Total Anomalous Pulmonary Venous ConnectionClinical Perspective
Morphology and Outcome From an International Population-Based Study
Background—Late mortality after repair of total anomalous pulmonary venous connection is frequently associated with pulmonary venous obstruction (PVO). We aimed to describe the morphological spectrum of total anomalous pulmonary venous connection and identify risk factors for death and postoperative PVO.
Methods and Results—We conducted a retrospective, international, collaborative, population-based study involving all 19 pediatric cardiac centers in the United Kingdom, Ireland, and Sweden. All infants with total anomalous pulmonary venous connection born between 1998 and 2004 were identified. Cases with functionally univentricular circulations or atrial isomerism were excluded. All available data and imaging were reviewed. Of 422 live-born cases, 205 (48.6%) had supracardiac, 110 (26.1%) had infracardiac, 67 (15.9%) had cardiac, and 37 (8.8%) had mixed connections. There were 2 cases (0.5%) of common pulmonary vein atresia. Some patients had extremely hypoplastic veins or, rarely, discrete stenosis of the individual veins. Sixty (14.2%) had associated cardiac anomalies. Sixteen died before intervention. Three-year survival for surgically treated patients was 85.2% (95% confidence interval 81.3% to 88.4%). Risk factors for death in multivariable analysis comprised earlier age at surgery, hypoplastic/stenotic pulmonary veins, associated complex cardiac lesions, postoperative pulmonary hypertension, and postoperative PVO. Sixty (14.8%) of the 406 patients undergoing total anomalous pulmonary venous connection repair had postoperative PVO that required reintervention. Three-year survival after initial surgery for patients with postoperative PVO was 58.7% (95% confidence interval 46.2% to 69.2%). Risk factors for postoperative PVO comprised preoperative hypoplastic/stenotic pulmonary veins and absence of a common confluence.
Conclusions—Preoperative clinical and morphological features are important risk factors for postoperative PVO and survival.
Total anomalous pulmonary venous connection (TAPVC) is a rare form of congenital heart disease in which all pulmonary veins connect to the systemic veins, right atrium, or coronary sinus. TAPVC can occur in conjunction with a wide variety of cardiac malformations, especially atrial isomerism. Before the advent of cardiac surgery, almost all of these children died in the first few months of life. With treatment, there has been continued improvement in the mortality and morbidity of isolated TAPVC.1,–,4 This can be attributed to progress in surgical expertise and to developments in intensive care such as use of nitric oxide and extracorporeal membrane oxygenation, which have led to salvage of the sickest neonates.
Clinical Perspective on p 2726
Postoperative PVO may be a consequence of an inadequate anastomosis between the left atrium and pulmonary venous confluence, inadequate postoperative growth of the surgical anastomosis, or a reaction of tissues to prosthetic material used. In some cases, the obstruction is proximal to the operative site.10,11 There may be preoperative pulmonary vein hypoplasia/stenosis, or such changes may develop postoperatively. Jenkins et al12 showed in neonates with TAPVC a strong association between the size of the individual pulmonary veins at diagnosis and survival. Haworth13 showed that in infracardiac TAPVC, prenatal structural changes in the extrapulmonary veins may predispose to the later development of PVO despite a successful surgical repair. Other groups have suggested that a small left atrium14 or confluence morphology1,15 may predispose to subsequent PVO.
Owing to the relative rarity of the condition, previous studies have had to span a wide range of surgical eras to obtain sufficient patients for analysis. By designing an international population-based study, we have collected a large cohort of contemporary patients to examine the morphology, outcome, and risk factors for poor outcome.
The UK, Ireland and Sweden Collaborative Study of TAPVC is a retrospective population-based study of all identified cases in infants born between January 1, 1998, and December 31, 2004, in the United Kingdom, Republic of Ireland (hereafter denoted as Ireland), and Sweden. Prenatally diagnosed cases whose estimated dates of delivery were January 1, 1998, through December 31, 2004, were also collected. Cases with functionally univentricular circulations or atrial isomerism were excluded, because these additional lesions have a highly significant influence on overall outcome. Ethics committee approval was obtained.
