) findings, and evidence of pulmonary hypertension. Postoperative management included various combinations of inotropes and pulmonary vasodilators (i.e. nitric oxide), as needed. Patients with echocardiographic evidence of residual obstruction after reoperation were subjected to cardiac catheterisation in preparation for additional reoperations.
|
3. Results |
|---|
|
TopAbstract1. Introduction2. Methods3. Results4. DiscussionAppendix A. Conference...References |
|---|
|
|
3.2. Mortality
The overall mortality was 25% (5/20). Of the five patients who did not survive re-intervention (Table 1), two received adequate relief of anastomotic obstruction but succumbed in the immediate postoperative period to severe pulmonary hypertension (#10 and #17). They both had supracardiac TAPVC and both presented at approximately 3 months after primary repair. The remaining three suffered from various combinations of postoperative residual obstruction and pulmonary hypertension (#4, #7, and #9) (Tables 1 and 2). Patient #4 had two reoperations (enlargement with native atrial tissue and Gore-tex) for anastomotic and ostial obstruction of the right-sided PVs, after which mild obstruction and pulmonary hypertension (0.7 systemic) persisted. He expired 3 weeks postoperatively with CHF. Patient #7 initially presented with anastomotic PVO, which was corrected by Gore-tex patch enlargement. He then developed obstruction of the right-sided PVs, so, a second reoperation was required 4 months later. Bovine pericardium was used for the repair. Postoperatively, some degree of obstruction persisted and RV pressures remained elevated (0.8 systemic). He expired 3 months later with CHF. A third patient (#9) underwent reoperation for anastomotic stenosis using a Gore-tex patch. The obstruction was only partly relieved, and rapidly progressed to involve the right-sided PVs. RV pressures remained moderately elevated (0.5 systemic). As additional surgery was being postponed for his very poor clinical condition, he succumbed 3 months later to CHF and RSV bronchiolitis.
3.3. Postoperative course and follow-up
Mean length of hospital stay was 14.9 days (range 6–28). Excluding death and failure to relieve the obstruction, the only notable complication observed was diaphragmatic paralysis. This was unilateral in three patients and bilateral in one, and was treated by diaphragmatic plication. Other details regarding the outcome of our patients are shown in Table 1.
Follow-up ranged from 1 month to 15 years (mean 26.8 months), and was limited in four patients (Table 1). As Table 3 shows, residual obstruction after reoperation was noted in seven patients (7/20; 35%); it was mild in three and moderate in four.
3.4. Risk factors for death
As Table 5 shows, mode of presentation, site of obstruction, RV pressure prior to reoperation, and number of reoperations did not appear to have an impact on survival. Conversely, a trend for a higher mortality was noted in patients with residual obstruction in those who received Gore-tex as part of the repair, and especially in those who presented early (<6 months) after primary TAPVC repair. Results were especially poor in patients who presented within the first 3 months of TAPVC correction, as mortality approached 80% (4/5) (Table 5). Similarly, the presence of postoperative pulmonary hypertension (RVP 
0.5 systemic) strongly influenced survival. As Table 5 shows, all of the 13 patients with a postoperative RV pressure of less than 0.5 systemic survived, whereas 71% (5/7) of those with a postoperative RV pressure greater than 0.5 did not. The impact of residual obstruction and pulmonary hypertension on outcome is explained further in Table 6. Unfortunately, our series was too small to allow a meaningful statistical comparison.
|
|
|
|
4. Discussion |
|---|
|
TopAbstract1. Introduction2. Methods3. Results4. DiscussionAppendix A. Conference...References |
|---|
Although in recent years knowledge of this disease has accumulated and surgical outcomes have improved, PVO after correction of TAPVC remains incompletely understood. In 1996, Bando et al. [7] reported eight patients who developed PVO after TAPVC repair and were managed surgically; five of them survived the operation (mortality 37%), while three succumbed to pulmonary hypertensive crises. The authors maintained that the presence of a small pulmonary venous confluence at primary correction strongly predicted both early death and need for additional procedures due to subsequent PVO. Although their technique of TAPVC repair was similar to others and that of ours, none of the eight patients presented with anastomotic stricture but all had diffuse involvement of the PVs. Most notably, these findings contrasted with those of others, who emphasised the importance of risk factors such as anatomic arrangement of TAPVC, presence or absence of obstruction, diameter of individual PVs, suture material, and surgical technique at primary repair [4].
Similar data were recently produced by van Son et al. [4] and Hyde et al. [3]. Both studies suggested that patients with ostial PVO tended to present earlier (<6 months) and to have a worse prognosis than those with anastomotic PVO. A larger series was reported by Lacour-Gayet et al. [6] who described 16 such patients, seven of whom were treated by a sutureless in situ pericardial repair. The overall mortality was 31% (5/16), and the only significant risk factor for death appeared to be the presence of bilateral vs. unilateral PVO. Their findings seem to suggest that surgical results had improved when their innovative surgical technique was used.
