Effect of Prenatal Diagnosis on Outcome in Patients With Congenital Heart Disease

After completing this article, readers should be able to: Congenital heart disease (CHD) continues to be a major source of neonatal mortality. Since the development of echocardiographic screening for fetal cardiac anatomy in the 1980s, an increasing number of affected newborns are identified prior to delivery. (1) A recent study in a population that had risk factors for CHD found that the sensitivity of fetal echocardiography was 38%, with a false-positive rate of only 0.05%. (2) However, when performed by skilled operators viewing outflow tracts in addition to the four-chamber view, an accuracy as high as 96% has been reported. (3) These prenatal diagnoses are heavily skewed toward more severe CHD, with disorders of the atrioventricular junction (endocardial cushion defects), univentricular hearts, and hypoplastic left heart syndrome (HLHS) being the most common (Table 1).The ability to identify the most severe forms of CHD prior to delivery has changed the population of liveborn infants who have CHD. Prenatal diagnosis allows for extensive parental counseling and coordination of care. The fetal diagnosis of CHD prompts genetic testing to identify chromosomal abnormalities, which are found in 10% of cases, and a detailed anatomic evaluation to assess associated extracardiac anomalies, which coincide in 15% of cases. (1) In cases of poor prognosis, early prenatal diagnosis gives the parents the option of terminating the pregnancy. The rates of pregnancy termination following prenatal diagnosis of severe CHD vary widely, based on regional, socioeconomic, and religious differences, and may decrease the incidence of some disorders at birth. In mainland Great Britain, the incidence at birth of pulmonary atresia with intact ventricular septum (PAIVS) has decreased by 27% due to terminations following prenatal diagnosis. (5) Parents who choose to continue to carry fetuses that have isolated severe CHD tend to opt for more aggressive management. This includes delivery at high-level hospitals and expedient referral and transfer to a pediatric cardiology center. Thus, prenatal diagnosis allows optimization of preoperative care, providing hope that the perinatal management will improve both mortality and neurologic morbidity.Indications for fetal echocardiography include fetal, maternal, and familial factors. (6) Fetal risk factors for CHD include evidence of structural malformations on ultrasonography, aneuploidy, hydrops, and fetal arrhythmias. Along with anomalies such as omphalocele, duodenal atresia, VACTERL syndrome, and spina bifida, increased nuchal translucency has been associated with structural heart disease. (7)(8) Echocardiography is an important diagnostic tool in the evaluation of fetal hydrops due to the cardiac origin in 15% to 25% of cases. Structural heart disease is an important cause of fetal bradycardia, accounting for 50% of in utero complete heart blocks. In contrast, fetal tachycardias, particularly supraventricular tachycardia, rarely are associated with structural heart disease. Echocardiography also should be performed in multiple gestations when there is suspicion of twin-twin transfusion syndrome, not only due to the cardiomyopathy that can develop and progress to hydrops with fetal demise, but also due to the potential of progression to pulmonary infundibular stenosis and even pulmonary atresia in the recipient twin. (9)Maternal risk factors that should prompt fetal echocardiography include maternal disease, teratogen exposure, and exposure to certain medications. In pregestational maternal diabetes, the overall rate of congenital malformations may be as high as 6% to 10%, of which 40% to 50% are cardiac. If present during fetal organogenesis, maternal diabetes has been associated with transposition of the great arteries, ventricular septal defects, and abnormalities of cardiac looping, such as double-outlet right ventricle and heterotaxy syndromes. When poor glycemic control, as seen in gestational diabetes, occurs later in gestation, fetal and neonatal myocardial hypertrophy can develop. Maternal phenylketonuria also may have a teratogenic effect on the developing heart, resulting in a seven-fold increase in CHD, particularly coarctation of the aorta and HLHS. The maternal autoantibodies anti-Ro/SSA and anti-La/SSA seen with Sjögren syndrome and systemic lupus erythematosus are associated with fetal bradycardia and heart block. A history of teratogen exposure, such as to vitamin A analogs, antiseizure medications, alcohol, and