|
|
||||||||
Ann Thorac Surg 2006;82:1629-1636
© 2006 The Society of Thoracic Surgeons
a Divisions of Cardiology and Cardiothoracic Surgery, The Cardiac Center at The Children's Hospital of Philadelphia, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania
b Departments of Pediatrics and Surgery, The Cardiac Center at The Children's Hospital of Philadelphia, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania
Accepted for publication May 11, 2006.
* Address correspondence to Dr Cohen, Division of Cardiology, The Children's Hospital of Philadelphia, 34th St and Civic Center Blvd, Philadelphia, PA 19104 (Email: cohenm{at}email.chop.edu).
| Abstract |
|---|
|
|
|---|
METHODS: Our database was reviewed from January 1996 to December 2004 for patients with heterotaxy syndrome. Assessment was made for prenatal diagnosis and echocardiographic characteristics of heterotaxy syndrome. We sought to assess the accuracy of fetal echocardiography in the diagnosis of heterotaxy syndrome and determine whether prenatal diagnosis and other risk factors have an impact on survival in patients with heterotaxy syndrome.
RESULTS: Of 81 patients that met criteria, 43 (53%) had prenatal diagnosis. Prenatal diagnosis had high specificity and positive predictive value for all findings but had low sensitivity for anomalous pulmonary veins. Among the 70 patients born alive, survival was 60% with median follow-up of 51.4 months (range, 6.5 to 109.7 months). Prenatal diagnosis did not improve survival (p = 0.09). None of the 11 patients with complete heart block (CHB) survived past 3 months of age. Two patients underwent heart transplantation as their first intervention and have survived. CHB and anomalous pulmonary venous connection were associated with shorter duration of survival.
CONCLUSIONS: Prenatal diagnosis of heterotaxy syndrome does not improve survival in patients who undergo single ventricle reconstruction. The most potent risk factors for poor outcome (CHB, anomalous pulmonary veins) are likely not impacted by identification in utero. In light of the poor outcome, cardiac transplantation as an initial therapy may be a viable option for some patients.
| Introduction |
|---|
|
|
|---|
Whether prenatal diagnosis of congenital heart disease has a beneficial impact on outcome has been controversial. Some studies have shown no incremental benefit in survival with prenatal diagnosis [1517]; however, recent reports suggest that significant morbidity such as metabolic acidosis can be avoided if a prenatal diagnosis is made [15, 18]. Prenatal diagnosis allows for prompt administration of prostaglandin when ductal-dependency is suspected. This may contribute to improved outcome for the most severe forms of congenital heart disease, as recently demonstrated for infants diagnosed prenatally with hypoplastic left heart syndrome [19]. However, factors specific to heterotaxy syndrome, such as complex anatomic substrate, congenital complete heart block (CHB), and extracardiac anomalies, may counteract these benefits and negatively affect survival regardless of prenatal recognition of disease [12, 2022].
The complexity of anatomy in heterotaxy syndrome, which commonly includes unbalanced atrioventricular canal with contralateral ventricular hypoplasia, complex systemic and pulmonary venous anatomy, and conotruncal abnormalities, frequently dictates a course of management that culminates in a single ventricle reconstruction. We hypothesized that patients with heterotaxy syndrome who require single ventricle reconstruction might have a better outcome when an accurate prenatal diagnosis is made.
With this in mind, we sought to assess the accuracy of fetal echocardiography in the diagnosis of heterotaxy syndrome, with particular emphasis on anatomic variables that may negatively affect surgical outcome, such as total anomalous pulmonary venous connection (TAPVC) and atrioventricular valve regurgitation. We also sought to determine whether prenatal diagnosis improved survival in this highest-risk group and to assess the risk factors for mortality.
| Patients and Methods |
|---|
|
|
|---|
The surgical and echocardiographic databases identified all patients with a postnatal diagnosis of heterotaxy syndrome in the same time period. The diagnosis in each patient was defined by evidence of combination of situs solitus and situs inversus or midline liver, or both; systemic or pulmonary venous anomalies, or both, characteristic of the disease; confirmation of splenic dysfunction after birth, when possible; and the association of common atrioventricular canal with conotruncal abnormalities [1, 2].