A research fellow (A.N.S.) and nurse (L.S., S.J.) visited all 17 pediatric cardiology centers in the United Kingdom and Ireland on multiple occasions to collect patient data. Cases were identified by use of hospital databases and intensive care and operative records. Particular care was taken to identify those cases who had died before surgery. Our data were cross-checked against the United Kingdom PICANet (Paediatric Intensive Care Audit Network) database and mortality data from the Office of National Statistics. In addition, local pathologists were contacted. Clinical information was collected from the 2 Swedish centers in similar fashion by the 2 Swedish coauthors (J.S., K.H.).
Clinical information was extracted from the medical records and entered into a customized database. A pro forma was devised to ensure homogeneity of the data collected, and 2 authors (A.S., P.E.F.D.) visited the collaborators in Sweden to ensure standardized data collection.
Preoperative and postoperative morphological features of each case were determined by direct review of all available imaging and reports by the first author (A.N.S.) and a cardiac radiologist (J.P.). Directly reviewed data comprised 83% of all angiography, 67% of all magnetic resonance imaging and computed tomography angiography, and 25% of all initial echocardiography. Digital subtraction was performed on all suitable angiograms (90%).
The pulmonary venous connection was considered to be obstructed on echocardiography if the pulsed-wave pulmonary venous Doppler pattern showed nonphasic flow or velocities >2 m/s.16 Angiographic diagnosis of pulmonary vein stenosis was made either from the levophase of pulmonary arterial injection or from direct injection of contrast into the individual pulmonary veins. Pulmonary veins were considered to be obstructed if the pulmonary vein diameter was reduced by 50% or more from the largest measured dimension. Slow transit of contrast through the pulmonary vascular tree was also taken as an indicator of PVO. The hemodynamic criterion for PVO was a mean gradient across the stenosis of 4 mm Hg or more.7 All available pathological reports were also studied.
On the basis of the evaluation of all available data (including operative reports), patients were classified as either having or not having evidence of PVO at presentation. A severity grade was ascribed as follows: 1, Severe obstruction requiring immediate intervention; 2, obstruction present but clinically stable, not requiring immediate intervention; and 3, no significant obstruction. Preoperative morphological features assessed comprised (1) presence of associated congenital heart lesions, (2) site to which the anomalous pulmonary veins connected (with mixed connection defined as drainage of pulmonary veins or confluence to more than 1 anomalous site), (3) presence of a common confluence into which all the pulmonary veins drained, (4) presence of preoperative obstruction, (5) presence of diffuse hypoplasia of the individual pulmonary veins, and (6) any other unusual morphological features.
All patients who had surgical repair were assessed for postoperative PVO with the same criteria described above. Pulmonary hypertensive episodes that led to cardiac arrest or the need for mechanical circulatory support were noted.
Continuous risk factors are expressed as median (limits), and categorical risk factors are expressed as number (%). The Poisson regression model was used to compare incidence across the 3 countries. Patient survival was described with Kaplan-Meier curves. Outcome measures examined were death and presence of postoperative PVO. Cox proportional hazards modeling, stratified by country, was used to test the association between potential risk factors and these outcomes. Variables for the multivariable analysis were chosen if P<0.1 on univariable analysis and there were <5% missing data. P<0.05 was considered significant (STATA 10).
Four hundred twenty-two live births with TAPVC were identified: 367 in the United Kingdom, 37 in Sweden, and 18 in Ireland. This equates to an incidence of 7.1 per 100 000 live births. Median birth weight was 3.28 kg (limits 0.96, 5.3 kg); 266 (63%) were male.
There was no difference in incidence of TAPVC between the United Kingdom and Sweden, whereas Ireland had a significantly lower incidence, with an incidence rate ratio of 0.6 (95% confidence interval [CI] 0.37 to 0.96, P=0.032).
Two cases were diagnosed prenatally but died in utero; neither had a postmortem examination. Of the live-born cases, there was prenatal suspicion of a cardiac lesion in 10 (2.4%).
