2020 年 84 巻 12 号 p. 2275-2285
Background: Congenital heart disease (CHD) is often diagnosed prenatally using fetal echocardiography, but few studies have evaluated the accuracy of these fetal cardiac diagnoses in detail. We investigated the discrepancy between pre- and postnatal diagnoses of CHD and the impact of discrepant diagnoses.
Methods and Results: This retrospective study at a tertiary institution included data from the medical records of 207 neonates with prenatally diagnosed CHD admitted to the cardiac neonatal intensive care unit between January 2011 and December 2016. Pre- and postnatal diagnoses of CHD differed in 12% of neonates. Coarctation of the aorta and ventricular septal defects were the most frequent causes of discrepant diagnosis. Unexpected treatments were added to 38% of discrepant diagnostic cases. However, discrepant diagnoses did not adversely affect the clinical course. The 9% of the 207 neonates who required invasive intervention within 24 h of delivery were accurately diagnosed prenatally.
Conclusions: Pre- and postnatal diagnoses differed in only a few neonates, with differences not adversely affecting the clinical course. Neonates who required invasive intervention immediately after delivery were accurately diagnosed prenatally. Prenatal diagnosis thus seems to contribute to improved prognosis in neonates with CHD.
Congenital heart disease (CHD) is often detected prenatally by fetal echocardiography, and most CHD can be diagnosed within ~20 weeks of gestation. The Japanese Society of Fetal Cardiology and the Japanese Society of Pediatric Cardiology published guidelines for fetal echocardiography in 2006,1,2 and during 2014 the American Heart Association published a scientific statement about the diagnosis and treatment of fetal cardiac disease.3 The rate of prenatal CHD diagnosis increased rapidly after public medical insurance in Japan started covering the prenatal diagnosis of CHD determined using fetal echocardiography during 2010.
However, the benefit of prenatal diagnosis of CHD in terms of patient outcomes remains controversial.4–9 It has improved the clinical course of neonates with CHD,6,7,9,10 but whether prenatal diagnosis improves mortality rates among fetuses with CHD remains unclear. Undiagnosed neonates who die before transportation to tertiary medical care institutions have not been assessed. Prenatal diagnosis might improve mortality among neonates who need treatment immediately after birth.
In addition, few reports have described differences in diagnostic details between pre- and postnatal diagnoses of CHD and their impact.10–14
Our hospital is a tertiary institution specializing in cardiovascular disease and started expert fetal echocardiography as a routine examination from 2011. The present retrospective study aimed to investigate the characteristics of neonates with prenatally diagnosed CHD who were admitted to the cardiac neonatal intensive care unit (NICU), the accuracy of prenatal diagnoses, and the effect of discrepant pre- and postnatal CHD diagnoses on the clinical course of neonates.
We also retrospectively assessed the characteristics of neonates with prenatal diagnoses of CHD who required invasive intervention within 24 h of delivery to determine the effects of discrepant diagnoses and whether prenatal CHD diagnosis contributed to the prognosis.
The Ethics Committee at the National Cerebral and Cardiovascular Center approved the protocol of this retrospective study at a Japanese tertiary institution (M29-068).
We collected the following information from the medical records of neonates born with prenatally diagnosed CHD and who were admitted to the cardiac NICU between January 2011 and December 2016: sex, birth weight, gestational age (GA) at birth, mode of delivery, Apgar scores at 1 and 5 min, maternal age at birth, assisted reproductive technology (ART), GA at detection of cardiac abnormality, GA at first and last expert fetal echocardiography, frequency of fetal echocardiography, fetal diagnosis of heart disease, explanation of findings to parents, definitive postnatal diagnosis of heart disease, dominant cardiac defect (complex cardiac defects were classified according to the dominant defect; details are given later), functional single ventricle, heterotaxy, chromosomal anomaly, discrepancy between pre- and postnatal diagnoses, changes in treatment strategies between before and after birth, medical therapy and catheter intervention during neonatal period, neonatal surgery, duration of hospitalization, neonatal death and hospital death.
We compared the prenatal parameters and the dominant cardiac defects between patients who were assigned to groups according to the presence or absence of a discrepancy between pre- and postnatal diagnoses.
We investigated the characteristics of neonates with discrepant pre- and postnatal diagnoses and assessed the effect of discrepant diagnoses on treatment plans, duration of hospitalization and mortality.