lithium, should prompt a fetal cardiac evaluation due to associated heart defects (Table 2). In addition, fetal echocardiography may be indicated if there is maternal use of medications such as indomethacin or other prostaglandin synthetase inhibitors that could cause premature in utero ductal closure.Familial indications for fetal echocardiography include a family history of CHD and the presence of some mendelian syndromes. The fetus should be evaluated if there is a maternal or paternal history of CHD or a previous child who has CHD. Mendelian syndromes that may be associated with CHD include tuberous sclerosis, Marfan syndrome, Ellis-van Creveld syndrome, Noonan syndrome, and DiGeorge/velocardiofacial syndrome (Table 3).The prognosis for a fetus at the time of in utero diagnosis is substantially worse than a similar fetus diagnosed postnatally. The four-chamber view routinely used to assess the fetal heart has the greatest sensitivity for disorders of the atrioventricular junction and for critical left ventricular outflow tract obstruction, (4) conditions that are associated with a poor outcome. Additionally, reasons for referral and evaluation, including extracardiac anomalies, chromosomal abnormalities, fetal hydrops, and fetal arrhythmias, are associated with poor fetal outcome. Fetuses that have CHD, even in the setting of normal findings on chromosomal studies, have an increased risk of intrauterine death because the anatomy may become more unfavorable in utero with advancing gestation. (10) A recent study calculated the survival rate of fetuses diagnosed early in gestation with HLHS to be less than 40%. In contrast, fetuses that survived until delivery and were diagnosed postnatally had a survival of greater than 70%. (11)Several retrospective studies from surgical referral centers have compared newborns who did and did not have prenatal diagnoses of CHD. These studies, which encompassed a variety of heart defects, have failed to show a survival advantage with prenatal diagnosis. Copel and associates (12) performed a cohort analysis in patients who have prenatally versus postnally diagnosed CHD and failed to find cost savings, decreased length of hospitalization, or improved survival with the prenatal diagnosis. Verheijen and colleagues, examining a population of infants who had diverse heart defects, compared 81 patients who received prenatal diagnoses with 327 diagnosed postnatally. (13)(14) The prenatally diagnosed group was started on prostaglandin therapy significantly earlier, had less severe metabolic acidosis, and had lower lactate levels, especially among infants who were ductal-dependent. The authors hypothesized that the lower lactate levels were most likely due to earlier initiation of prostaglandins. However, there were no differences in length of hospital stay or overall mortality, even when evaluated by subgroups of ductal dependency, univentricular or biventricular repair, and presence and type of heart obstruction. Because the timing of postnatal presentation varies, the greatest effect of prenatal diagnosis on outcome might be seen in left ventricular outflow tract obstruction, including HLHS, in which patients may not be diagnosed in the immediate neonatal period but may present in extremis after ductal closure.The effect of prenatal diagnosis on outcome in left-sided obstructive lesions is similarly unclear. Eapen and associates (15) compared 23 cases of prenatally diagnosed left ventricular outflow tract obstruction with 45 cases that were diagnosed postnatally. Diagnoses included coarctation of the aorta, aortic stenosis, subaortic stenosis, or variants of HLHS, all dependent on ductal patency for systemic circulation. There was a significant delay in presentation of the postnatal group and subsequent greater acidosis, hemodynamic instability, noncardiac organ dysfunction, and delay in surgery, but there was no difference in surgical mortality. This result differs from a retrospective review of all cases of critical aortic coarctation that included coarctation diagnosed at autopsy in the postnatal group. (16) These investigators found that cardiovascular collapse and death were more common in the postnatal group, as was greater preoperative illness severity.HLHS continues be one of the most frequently prenatally diagnosed types of CHD and carries a poor prognosis, with a high rate of late neurologic sequelae. Theoretically, prenatal diagnosis followed by early initiation of prostaglandin therapy could prevent hypoxemia, hypoperfusion, and acidosis, potentially improving both