Although our data confirm these recent trends in terms of improved outcomes, some important differences were noted. Contrary to others [3], in our series, early presentation (<6 months) was not the most frequent mode of presentation, as many of our patients presented beyond 6 months after primary repair (Table 4). Also, the associations noted by others [3,4] between early PVO and involvement of the individual PVs and that between late PVO and anastomotic stenosis were not validated by our study. In fact, our observations contradicted this view, as both types of PVO presented either early or late (Table 4). However, when PVO was diagnosed early after primary correction, irrespective of its extent, it did generally predict a poor outcome (Table 5).
With regard to the progression of disease from the anastomosis to the individual PVs, we observed this phenomenon in four of the ten patients who initially presented with isolated anastomotic stenosis (Tables 1 and 3). We agree with others [6] that in some patients, there seems to be a pattern of progression in which PVO begins at the anastomosis and subsequently extends peripherally to involve the wall of individual PVs. This progression may occur either before reoperation, in which case patients present with a mixed picture of anastomotic and ostial obstruction (six of our patients), or after reoperation, as patients may develop ostial PVO after correction of isolated anastomotic stenosis (four of our patients) (Table 3). At this time, the mechanisms responsible for this phenomenon remain unclear and the course of events is unpredictable. It is unknown whether traumatic manipulations of the PVs at surgery or ‘minimal’ obstructive changes, occasionally seen after TAPVC repair, could play a role in inducing or facilitating this phenomenon, and if so, to what extent. It could be hypothesised that the abnormal architecture of the pulmonary vasculature typically observed in patients with TAPVC [4,8] could render these vessels more vulnerable to developing intimal hyperplasia, and ultimately obstruction, in response to distal anastomotic stenosis. Such structural changes involving the intima and media of pulmonary vessels are thought to be reversible after a successful primary repair. However, wall thickening could persist and evolve to cause obstruction in the presence of anastomotic stenosis and increased pulmonary venous pressure. This hypothesis would be substantiated by the observation that, at least in our series, narrowing of PVs was very rare in isolation. In contrast, it was nearly always found concomitantly with anastomotic stenosis (Table 3).
When this progression of disease occurs, patients may require cumbersome re-interventions to relieve PVO that progressively extends away from the anastomosis. Although one would expect less favourable surgical results in this setting, the need for multiple re-interventions does not seem, by itself, to preclude survival, as also shown by others [6]. Thus, it has become our policy to aggressively attempt to relieve PVO, even if this requires numerous and difficult surgical attempts.
As the pattern of PVO was diverse, surgical techniques were individualised to each patient, and definitive conclusions regarding their efficacy were impossible to draw. Generally, we preferred to enlarge the pulmonary venous pathway using native atrial tissue whenever possible, as described by others [5,9]. Unfortunately, our experience with the sutureless repair using in situ pericardium, as advocated by Lacour-Gayet et al. [6], was limited and we cannot comment on its validity. Nonetheless, preliminary evidence suggests that this strategy may be useful, especially in patients presenting with PVO that involves individual PVs, particularly on the right side. In one of our patients, we used an alternative technique in which the sutureless repair was carried out with a free pericardial patch. This was sewn away from the edges of the pulmonary venous pathway so as to avoid a suture line directly in contact with the pulmonary venous endothelium, and possibly reduce the risk of subsequent re-stenosis.
In eight of our patients, the repair was difficult, and a Gore-tex patch was used to enlarge the pulmonary venous pathway. Although this strategy did not seem to alter either the need for multiple re-interventions or the risk for postoperative residual obstruction, it is not known whether these findings would have reached significance in a larger series of patients. The trend for a higher incidence of postoperative residual obstruction and mortality noted in patients receiving Gore-tex remains a concern. Based on our and others' observations [4], we now advise against the use of any artificial material, favouring whenever possible the use of native tissue (i.e. native atrium, sutureless repair with in situ or free pericardium).
Although the results of re-interventions for PVO have improved in recent years [3,4,7], there is still limited understanding of the factors associated with failure. This is probably due to the fact that the disease is rare, and reported series are small [1,3,4,7]. Unfortunately, the small sample size also limited the power of our review. Contrary to others [3–5], our data failed to show a clear negative prognostic significance of ostial PVO as compared to anastomotic PVO. In fact, the former did not seem to clearly hinder survival, nor did it expose patients to a greater need for multiple re-interventions. These data, however, should be interpreted with caution as, in three of the 13 patients with ostial PVO, the follow-up period was limited. Conversely, early presentation (<6 months) was found to be a strong risk factor for death, irrespective of the extent of PVO.