Because accurate prenatal determination of atrial appendage anatomy is often not possible, patients were designated as having asplenia or polysplenia syndrome from the typical characteristic findings of these two entities or splenic anatomy, or both. Patients were included if they had a diagnosis destined for a single ventricle reconstructive strategy with intent to perform a bidirectional Glenn shunt and subsequent Fontan completion
All fetal echocardiographic studies were performed using an Acuson Sequoia ultrasound system (Mountain View, CA) with appropriate transducers for the mother's size/body habitus and fetal gestational age. Postnatal studies were performed on a Philips 2500 5500 or 7500 ultrasound system (Andover, MA) using appropriate transducers for patient size. All prenatal and postnatal echocardiographic reports, surgical operative notes, and autopsy reports (when performed) were reviewed.
To determine accuracy of prenatal diagnosis, the presence of systemic or pulmonary venous anomalies, or both; systemic or pulmonary outflow tract obstruction, or both; and the presence of significant (moderate or greater) atrioventricular valve regurgitation were compared on the prenatal and postnatal echocardiogram for 43 subjects with a prenatal diagnosis. TAPVC was defined as a confluence connecting to an extracardiac site, and all cases had evidence of at least mild pulmonary venous obstruction at birth. Accuracy of prenatal prediction of postnatal ductal-dependent circulation to systemic or pulmonary circulations was also assessed. If the fetus survived to birth, medical records were reviewed for sex, type of surgery, survival, and date of last follow-up. The frequency of anatomic variables associated with heterotaxy syndrome was assessed on the prenatal and postnatal echocardiograms to determine if the populations were similar.
Statistical Analysis
To assess accuracy of prenatal diagnosis, sensitivity, specificity, and positive and negative predictive value were calculated for each identified characteristic finding. The postnatal echocardiograms and the surgical operative notes were used as the gold standard for comparison.
Univariate comparisons between subsets of the cohort were made using
2 tests or the Fisher exact test if any expected value was less than 5. Among the subjects born alive, we examined the relationship between survival and potential risk factors for poor outcome. Survival-time analysis was performed using the Cox proportional hazards model. We excluded prenatally diagnosed subjects who did not survive to birth from the survival analysis because of the unknown number of fetal deaths in those without prenatal recognition of heart disease.
For purposes of analysis, 2 patients who underwent cardiac transplantation were considered nonsurvivors because they had heart disease considered "untreatable" and would have died without a transplant. Duration of survival was defined as birth to death, cardiac transplantation, or last known follow-up. Variables were selected for testing in the Cox proportional hazard model from the results of the univariate analysis, clinical experience, and known associations in the literature. The potential risk factors assessed for the model were prenatal diagnosis, sex, type of first surgical intervention, and presence of the following cardiac abnormalities: CHB, TAPVC, type of outflow tract obstruction, dextrocardia, and moderate or greater atrioventricular valve regurgitation at the postnatal study.
Nested models were compared with the likelihood ratio test. The most parsimonious model that maintained maximal explanatory power was selected. Prenatal diagnosis was then added back to this model to determine whether it added any explanatory power.
Univariate analyses were performed using STATA 7.0 (STATA Corp, College Station, TX). Cox proportional hazard model was performed using SAS 9.1 (SAS Institute, Inc, Cary, NC). Statistical significance was defined as p
0.05.
| Results |
|---|
|
|
|---|
The spectrum of anatomic findings for prenatal and postnatal diagnosis is summarized in Table 1. TAPVC was more frequently observed in those with prenatal diagnosis (p = 0.05). CHB was also more common in the prenatally diagnosed group but did not reach statistical significance (p = 0.20).
|
|
2, p = 0.025). There were 11 fetal deaths: 7 died in utero and four pregnancies were terminated. Four of the 7 fetuses that died in utero had CHB, making it the only risk factor associated with fetal death (p = 0.02), and all 4 with CHB had hydrops fetalis. Only 1 fetus (with CHB) that died in utero had significant atrioventricular valve regurgitation. Six of the fetuses that died in utero had polysplenia syndrome (of which 4 also had CHB) and 5 had asplenia syndrome.
|
2, p = 0.05). Of the 29 patients who died after birth (15 with prenatal diagnosis, 14 with postnatal diagnosis), 6 (20.7%) died before cardiac surgical intervention (although 2 of these 6 had pacemaker implantation for CHB), 19 (65.5%) died after the initial surgical procedure, and 4 (13.8%) died after a bidirectional Glenn shunt. No late deaths occurred after the Fontan operation. No patient is known to have died from extracardiac disease, except for one who died shortly after birth. This patient had multiple congenital anomalies and the parents declined intervention. Of the survivors, 2 Fontan patients presently have mild pulmonary venous obstruction and 2 have not undergone bidirectional Glenn shunt because of severe abnormalities (biliary atresia, poor anatomic substrate for Fontan physiology) that preclude further cardiac surgical intervention.