The most common symptoms were cyanosis (n=182, 43.1%), respiratory distress (n=134, 31.8%), failure to thrive (n=47, 11.1%), circulatory collapse (n=19, 4.5%), murmur (n=2, 0.5%) and supraventricular tachycardia (n=2, 0.5%). The remaining cases either were identified incidentally (n=14, 3.3%) or the details of presentation were unknown (n=12, 2.8%). Only 165 (39%) presented after birth but before hospital discharge; there were no cases in which the diagnosis was only made at postmortem examination.
Table 1 describes additional features at presentation of the 422 live-born cases. Seventeen patients (4%) had a syndrome, cat's-eye being the most common. All subsequent descriptions and analyses refer to the cohort of live-born patients.
Range of Morphology
Associated Cardiac Abnormalities
Sixty patients had 1 or more cardiac lesions in addition to TAPVC (Figure 1). The vast majority of these were small ventricular septal defects that did not require any additional intervention, but 26 did.
In 19 of the 26 cases that required intervention, there was complete repair of the associated cardiac lesion at the time of TAPVC repair. In 7 of the 26, the associated cardiac lesions were repaired (or were planned to be repaired) in a separate procedure. In 5, pulmonary artery banding with or without aortic arch repair was performed at the time of TAPVC repair. TAPVC repair only was performed on 1 child with associated tetralogy of Fallot; in another child, coarctation and ventricular septal defect were repaired before TAPVC surgery.
Pulmonary Venous Connection
Two hundred five neonates (48.6%) had supracardiac pulmonary venous connection, 110 (26.1%) had infracardiac pulmonary venous connection, 67 (15.9%) had cardiac pulmonary venous connection, and 37 (8.8%) had a mixed connection. There were 2 cases (0.5%) of common pulmonary vein atresia in which no connection to the systemic venous circulation could be identified. In 1 case (0.2%), no information on the type of connection was available.
Of the 205 patients with the supracardiac type, there were 150 (73.2%) with connection to the innominate vein, 43 (21.0%) with connection to various parts of the superior vena cava, 6 (2.9%) with connection to the azygos vein, and 6 (2.9%) unknown. Among the 67 with a cardiac type, there were 58 patients (86.6%) with connection to the coronary sinus, 8 (11.9%) with connection to the right atrium, and 1 (1.5%) unknown. Patients with mixed connections usually had a combination of cardiac and supracardiac connections with no common pulmonary venous confluence; however, in 5 cases there was a common confluence that connected to more than 1 anomalous site. Infracardiac connection was to the portal system in the vast majority of cases, but data on the exact connection were not always available. Two cases were identified in which the confluence connected directly to the inferior vena cava, 1 above the diaphragm and 1 below. There were 5 cases in which the right and left descending veins joined below the diaphragm, with no intrathoracic common confluence.
Both cases of common pulmonary vein atresia presented immediately after birth, cyanosed and acidotic. Surgery was performed as an emergency procedure, and in both cases, the surgeon found no connection between the confluence and the systemic venous system, and the individual pulmonary veins were small. Both neonates died due to severe postoperative pulmonary hypertension despite an adequate surgical anastomosis. Postmortem examination in 1 showed diffuse hypoplasia of the pulmonary veins and pulmonary lymphangiectasia.
Obstruction was either within the anomalous connecting vein or at its connection to the systemic circulation. Rarely, it was at the entrance of an individual pulmonary vein(s) to the confluence (see below).
Among the whole group, 215 (50.9%) of 422 cases had obstruction to pulmonary venous return and 197 (46.7%) did not. In 10 (2.4%), it was unclear whether obstruction was present. Of those having surgical repair, preoperative obstruction was graded as severe in 32.3%, present but not severe in 18.7%, absent in 47.3%, and unknown in 1.7%. Median age at presentation was 3 days (0, 18 days), 25 days (11, 65 days), and 38 days (2, 90 days), respectively.
On multivariable analysis, infracardiac connection was a significant risk factor for preoperative PVO (odds ratio 2.21, 95% CI 1.16 to 4.21, P=0.016). A cardiac connection appeared protective (odds ratio 0.45, 95% CI 0.22 to 0.93, P=0.03). Syndromic patients were less likely to be obstructed (odds ratio 0.29, 95% CI 0.09 to 0.96, P=0.04). In such patients, TAPVC was frequently discovered early during assessment for multisystem disease.