With regard to the impact of residual obstruction and postoperative pulmonary hypertension on outcome, their importance has been recognised by others [6]. In some patients, it may be difficult to establish whether failures are due to incomplete relief of PVO, postoperative pulmonary hypertension with RV failure, or a combination of these [10]. Not surprisingly, our data show that failure to completely relieve PVO at surgery tends to adversely affect survival. However, the impact of postoperative pulmonary hypertension on outcome seems to be greater. While one would expect pulmonary hypertension to be the consequence of residual obstruction, this does not always appear to be the case. Our study suggests that postoperative pulmonary hypertension can occur independently, irrespective of the presence of residual obstruction, and have a major impact on outcome. This was the case in at least four of our patients, all of whom succumbed in the absence of significant residual PVO.
Conversely, four other patients, in whom the presence of some degree of obstruction of the pulmonary venous system (mild or moderate) did not result in severe pulmonary hypertension, all survived. Furthermore, the longest survivor in our series is a patient with totally occluded left-sided PVs, but without pulmonary hypertension, who has remained in good health 15 years following surgery. Thus, these findings support the observation that the occlusion of one or even two PVs may be compatible with long-term survival, if it does not result in an excessive raise in pulmonary artery pressure. They also emphasise the importance of achieving a low pressure pulmonary vascular bed, as already proposed by others [10].
Unfortunately, our study presents several limitations. It is a retrospective review of a small group of patients treated over a considerable period of time. The limited number of patients precluded meaningful statistical comparisons among patients. Thus, the impact of individual risk factors on outcome was difficult to assess. As surgical techniques and the perioperative care of these patients have evolved over the years, an important bias might have been introduced. Also, surgical techniques were so diverse, and in many cases used concomitantly, that their impact on outcome could not be established.
In summary, our findings support the conclusion that early presentation and postoperative pulmonary hypertension have the greatest adverse impact on outcome. Of these, failure to achieve a low-pressure pulmonary vascular system seems to be the variable that most strongly prevents survival. In our series, neither ostial PVO nor multiple re-interventions appeared to increase surgical risk significantly. The negative impact of postoperative residual obstruction on outcome was not striking. However, an aggressive surgical approach to this disease is still warranted. Although the role of each technique in obtaining long-lasting relief of PVO remains to be established, the use of artificial material seems unwise.
|
Acknowledgments |
|---|
Research at the Institute of Child Health and Great Ormond Street Hospital for Children NHS Trust benefits from Research and Development funding received from the NHS Executive.
|
Footnotes |
|---|
|
Appendix A. Conference discussion |
|---|
|
TopAbstract1. Introduction2. Methods3. Results4. DiscussionAppendix A. Conference...References |
|---|
Dr Ricci: Well, I think that's more than a question. It's perhaps a statement. I agree with you that there might be some predominance in the infracardiac type, but I think our series as well as perhaps others that have been published recently clearly show that this disease can occur after any type. One of the reasons, or at least even from the literature, is that perhaps the infracardiac type is more commonly associated with a small common confluence, and that perhaps might be one of the factors responsible for that. It's a good point.
Dr H. Najm (Riyadh, Saudi Arabia): In your analysis, you looked at the ostial lesions, as we know, ostial lesions are usually isolated and the ostium of four veins do not get obstructed together. Have you looked at unilateral obstruction versus bilateral obstruction? And have you seen progression from a unilateral obstruction to bilateral and what is the outcome in each subset?
Dr Ricci: That's another very good point. I know some authors previously have made a distinction. We haven't done it. We have seen that 17 or 18 of our patients out of 20 presented with bilateral disease, either because it was involving the anastomosis, which in that case you would have obstructed venous flow from both lungs, or because it did involve, for example, one pulmonary vein on the right and the other one on the left. So we did not look at that because essentially our data were not sufficient.
|
References |
|---|
|
TopAbstract1. Introduction2. Methods3. Results4. DiscussionAppendix A. Conference...References |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Takabayashi, H. Shimpo, K. Yokoyama, M. Kajimoto, and K. Onoda Congenital Pulmonary Vein Stenosis with Anomalous Pulmonary Venous Connection Asian Cardiovasc Thorac Ann, October 1, 2007; 15(5): 438 - 440. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Y. Alton, C. M.T. Robertson, R. Sauve, A. Divekar, A. Nettel-Aguirre, S. Selzer, A. R. Joffe, I. M. Rebeyka, D. B. Ross, and Western Canadian Complex Pediatric Therapies Proje Early childhood health, growth, and neurodevelopmental outcomes after complete repair of total anomalous pulmonary venous connection at 6 weeks or younger J. Thorac. Cardiovasc. Surg., April 1, 2007; 133(4): 905 - 911. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. J. Devaney, A. C. Chang, R. G. Ohye, and E. L. Bove Management of Congenital and Acquired Pulmonary Vein Stenosis. Ann. Thorac. Surg., March 1, 2006; 81(3): 992 - 996. [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 |