|
Risk Factors Associated With Outcome
The Cox proportional hazard model, without adjusting for other risk factors, showed that prenatal diagnosis was not associated with longer duration of survival (p = 0.09; Fig 3). Univariate relationships between failure to survive and potential risk factors among patients who survived to term are summarized in Table 3. Significant atrioventricular valve regurgitation at birth and CHB were associated with poor outcome, whereas dextrocardia was associated with better survival.
|
|
|
| Comment |
|---|
|
|
|---|
Previous reports of prenatal diagnosis of heterotaxy syndrome have suggested that prognosis is poor despite recognition in utero [2729]. Lim and colleagues [29], who assessed a group of patients with left and right atrial isomerism, reported that prenatal diagnosis did not affect survival even if immediate and aggressive care was given after birth. Our study confirmed this: Patients with prenatal diagnosis did worse than those with postnatal diagnosis if fetal deaths were included. A fetus recognized as having heterotaxy syndrome prenatally may have more severe disease than an infant diagnosed postnatally. For example, a fetus with complete heart block is likely to be identified in utero because fetal bradycardia can be routinely recognized at the obstetrical visit, prompting further investigation for congenital heart disease. In our cohort, CHB was more common in the prenatally diagnosed group, although it did not reach statistical significance.
Risk Factors for Poor Outcome
CHB in association with heterotaxy syndrome was uniformly fatal and was therefore a significant risk factor for poor outcome, with a hazard ratio of 11.35 (Table 4). No patient survived past 3 months of age despite (in some cases) aggressive management with pacemaker placement and palliative surgical intervention for the heart defect. Others have reported similarly poor outcome for patients with heterotaxy syndrome in association with CHB [11, 21, 2831].
Isolated fetal CHB is also associated with high morbidity and mortality and can be associated with a dilated cardiomyopathy [20, 21, 32]. It is likely that the myopathic process that occurs with CHB has an even more profound impact on patients with complex congenital heart disease, making it a lethal combination. Of interest, those patients with polysplenia syndrome were less likely to survive than those with asplenia. This finding is related to the fact that CHB is exclusively seen in patients with polysplenia syndrome. Others have shown better outcome for patients with polysplenia; however, children who undergo two-ventricle repair were included in those analyses [27, 29].
Amongst those who survived to term, TAPVC was another risk factor for poor outcome (Table 4). In heterotaxy syndrome, TAPVC to an extracardiac site is frequently accompanied by pulmonary venous obstruction. Such patients often present with marked, life-threatening cyanosis in the newborn period. In addition, reintervention on the surgical site for recurrent pulmonary venous obstruction is common [33]. Extremely poor outcome has been reported for patients with heterotaxy syndrome in association with a single ventricle and TAPVC [13, 14, 34]. Gaynor and colleagues [34] reported a survival rate of 19% at 5 years.
In heterotaxy syndrome in particular, the mortality rate for repair of TAPVC has been reported as high as 95%, with obstructed pulmonary veins an independent risk factor for death [12]. A more recent assessment suggests that short-term outcome for TAPVC with single ventricle is improving, although survival remains only 50% [35]. In our present study cohort, 10 (45.4%) of 22 heterotaxy syndrome patients with TAPVC survived, 2 of whom have residual pulmonary venous obstruction.
Significant atrioventricular valve regurgitation was a significant risk factor for poor outcome in univariate analysis (Table 3). In our cohort, only 4 of 14 patients with significant atrioventricular valve regurgitation have survived without cardiac transplant. Significant atrioventricular valve regurgitation in the face of single ventricle physiology has been previously reported to be associated with poor outcome in patients undergoing single ventricle reconstructive surgery [36, 37].
Dextrocardia was assessed as a risk factor because we hypothesized that it would be associated with poor outcome. Surprisingly, we found that dextrocardia did not negatively impact survival (Tables 3 and 4). Although no association was found between dextrocardia and any other risk factor (ie, CHB, TAPVC), it may be a surrogate of some unknown protective mechanism.
Accuracy of Prenatal Diagnosis
In our assessment of accuracy of prenatal diagnosis of heterotaxy syndrome, we found that identification of systemic and pulmonary venous anomalies in utero has a very low false-positive rate and a high positive predictive value. However, the false-negative and negative predictive values were poor for anomalous systemic and pulmonary venous connections. Although recognition of systemic venous anomalies is important for surgical intervention in heterotaxy syndrome (particularly for bidirectional Glenn), prenatal diagnosis of systemic venous anomalies does not significantly affect outcome because these anomalies do not cause hemodynamic compromise after birth.