Abnormalities of the Individual Pulmonary Veins Before Surgery
Pulmonary venous hypoplasia is a challenging diagnosis preoperatively and was only recorded when unequivocally present; it was found in 23 of 422 patients (5.5%). In 3, the hypoplasia was extreme and deemed inoperable. Of the 20 of 23 patients who underwent surgical repair, 12 (60%) died. Nine (45%) developed postoperative PVO, and all but 2 died. Seven of the 23 cases with severely hypoplastic pulmonary veins underwent postmortem examination or lung biopsy and were reported to have diffuse thickening of the vessel walls, including 5 with pulmonary lymphangiectasia.
Discrete stenosis of a pulmonary vein as it entered the confluence was never diagnosed before surgery. In 5 cases (1.2%), the surgeon described a discrete narrowing of 1 or more pulmonary veins as they entered the common pulmonary venous confluence. Two of these cases went on to develop postoperative PVO and died.
Morphology of the Pulmonary Venous Confluence
There was great heterogeneity in the size, orientation, and shape of the pulmonary venous confluence; several of these variations are illustrated in the online-only Data Supplement Figure.
Outcome Before Surgery
Sixteen of the 422 cases (3.8%) did not undergo surgical repair (Figure 1): 8 supracardiac (50%), 5 infracardiac (31%), 2 mixed (13%), and 1 cardiac (6%). There were 4 main reasons for this: (1) Condition at presentation too poor to survive surgery; (2) irreparable anatomy, such as very hypoplastic pulmonary veins; (3) significant extracardiac conditions; and (4) those with significant associated cardiac abnormalities (eg, tetralogy with cardiac ectopia). There was overlap between the first 2 groups because patients with pulmonary venous hypoplasia presented in very poor condition; the local teams judged the situation to be futile. These 16 patients died between the first and 136th day of life (median 3 days).
Four hundred six of the 422 patients had a surgical repair. Twenty-six required additional procedures for other forms of congenital heart disease (Figure 1; online-only Data Supplement Table I).
Surgery was performed in multiple institutions by many different surgeons, so there was variable surgical technique. For infracardiac and supracardiac TAPVC, repair usually involved lifting the heart to the right and performing a side-to-side anastomosis of the pulmonary venous confluence to the morphological left atrium. In 36 cases, a right atrial approach was taken. We found no significant difference in outcome between the 2 approaches (P=0.7). Repair of TAPVC to the coronary sinus invariably involved unroofing the coronary sinus, followed by pericardial patch closure of the atrial septal defect that committed the pulmonary veins to the left atrium. It was not always clear whether any of the cardiac veins were committed to the right atrium once the coronary sinus ostium was closed. There was marked variation in the procedures used to repair mixed TAPVC because of the diverse morphology.
Surgical Outcomes and Risk Factors for Death
Seventy-four (17.5%) of the 422 patients with TAPVC died. Four hundred six underwent surgical repair, of whom 58 died (14.3%).
Median follow-up of all 406 cases undergoing surgery was 2.9 years (1 day to 8.4 years); 3 patients were lost to follow-up. Estimates of 30-day, 1-year, and 3-year survival for all surgically treated patients were 94.3% (95% CI 91.5% to 96.2%), 85.9% (82.1 to 89.0%), and 85.2% (81.3% to 88.4%), respectively (Figure 2). Only 6 patients died beyond 6 months after TAPVC repair, and none died after 3 years. Two patients died of causes unrelated to the heart.
Three patients had failure to wean from bypass and died in the operating room. Nineteen patients (4.7%) underwent postoperative extracorporeal membranous oxygenation or left ventricular assist device implantation, of whom 8 (42%) survived (median time on support 6 days, limits 2 to 20 days). Details of the group who required either postoperative mechanical support or who failed to wean from bypass are shown in Figure 3. Postoperative complications and their frequency are shown in Table II in the online-only Data Supplement. In addition, 5 patients with TAPVC to the superior vena cava developed postoperative superior vena cava obstruction. Median length of stay in the cardiac center was 13 days (4 to 154 days). Risk factors for death by univariable and multivariable analysis are shown in Table 2.
