Prenatal echocardiography – the impact on neonatal management

Congenital heart disease (CHD) results in neonatal morbidity and mortality. Prenatal diagnosis allows preparing an appropriate perinatal and postnatal care. Babies born in low-risk level sites with unexpected CHD may have poorer outcomes. The purpose of this study was to compare results of foetal echocardiography to postnatal findings and assess the impact of antenatal suspicions of CHD on postnatal management. Medical records of mother-infant pairs with CHD admitted to the Neonatal Intensive Care Unit (NICU) of the Medical University of Gdansk from 01.01. to 31.12.2013 were reviewed. We analysed if the defect was detected preor postnatally, and if the diagnosis was made by the obstetrician from low-risk level sites (level I) or from a tertiary care centre (level II sonography). The overall incidence of CHD was 68 (3,4%). Critical congenital heart defects (CCHD) were found in 24 neonates (1,2%), 21 were diagnosed prenatally, 3 were transferred from 1st level units. Correlation between prenatal diagnosis made at our centre and postnatal findings was achieved in 47,7%. Accuracy in all prenatal and postnatal findings for both I and II sonography levels was 35,2%. There were major differences in the disproportion of the great vessels and postnatal confirmation of coarctation of the aorta (CoAo) (7,1%). We obtained a high accuracy of prenatal and postnatal findings in detection of lesions such as Tetralogy of Fallot (ToF), transposition of the great arteries (TGA), DORV (double outlet right ventricle) and Critical Pulmonary Stenosis, which require an outflow tract view (92,9% of cases). Conclusions: We confirmed increasing diagnostic rates when the diagnostics is performed at a tertiary care centre. These results are in agreement with literature stating that prenatal detection of CoAo is still challenging. Despite the high rates of misdiagnosis, majority of infants benefited from prenatally diagnosed CCHD.


INTRODUCTION
Congenital heart disease (CHD) results in significant neonatal morbidity and mortality.(1,2) Prenatal diagnosis of CHD is increasingly common; it varies from 16 to 65% depending on the experience of the centre and that of the physician.(3) Previous studies have shown big disparities in detection rate between university centres and peripheral practices.Many countries attempted to introduce guidelines for foetal screening cardiac examination.Using a four chamber view is advocated for obstetric scan.But some lesions are not evident from this scanning plan.Without using the extended scan like three vessel view, some of severe CHD, such as coarctation of the aorta (CoAo) remains undetected before birth.(4)(5)(6) Prenatal diagnosis allows to refer the parents to tertiary medical centres and prepare for planned delivery, as well as establish an appropriate perinatal and postnatal therapeutic plan.(7)(8)(9) Duct dependent heart disease used to be considered as critical CHD (CCHD) in paediatric cardiology.In case of defects diagnosed after birth, after ductus arteriosus (DA) is closed, the infant that did not receive prostaglandin (PGE) was in a life-threatening situation.(10) When the defect is detected prenatally and the infant is given PGE as planned, the prognosis for good outcomes of the treatment becomes remarkably better.(8,11,12) Prenatal assessment of severity of a cardiac disease seems to be the most important.On account of that, it can be planned whether the infant may require immediate intervention in the first hours after birth or can be stabilized with a PGE infusion and may not require a surgery in the first hours of life.(8,11,12) Researches have recently approached to creating a classification of the management of foetuses in the delivery room, based on the prenatal diagnosis and the expected diagnosis at birth.In Poland, Responek and Słodki defined four groups of CHD: the severest heart defects, severe urgent heart defects, severe planned heart defects and planned heart defects.(11) Berkley et al. proposed 5 care plans, depending on the expected severity of CHD, starting from comfort care in case of lethal defects up to a delivery at a tertiary centre with planned delayed surgery at a tertiary cardiac centre or maternal transport with delivery at a tertiary cardiac centre.In case when the prenatally expected severity of a defect and the suggested perinatal care plan are corresponding to the after-birth diagnosis, even if there are some differences in the anatomy of the defect, they can be classified as minor variations in prenatal and postnatal diagnosis.When the previously planned care program needs to be modified, it has to be classified as major variation.(12) Babies born in low-risk level sites with unexpected CHD may have significantly poorer outcomes due to lack of immediate access to appropriate speciality care.(13) Prenatal suspicion of CHD, even if it turns out false positive, always instigates parental anxiety and requires detailed paediatric cardiac evaluation.The purpose of this study was to compare results of foetal echocardiography (FE-CHO) to postnatal findings and to assess the impact of antenatal suspicions of CHD on postnatal management.