neurologic morbidity and mortality. In a review of 217 cases of HLHS that were referred to a surgical center, 39 of which were diagnosed prenatally, there was a delay in presentation in the postnatal group, with 23% presenting after discharge. (17) The prenatal group experienced less acidosis and hypoxemia and fewer indications of kidney impairment preoperatively. However, there was no improvement in the postoperative course, with no difference in operative mortality, time ventilated, or length of hospital stay compared with infants in which HLHS was diagnosed postnatally. This result differs from findings of another study of 88 cases of HLHS, 33 of which were diagnosed prenatally. (18) In this study, prenatally diagnosed patients had an increased survival, a lower incidence of tricuspid regurgitation, less ventricular dysfunction, and lower requirements for alkali or inotropes.Daubeney and colleagues (5) studied the effect of prenatal diagnosis on the outcome of patients who had PAIVS in a population-based study from the United Kingdom. In Great Britain, more than 90% of pregnant woman undergo four-chamber fetal echocardiography at 16 to 20 weeks’ gestation. In this study, PAIVS was detected prenatally in 86 patients, and there were 183 live births of infants who had PAIVS. The 29 liveborn prenatally diagnosed patients’ hearts had similar morphologic features and spectra of severity as found in the 154 liveborns diagnosed with PAIVS postnatally. Although 4.5% of postnatal patients were discharged from the hospital prior to diagnosis, there was no difference in 1- and 2-year mortality between the two groups.D-transposition of the great arteries (D-TGA) usually presents at birth with severe hypoxia that requires prostaglandin infusion or atrial septostomy to prevent rapid hemodynamic collapse and death. Bonnett and associates (19) reviewed a series of patients who had prenatally versus postnatally diagnosed D-TGA, were referred for surgical repair, and had similar risk factors for operative mortality. The postnatally diagnosed group had an increased frequency of metabolic acidosis, multiorgan failure, or neurologic distress with increased risk of hemodynamic compromise at admission. This increased preoperative morbidity translated into increased preoperative mortality (6% versus 0%), increased postoperative mortality (20% versus 0%), and longer hospital stay in the postnatal diagnosis group. Another review of outcome in prenatally versus postnatally diagnosed D-TGA also showed improved preoperative condition in the prenatally diagnosed patients, but this did not translate into decreased mortality, duration of mechanical ventilation, or length of hospitalization. (17) The failure to show a difference in mortality may have been due to the extremely low overall mortality (0.5%) in this series. The only two deaths were in the postnatal group, and it is not known if rapid access to emergency septostomy, which was available to the prenatal diagnosis group, would have prevented these deaths.The impact of prenatal diagnosis on adverse neurologic events and survival in HLHS was examined in an investigation of 216 patients referred for palliative surgery, 79 of whom were diagnosed prenatally and all of whom were enrolled in a neuroprotection trial that provided resources for prolonged developmental follow-up. (20) The recorded adverse neurologic outcomes included clinical seizures and coma prior to discharge after stage I palliation. Seizures were selected as an endpoint because of an association with poor performance on cognitive tests in school-age children who had undergone CHD repair. The prenatal group had less preoperative acidosis, but there was no change in preoperative, operative, or overall mortality. However, there was a trend toward reduced preoperative and postoperative adverse neurologic events. Although the study was limited by small sample size, the results are consistent with the hypothesis that prevention of early acidosis and ischemia may decrease neurologic injury. The authors plan to follow this cohort long-term to evaluate the effects of prenatal diagnosis on neurodevelopment.Only one study has compared neurodevelopmental outcomes in prenatally and postnatally diagnosed CHD. Bartlett and colleagues (21) compared 25 patients prenatally diagnosed with D-TGA with 321 diagnosed postnatally. Prenatal diagnosis was associated with a lower rate of spontaneous labor, lower birthweight, lower 5-minute Apgar scores, and a higher intubation rate preoperatively. These differences, in part, were hypothesized to be related to the