Postoperative Pulmonary Hypertension
There was variable documentation of the pulmonary artery pressure after bypass and of the presence and severity of pulmonary hypertensive crises. Thirty of the 406 cases (7.4%) had severe pulmonary hypertension that led to cardiac arrest or need for mechanical circulatory support. There was a small subgroup of 7 patients (1.7%) with an adequate anastomosis but postoperative pulmonary hypertension refractory to any form of treatment; all died.
Seventy-one of the 406 patients undergoing surgery (17.5%) had evidence of postoperative PVO. Of these 71, 44 (62%) also had preoperative PVO. In most cases, the morphology was complex, with narrowing at the anastomosis between the confluence and the left atrium resulting in retraction and fibrosis of the surrounding tissues, which produced narrowing at the entrance of the pulmonary veins.
Presentation, morphology, intervention, and outcome of the individual interventions will be detailed in a forthcoming report (not yet published). For cases with postoperative PVO, 30-day, 1-year, and 3-year survival rates were 95.8% (95% CI 87.5% to 98.6%), 62.0% (95% CI 49.7% to 72.1%), and 58.7% (95% CI 46.2% to 69.2%), respectively. Risk factors for presence of postoperative PVO are shown in Table 3.
The UK, Ireland and Sweden Collaborative Study of TAPVC is an international population-based study that allows analysis of contemporary data in 3 defined geographic areas. The present findings reflect the entire spectrum of morphology and represent current practice.
We report a higher incidence of TAPVC than previous studies at 7.1 per 100 000 live births compared with 5.9 in the New England Regional Infant Cardiac Program17 and 6.8 in the Baltimore-Washington Infant Study.18 Despite great care with case identification, it is still likely that a small number of cases were missed, and in some, a diagnosis may never have been made, either in life or death.19,20 It is unclear whether the lower incidence in Ireland was true or due to incomplete collection.
The present study shows great morphological heterogeneity and provides an accurate assessment of how frequently each subtype occurs in a population. There was variation in (1) associated congenital heart disease; (2) size, shape, and position of the pulmonary venous confluence; (3) number and size of the individual pulmonary veins; (4) connection to the systemic circulation; and (5) severity of preoperative PVO. These factors will significantly affect the complexity of the surgical procedure.
The present findings highlight the morphological variation of the pulmonary venous confluence. Data were not available in all cases to describe all the anatomic subtypes and their incidence accurately. This could be assessed in a prospective study, especially with the advent of 3-dimensional reconstructions from magnetic resonance and computerized tomography. These techniques were just becoming established at the earliest part of the study but can provide surgeons with better detail of complex anatomy.21
Do the individual pulmonary veins have discrete stenoses before surgery? We were unable to find convincing evidence to support this, although again, this may simply reflect limitations of the imaging techniques used. Certainly, surgeons described discrete stenosis in 5 cases. In contrast, diffuse hypoplasia was diagnosed both before and during surgery.
Overall Survival and Risk Factors for Death
Karamlou et al3 recently reported an operative 5-year survival of 97% for patients undergoing repair at the Hospital for Sick Children in Toronto since 2000. The 3-year operative survival reported in the present study was seemingly poorer at 85%; this may reflect the population basis rather than institutional basis for the present study, which comprised the full range of morphology and/or surgical and medical management.
In common with the Toronto study,3 we also show the importance of preoperative morphology on outcome. Associated heart lesions for which definitive repair was not performed at the time of TAPVC surgery were an independent risk factor for death. By contrast, when the associated heart lesion was repaired during TAPVC surgery, it was not a risk factor for death in the multivariable analysis.
Preoperative PVO forms a clinical spectrum, with some patients having mild obstruction and others having severe obstruction with marked pulmonary edema. We differentiated these groups using a simple grading system and found severe preoperative PVO to be a risk factor for both death and postoperative PVO in the univariable analyses. Younger age at surgery was a significant risk factor for death in the multivariable analysis; this could be considered a surrogate for preoperative obstruction. Many others3,22,–,24 have also found that PVO at presentation is associated with an increased risk of death.