METHODS
Medical records of mother-infant pairs with CHD or other cardiac disorders admitted to the Neonatal Intensive Care Unit (NICU) of the Medical University of Gdansk from 01.01. to 31.12.2013 were reviewed.We analysed if the defect was detected preor postnatally, and if the diagnosis was made by the obstetrician from low-risk level sites (level I) or from a tertiary care centre (level II sonography).Foetuses referred to our tertiary perinatal care centre had been scanned by a perinatologist specialized in foetal medicine.The types of CHD recognized prenatally were divided into groups based on the echocardiographic views required to identify the disease in the foetus.Group 1 were those malformations identified by a fourchamber view alone.Group 2 were those identified by the addition of an outflow tracts view.Group 3 were those defects which needed other views (table 1).Disproportion of the great vessels was described separately.There were 28 prenatal suspicions of these anomalies.Additionally, a separate group of others findings, like: cardiomegaly -2, dextrocardia -1, normal echo -2, was isolated.Postnatal confirmation was obtained on the basis of neonatal echocardiography (NECHO), performed by paediatric cardiologist before discharging the infants home or transferring to surgery centre.Late outcomes were obtained by reviewing medical records of paediatrics cardiac centre.We compared the results of foetal and neonatal echocardiograms.Based on Berkley's suggestions, we classified differences as minor or major variations.A minor variation was the one that did not result in a treatment plan change, while a major variation did result in a treatment plan change.We looked at the accuracy of the prenatal and postnatal diagnosis in regards to guiding neonatal management after delivery.Arrhythmias without structural heart disease were excluded from the study.

STATISTICAL ANALYSIS
Accuracy of prenatal and postnatal diagnosis was evaluated by chi-square distri-bution and independence tests, using IBM SPSS 23.

RESULTS
The amount of cardiac disorders in neonates hospitalized in our facility (2008 infants) over the study period was 74 (3,7%).Of them 68 were CHD (3,4% of the whole group), in 6 newborns other cardiac disorders were detected (cardiac tumor in 1 infant, dextrocardia -1 infant, false chordae tendineae -1 and tricuspid insufficiency in 3 neonates).Severe and critical congenital heart defects were found in 24 neonates (1,2% of the series, 35,3% of all CHD), 21 were diagnosed prenatally at our centre, 3 were transferred from level I units.All infants with antenatal suspicion of cardiac lesions were admitted to the NICU and obtained cardiac assessment.There were 88 foetuses in which an obstetric scan suggested cardiac lesions.Of them 65 were verified at our centre.Remaining 23 were examined only at level I sonography.Neonatal findings were similar with prenatal diagnosis in 31 cases; all of them were recognized at level II facility.There were 29 of CHD recognized.Minor variations were detected in 6 cases, all described as accurate.Additionally, CHD were excluded in 2 cases referred from level I facilities for foetal ECHO; NECHO revealed normal heart anatomy and they were assessed as accurate.Correlation between prenatal diagnosis made at level II sonography and postnatal findings was achieved in 47,7% of cases.Accuracy in all prenatal and postnatal findings for both I and II sonography levels was achieved in 35,2% of cases (table 2).Accuracy of prenatal and postnatal diagnosis was 92,9% (13 cases from 14 foetuses) for heart defects requiring outflow view for detection, and 30,4% (7 cases from 23 assessed foetuses) for those presented in four chamber view, as well as requiring other views for recognition -33,3% (6 cases among 18 foetuses).Regarding other diagnosis represented by 5 cases, there were 3 corrected ones (60%) (tables 3.1-3.4).There were major differences in the disproportion of the great vessels and postnatal confirmation of CoAo; from 28 foetuses cardiac defects like Aortic Stenosis (AS) and CoAo were confirmed postnatally in only 2 cases (7,1%) (table 3.5).There were groups of 21 critical and serious CHD detected prenatally, defined on the basis of a baby's after-birth condition, survival and timing of a surgical inter-vention: the severest (CHD impossible to treat), severe urgent cardiac disease (CHD needing invasive procedures within the first hours after birth), planned cardiac disease (CHD needing invasive procedures within the first months after birth) and planned heart defects (which do not require cardiac surgery within the first months after birth) (table 4).In one case a lethal cardiac defect was detected prenatally; the neonate's status deteriorated rapidly after birth, the baby received comfort care, yet died 2 hours after delivery.Nineteen neonates were admitted to urgent or planned surgery.There was hypoplastic left heart syndrome (HLHS) with intact atrial septum (IAS) diagnosed prenatally in 1 case with a satisfying atrial flow after birth, thus previously planned care program needed to be modified and the classification of a defect changed from severe urgent to severe planned (major variation).However, the infant required prostaglandin and a planned surgical intervention.The infant with Ebstein's anomaly and pulmonary atresia suspected prenatally as criss-cross heart despite of having major variation between foetal and neonatal ECHO benefited from prenatal suspicion of CCHD, because the prostaglandin intake has started immediately after birth.The neonate with corrected transposition of the great arteries (cTGA) plus ventricular inversion and atrio-ventricular block Wenckebach's type required antiarrhythmic treatment but did not require a surgery.In cases of CCHD, prenatal diagnosis corresponded with postnatal findings in 19 of 21 foetuses diagnosed in our centre (accuracy 90,5%) (table 4).In the following prenatal examinations, 14 infants (13 of them examined at our centre) were diagnosed with cardiac disease with no instability expected including: mild pulmonary stenosis (PS) (1 baby); dextrocardia and Ventricular Septal Defect (VSD) (1 baby); dextrocardia (1 baby); VSD (3 babies); secundum atrial septal defect (ASD2) (1 baby); cardiac tumor without outflow obstruction (1 baby); tricuspid valve regurgitation (TVR) without other cardiac malformation (3 babies), dilated coronary sinus (1 baby) or bicuspid aortic valve (BAV) (1 baby).In one neonate referred from level I facility for delivery at our centre clinically nonsignificant CHD (ASD2) occurred.In 39 neonates cardiac defects were diagnosed after birth in neonatal echocardiography done for other reasons.Defects diagnosed postnatally are shown in table 5. Three neonates without antenatal diagnosis of CCHD were referred to our centre