We also assessed patients who had undergone invasive procedures such as catheter intervention or cardiac surgery within 24 h of delivery and compared the influence of pre- and postnatal diagnostic discrepancy.
Expert Fetal EchocardiographyPatients referred under suspicion of fetal cardiac disease underwent expert fetal echocardiography provided by pediatric cardiologists specializing in fetal and neonatal heart disease.
Ultrasound examination was performed under a transabdominal approach for all fetuses. A diagnosis of CHD was made according to guidelines,1,2 including color and pulsed-wave Doppler using a segmental approach. M-mode was used if the cardiac rhythm was abnormal. Examinations were performed using an Aplio XV, 500 (Toshiba Medical Systems, Tokyo, Japan), Prosound F75 (ALOKA, Tokyo, Japan), Voluson E8 (GE Healthcare, Chicago, IL, USA) or iE33 (Philips, Amsterdam, The Netherlands). Table 1 shows the morphological checklist of expert fetal echocardiography for CHD.
Variables | Findings | ||||
---|---|---|---|---|---|
Stomach | Left | Right | Ambiguous | ||
Liver | Right | Left | Bilateral | ||
Spleen | Left | Right | Ambiguous | ||
SVC | Right | Left | Bilateral | PLSVC to CS | PRSVC to CS |
Innominate vein | Left | Right | Retro aortic | None | |
IVC | Right | Left | Interruption with azygos or hemiazygos connection |
||
Viscero-atrial situs | Solitus | Inversus | RI (asplenia) |
Left isomerism (polysplenia) |
Ambiguous |
Pericardial effusion | Normal | Mild | Moderate | Large | |
Cardiomegaly | None | Mild | Moderate | Severe | |
Cardiac position | Levoposition | Mesoposition | Dextroposition | ||
Apex axis direction | Levocardia | Mesocardia | Dextrocardia | ||
Atrium | Solitus | Inversus | Right common | Left common | |
Pulmonary venous connection | Normal (to LA) |
TAPVR (draining vein and stenotic site) |
To bilateral LA (left isomerism) |
Common PV atresia |
PAPVR (right, left, upper, middle, lower) |
Atrial septum | Patent FO | Restrictive FO | Intact | ASD (primum, secundum, sinus venosus) | |
Atrioventricular connection | Concordance | Discordance | Tricuspid or mitral atresia |
Double-inlet single ventricle |
Common-inlet single ventricle |
Atrioventricular septum | Normal | Partial AVSD | Intermediate AVSD | Complete AVSD (Rastelli type A, B, C) | |
Ventricular loop | D-loop | L-loop | |||
Functional ventricle | Balanced 2 ventricles |
Left dominant 2 ventricle |
Right dominant 2 ventricle |
Left single | Right single |
VSD | Perimembranous | Outlet muscular | Inlet muscular | Travecular muscular | (small, moderate, large) |
Location of VSD with DORV | Subaortic | Subpulmonary | Doubly committed | Remote | |
Ventriculoarterial connection | Concordance | Discordance (TGA) |
Double-outlet RV | Single great arteries (e.g., Truncus) | |
Position of great arteries | Normal | d-MGA | l-MGA | ||
Aortic valve | Normal | Atresia | Stenosis (valvular, SAS, SVAS) (slight, moderate, severe) |
Over ride | |
Pulmonary valve | Normal | Atresia | Stenosis (valvular, infundibular, supra valvular) (slight, moderate, severe) |
||
Pulmonary artery | Normal | Hypoplastic | Main PA portion (present, absent, hypoplastic) |
Nonconfluent pulmonary artery |
|
Aortic arch | Left | Right | Double (right dominant, left dominant) |
Hypoplastic aortic arch |
Reverse flow |
Ductus arteriosus | Patent | Constriction | Reverse flow | Other findings (e.g., left DA from BCA with RAA) |
|
Aortic isthmus | Normal | Coarctation of aorta | IAA (type A, B, C) | ||
Abdominal aorta | Left | Right | |||
Valvular regurgitation | Trivial | Slight | Moderate | Severe | (aortic, mitral, pulmonary, tricuspid) |
Arrhythmia | Premature contraction (atrial, ventricular) |
Supraventricular tachycardia |
Ventricular tachycardia |
Sinus bradycardia | Atrioventricular block (I, II, III) |
ASD, atrial septal defect; BCA, brachiocephalic artery; CHD, congenital heart disease; CS, coronary sinus; DA, ductus arteriosus; DORV, double-outlet right ventricle; FO, foramen ovale; IAA, interrupted aortic arch; IVC, inferior vena cava; LA, left atrium; MGA, malposition of the great arteries; PAPVR, partial anomalous pulmonary venous return; PLSVC, persistent left superior vena cava; PRSVC, persistent right superior vena cava; PV, pulmonary vein; RAA, right aortic arch; RI, right isomerism; SAS, subaortic stenosis; SVAS, supravalvular aortic stenosis; SVC, superior vena cava; TAPVR, total anomalous pulmonary venous return; TGA, transposition of the great arteries; Truncus, truncus arteriosus.