We can speculate as to why preoperative obstruction has an influence on postoperative outcome. Previous studies have shown that pulmonary vein stenosis is a progressive disease.9,25,–,27 We recently showed that a discrete obstruction of the pulmonary veins leads to pulmonary venous remodeling, small pulmonary veins, and even development of pulmonary vein atresia.28 TAPVC may be obstructed in utero. If this occurs early in gestation, generalized hypoplasia of the pulmonary vascular bed may result, with arterialization of the pulmonary veins and pulmonary lymphangiectasia. This may have accounted for those cases in which tiny veins were seen at surgery. Haworth13 has shown that in pathological studies of obstructed infracardiac TAPVC, the entire pulmonary venous system may be small at birth. Seven of the cases in the present study with severely hypoplastic pulmonary veins were reported to have diffuse thickening of the vessel walls, and in 5, pulmonary lymphangiectasia was also present at postmortem examination or on lung biopsy.
In common with a previous report,1we found a subgroup of patients who presented in the first few days of life in poor condition and who developed severe postoperative pulmonary hypertension unresponsive to treatment despite an apparently adequate anastomosis. These patients often had diffusely small pulmonary veins at surgery and a lung biopsy that showed lymphangiectasia. We can speculate that these patients had in utero obstruction that led to progressive changes in the vascular bed. During gestation, the lymphatic system often regresses after approximately 20 weeks29 unless there is obstruction to the pulmonary veins. This would suggest that in cases in which pulmonary lymphangiectasia is seen, fetal PVO has been present from before 20 weeks' gestation.
In some patients, a residual atrial septal defect and/or the vertical vein was left open; these patients did not fare as well. This may reflect a raised pulmonary artery pressure, possibly secondary to PVO; the decision at surgery was therefore to leave a residual right-to-left shunt to improve cardiac output.
Postoperative PVO was an independent risk factor for late death, with 40% of patients dying by 3 years. Preoperative morphological features of the pulmonary veins and confluence were the most important risk factors for the presence of such postoperative obstruction, which confirms the results of other studies.1,12,14,15 Hancock Friesen et al7 demonstrated that none of the current perioperative or operative strategies have had an impact on the incidence of postrepair PVO, which implies that the process may reflect an underlying predisposition.
The present findings concur with those of Bando et al,1 who found that a small venous confluence and diffuse pulmonary venous narrowing were risk factors for adverse outcome. We found that patients who did not have a single confluence were at greater risk of postoperative PVO. Most cases of mixed TAPVC fall into this category. Karamlou et al3 also found that this group of patients was more likely to require reoperation.
Clinicians should be aware that patients with mixed connection and/or hypoplastic pulmonary veins and confluence are particularly at risk of postoperative PVO and should be cautious when predicting outcome. Such patients should be monitored carefully after TAPVC repair and actively investigated on the slightest suspicion of postoperative PVO. In some cases, a primary repair by use of a modified sutureless technique may be indicated.30
The present study is retrospective and dependent on the quality of data available in each cardiac center. The study was observational, with no attempt to standardize preoperative morphology or influence treatment strategies. Because the study was multi-institutional, there will have been variation in the management protocols and surgical technique. There was considerable variation in the morphology and complexity of the cases, and numbers in each center were small. This precludes meaningful comparison of outcomes between centers. The time scale of the study (1998–2004) meant that it predated the recent popularity of the modified sutureless technique8,9; analysis of this technique was therefore not possible. Despite overlapping data collection techniques, inevitably, some cases of TAPVC will have been missed. We estimate this effect to have been small, particularly because our documented incidence was higher than previous estimates. Finally, we had limited access to postmortem specimens and were therefore unable to study histology.
We have described the range of morphology, outcome, and factors predictive of poor outcome in one of the largest contemporary international, population-based studies of TAPVC. We have identified a cohort of patients with PVO after surgical repair and identified risk factors for this. PVO after repair of TAPVC remains a difficult and complex disease to treat, with 40% of patients dying by 3 years. Preoperative morphological features are important risk factors for its development; cases with mixed connection or hypoplastic pulmonary veins and confluence are particularly at risk.