obstetrician’s desire to induce delivery in term infants based on availability of staff and intensive care unit beds. Additionally, knowledge of the baby’s cardiac condition and cyanosis likely influenced the 5-minute Apgar score for color and may have lowered the threshold for intubation. Despite the prenatally diagnosed group undergoing corrective surgery earlier than the postnatal group, there was no difference in Psychomotor Development Index and Mental Development Index scores at 1 year of age.Fetal cardiology is a rapidly evolving field in which sensitivity for and earlier detection of congenital heart disease is increasing. There are many advantages of prenatal diagnosis that are difficult to quantify in a study. Paramount is the parents’ ability to be informed of their unborn child’s medical condition and to gather additional information to make decisions for their child’s future. Critical to this decision making process is further detailed anatomic scans and genetic counseling because extracardiac anomalies and aneuploidy denote a worse prognosis. The ability to diagnose CHD prior to 20 weeks’ gestation allows parents the choice to terminate a pregnancy, and in some populations, this option has decreased the incidence of severe CHD at birth.If parents decide to continue with the pregnancy, there is a period of intense preparation, both to develop a plan of medical management for the child and to prepare the parents for the emotional, psychological, and potentially financial rigors of what is to come. Increasingly this occurs at fetal cardiac care centers, where a multidisciplinary team confirms the diagnosis; educates the family about the child’s lesion, management, and prognosis; and helps to lessen parental anxiety. (9) At our institution, prenatally diagnosed patients are delivered increasingly at the adjoining hospital that has obstetric services and a consistent neonatology team. This allows for improved communication between the fetal cardiologists and neonatologists and development of a delivery plan. An attending neonatologist is present at the delivery of a fetus in which CHD was diagnosed prenatally, prostaglandin E1 therapy is initiated if indicated, umbilical lines are placed, and the patient is transferred directly to the cardiac intensive care unit. In cases of fetal arrhythmias, a cardiologist is present if emergent treatment is anticipated. Some patients who have HLHS and an intact atrial septum have been delivered at The Children’s Hospital of Philadelphia and placed directly on bypass for an atrial septostomy due to their high risk of mortality without rapid intervention.The studies reviewed in this paper offer conflicting evidence of whether prenatal diagnosis improves mortality or long-term outcomes in CHD. There are many potential reasons for this finding, including that most studies, even at large referral centers, are underpowered in the prenatal diagnosis arm to measure differences in outcome. Additionally, the postnatal population in most of these studies does not reflect the true spectrum of postnatally diagnosed disease because not all patients may be referred for surgery or even diagnosed prior to their death. The great variability in the anatomy of diseases makes it difficult to ascribe causality of outcome to the timing of diagnosis, especially in small populations. Furthermore, certain variables that affect outcome, such as patient size and anatomic subtype, are static and are not changed by prenatal diagnosis.Most findings in the literature support the hypothesis that prenatal diagnosis prevents preoperative hemodynamic compromise and acidosis by hastening efforts to maintain ductal patency or facilitating timely emergent surgical interventions if necessary. Although there is hope that prenatal diagnosis will reduce neurologic impairment by decreasing the frequency of acidosis and hemodynamic instability, little evidence of benefit presently is available. Definitive studies to examine this theory are yet to be performed.A major advantage of prenatal diagnosis is being realized. Fetal echocardiography has allowed us to track the development to CHD in utero and devise strategies to prevent its progression. This concept is most promising in left ventricular outflow tract obstruction, where balloon valvuloplasty of fetal aortic stenosis may promote forward flow through the ventricle and prevent the development of HLHS. Although such fetal interventions have not yet been perfected, fetal echocardiography eventually should provide both prenatal diagnosis and curative therapies for CHD.