Table 1 .
Division of congenital heart disease (CHD) into groups according to echocardiographic view required for prenatal diagnosis.

Table 2 .
Accuracy of foetal and neonatal echocardiography findings

Table 5 .
(3,20),33)fects identified postnatally.Infants with congenital heart disease without prenatal diagnosis, born in low-level facilities, are at risk of not receiving a needed care plan and, in the end, of worse treatment outcomes.(12,13,30)Delayeddiagnosis in patients who survive is the risk of hypoxic/ ischaemic brain injury.Periventricular leukomalacia has been reported on MRI imaging of the brain in up to 39% of neonates with critical CHD.(33)Prenatal diagnosis is associated with lower rate of acidosis, less frequent neurological complications and planned referral for surgery prior to development of life-threatening consequences of the defect, while a delay in diagnosis may lead to hypoxia, shock, multiorgan failure and worse outcomes of both early and late surgical actions.(13,30,33)Thesecanbeconfirmedby our observations.Infant with a CoAo and HAA diagnosed after birth was transferred to our unit in shock, with acidosis, needed to be intubated and required mechanical ventilation.Late outcomes of the treatment are not satisfying as it can result in cerebral palsy and periventricular leukomalacia.Notwithstanding the higher prenatal detection rate of CHD done at a tertiary setting, foetal cardiac assessment is still performed at level I sonography, as pregnant women are not referred for foetal echo from the primary services.(3,20)Falsepositivediagnosis of foetal cardiac abnormality results in parents' anxiety, in contrast to misdiagnosis of CCHD in foetus, thereby delaying the diagnosis and, consequently, adequate specialized prenatal follow-up and delivery management at a specialized centre.It should be emphasized that infants born at our centre had prenatal diagnosis of CCHD, in contrast to those born at other sites, which affected postnatal management and late outcomes.Neonates with CCHD born at non-tertiary centres weretransferred to our unit, which implies that availability of prenatal diagno-sis beyond university clinics is still insufficient.Newborns with late diagnosis of a CCHD were referred to cardiosurgical treatment in poor general and neurological status, which is another argument for identification of an accurate method for early detection of critical heart defect.ConclusionsThe prenatal diagnosis has significant impact on neonatal management, allowing appropriate medical care and planning surgery immediately after birth.We confirmed increasing diagnostic rates when the diagnostics is performed at a tertiary care centre.The results of our study are in agreement with literature which states that prenatal detection of CoAo is still challenging.Despite the high rates of misdiagnosis, majority of infants benefited from prenatally diagnosed CCHD.Improved accuracy in foetal diagnosis can be achieved through better organisation of perinatal care.