Sources of Funding
This study was supported in part by the Harrison Heart Foundation and the Joe Gandon Memorial Trust.
We would like to thank in particular the following doctors: Nick Archer, David Barron, Kate Brown, Frances Bu'Lock, Janet Burns, Julene Carvalho, Brian Craig, Desmond Duff, Helena Gardiner, John Gibbs, Ian Huggon, Ann Karimova, Ed Ladusans, Nilima Malaiya, John O'Sullivan, Trevor Richens , Graham Stuart, John Simpson, Andrew M. Taylor, Dirk Wilson, Chris Wren, and Robert Yates. Participating institutions in this study include: Alder Hey Children's Hospital, Liverpool; Birmingham Children's Hospital, Birmingham; Bristol Royal Hospital for Children, Bristol; Cardiothoracic Unit, Southampton General Hospital, Southampton; Children's Hospital, Oxford; Evelina Children's Hospital, London; Freeman Hospital, Newcastle; Glenfield Hospital, Leicester; Great Ormond Street Hospital, London; Leeds General Infirmary, Leeds; Lund University Hospital, Lund; Our Lady's Hospital for Sick Children, Dublin; Queen Silvia Children's Hospital, Gothenburg; Royal Belfast Hospital for Sick Children, Belfast; Royal Brompton Hospital, London; Royal Hospital for Sick Children, Edinburgh; Royal Hospital for Sick Children, Yorkhill, Glasgow; Royal Manchester Children's Hospital, Manchester; The Children's Hospital for Wales, Cardiff.
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.110.940825/DC1.
A list of participating institutions is included in the acknowledgments.
- Received February 22, 2010.
- Accepted September 27, 2010.
- © 2010 American Heart Association, Inc.
- Bando K,
- Turrentine MW,
- Ensing GJ,
- Sun K,
- Sharp TG,
- Sekine Y,
- Girod DA,
- Brown JW
- Lamb RB,
- Qureshi SA,
- Wilkinson JL,
- Arnold R,
- West MA,
- Hamilton DI
- Karamlou T,
- Gurofsky R,
- Al Sukhni E,
- Coles JG,
- Williams WG,
- Caldarone CA,
- Van Arsdell GS,
- McCrindle BW
- van de Wal HJ,
- Hamilton DI,
- Godman MJ,
- Harinck E,
- Lacquet LK,
- van Oort A
- Hyde JA,
- Stumper O,
- Barth MJ,
- Wright JG,
- Silove ED,
- de Giovanni JV,
- Brawn WJ,
- Sethia B
- Haworth SG
- Ando M,
- Takahashi Y,
- Kikuchi T
- Smallhorn JF,
- Burrows P,
- Wilson G,
- Coles J,
- Gilday DL,
- Freedom RM
- Sinzobahamvya N,
- Arenz C,
- Brecher AM,
- Blaschczok HC,
- Urban AE
- Latson LA,
- Prieto LR
- Seale AN,
- Webber SA,
- Uemura H,
- Partridge J,
- Roughton M,
- Ho SY,
- McCarthy KP,
- Jones S,
- Shaughnessy L,
- Sunnegardh J,
- Hanseus K,
- Rigby ML,
- Keeton BR,
- Daubeney PE
Totally anomalous pulmonary venous connection is a cyanotic congenital heart disease with ongoing morbidity and mortality particularly related to pulmonary venous obstruction. Our study is the largest population-based study of the disease to date and reports on the incidence, morphology, outcome, and factors predictive of poor outcome for all babies born with this anomaly in a fixed geographic area (United Kingdom, Ireland, and Sweden). Importantly, our report represents an exhaustive collaborative effort of all pediatric cardiac centers in the United Kingdom, Ireland, and Sweden. Unlike other studies, the population-based nature of our data reflects the entire spectrum of morphology, as well as contemporary practice. We have identified a cohort of patients with pulmonary venous obstruction after surgical repair and identified risk factors for this. After reading the article, the clinician should be aware of the wide spectrum of complexity in this condition, should be better able to predict outcome for an individual patient with total anomalous pulmonary venous connection, and should understand which patients are particularly at risk of developing postoperative pulmonary venous obstruction.