In contrast, prenatal identification of TAPVC allows the caretakers to evaluate and treat pulmonary venous obstruction early. Despite improvements in ultrasound technology, prenatal diagnosis of TAPVC remains challenging [3840]. Several reports have suggested that in isolation, TAPVC can be diagnosed with particular attention to the pulmonary venous flow pattern as well as visualization of a confluence and vertical vein behind the left atrium [41]. However, TAPVC in association with pulmonary outflow obstruction in patients with heterotaxy syndrome may be more difficult to interpret.
Our study showed that knowledge of a diagnosis of TAPVC before birth did not significantly improve survival, because those who were unrecognized before birth did not have worse outcome than those with prenatal identification of TAPVC. This is likely because all patients at our institution with asplenic type heterotaxy syndrome are assessed and treated aggressively knowing that TAPVC is a possibility. Despite early intervention, these patients generally do poorly because of recurrent pulmonary venous obstruction.
Fetal diagnosis of systemic and pulmonary outflow obstruction, as well as prediction of whether a lesion will be ductal dependent, was extremely accurate compared with postnatal diagnosis, with no errors in prenatal diagnosis (Table 2). These findings help dictate which palliative procedure will be required in the newborn period and help in the appropriate counseling of families with regard to delivery plan and outcome.
Significant atrioventricular valve regurgitation had a specificity and positive predictive value of 100%, but sensitivity and negative predictive values were poor (Table 2). Fetuses with significant atrioventricular valve regurgitation in utero are likely to have an anatomic defect of the atrioventricular valve. In contrast, atrioventricular valve regurgitation after birth may develop in those infants with a competent atrioventricular valve in utero as a result of hemodynamic compromise and changes in ventricular volume load when transitioning to the neonatal period. Certainly, identification of atrioventricular valve regurgitation prenatally, in the setting of heterotaxy syndrome, suggests that it will be present after birth as well.
Study Limitations
The retrospective study cohort design and small population may have made it difficult to determine some risk factors for poor outcome in patients with heterotaxy syndrome. Outcome for those heterotaxy syndrome patients who undergo biventricular repair was not addressed in this study, because the risk factors for these are considered much less than for those with single ventricle [29] and the intention of our study was to look at the impact of prenatal diagnosis in the highest risk patients. In a larger population, other known risk factors for poor outcome may have become significant, such as type of initial surgical intervention (ie, Norwood procedure versus aortopulmonary shunt).
In addition, assessment of morbidity such as neurologic injury, noncardiac surgical intervention, and reoperation rate were not addressed in this study. Complete identification of extra-cardiac anomalies was not available for patients who died in utero or neonatally, therefore this factor was not included in the analysis.
Although noncardiac surgery was performed on some patients for diagnoses such as malrotation and in one case, biliary atresia, no patient was known to have died from these interventions. Finally, selection bias may play a role in the findings of this study because our center is a referral for very complex fetal patients and so the most severe spectrum of disease may be seen.
Conclusions
Prenatal diagnosis of heterotaxy syndrome does not improve survival in patients undergoing single ventricle reconstruction. The most potent risk factors for poor outcome, CHB and TAPVC, are not positively impacted by identification in utero. Complete heart block in association with heterotaxy syndrome has the highest mortality in our series. The recognition that CHB with heterotaxy syndrome is a terminal disease may be helpful when counseling families who carry a fetus with this complex diagnosis.
Overall accuracy of prenatal diagnosis of heterotaxy syndrome is excellent. Particular attention is required when assessing systemic and pulmonary venous connections, because the sensitivity for these defects remains low. Because outcome is quite poor for heterotaxy syndrome and has not significantly improved within the past decade, cardiac transplantation at birth may be a viable option for select patients. Efforts to develop innovative treatment strategies and potential methods for fetal cardiac intervention in this high-risk group are also warranted.
| Notice From the American Board of Thoracic Surgery Regarding Trainees and Candidates for Certification Who Are Called to Military Service Related to the War on Terrorism |
|---|
|
|
|---|
Carolyn E. Reed, MD
Chair
The American Board of Thoracic Surgery
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
P. V. Anagnostopoulos, J. M. Pearl, C. Octave, M. Cohen, A. Gruessner, E. Wintering, and M. F. Teodori Improved current era outcomes in patients with heterotaxy syndromes Eur. J. Cardiothorac. Surg., May 1, 2009; 35(5): 871 - 878. [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 |