If CHD was diagnosed with total anomalous pulmonary venous return (TAPVR) from fetal echocardiography, we attempted to identify the draining vein stenosis and stenotic sites. We empirically judged draining vein stenosis by taking the following findings into consideration: morphologically localized stenosis, diameter <2.0 mm in pregnancy >30 weeks, flow disturbance and continuous velocity >1.0 m/s.
We similarly attempted to confirm the presence or absence of a restrictive foramen ovale in hypoplastic left heart syndrome (HLHS), TGA and other morbidities requiring interatrial communication after birth. We judged restrictive foramen ovale by taking the following findings into consideration: continuous forward flow with increased a-wave reversal or to-and-fro PV flow pattern in HLHS,15 fixed septum primum with little mobility, redundant septum primum or hypermobile septum in TGA.16,17
With regard to the diagnosis of coarctation of the aorta (CoA), we defined it as requiring surgery in the neonatal period. So-called “mild CoA” judged as not requiring neonatal surgery was not included. The diagnosis of CoA was subjectively made by putting the following findings together: disproportion of ventricles and great vessels, localized stenosis, posterior shelf at the isthmus, and flow disturbance.18,19
After fetal echocardiographic assessment, a pediatric cardiologist described the diagnosis, postnatal symptoms, probable course, treatment strategies and prognosis to the parents. It was also explained that predicting the postnatal outcomes of patent ductus arteriosus, CoA and atrial septal defect (ASD) in fetuses can be difficult.
DeliveryAfter the parents provided written informed consent, neonates were delivered and transferred to the cardiac NICU. The institution’s basic policy is to encourage spontaneous vaginal delivery, even for fetuses with CHD.20 Pediatric cardiologists specializing in fetal and neonatal heart disease were present at all deliveries, and examinations with necessary treatments were immediately started postpartum.
Postnatal Diagnosis of CHDPostnatal definitive cardiac diagnoses were based on postnatal echocardiography, computed tomography, magnetic resonance imaging, cardiac catheterization and intraoperative findings.
Complex cardiac abnormalities were classified according to the dominant heart lesion.21 For example, CoA with ventricular septal defect (VSD) was classified as CoA. Atrioventricular septal defect (AVSD), when coexisting with double-outlet right ventricle (DORV), was classified as AVSD. HLHS was defined as a heart with atrioventricular and ventriculoarterial concordance, small left ventricle, and reversed flow in the aortic arch.
Functional single ventricle was defined as connection of both atria to the same ventricle or as connection of both atria to separate ventricles, one of which was hypoplastic.22
Definition of Discrepancy Between Pre- and Postnatal DiagnosesWe defined a discrepancy between pre- and postnatal diagnoses as postnatal anatomy and diagnosis differing from the prenatal findings. We judged a discrepancy as present if there was a different morphological diagnosis without a difference between pre- and postnatal treatment plans, such as small VSD. Minor differences in the size of the septal defect or grade of regurgitation at the cardiac valve between pre- and postnatal diagnoses were accepted as not discrepant.
In terms of the diagnosis of CoA, we judged discrepancy between the pre- and postnatal diagnoses in the early neonatal period, because the aortic isthmus sometimes develops during the neonatal period.
Statistical AnalysisDescriptive data are expressed as numbers (n) and percentages (%). Associations between categorical variables were assessed using the chi-square and Fisher’s exact tests. Non-parametric continuous variables are presented as median with range, and associations were evaluated using the Mann-Whitney U-test. Values with P<0.05 were considered significantly different.
Among the neonates admitted to the cardiac NICU between January 2011 and December 2016, we investigated data from 207 neonates who had been diagnosed with CHD by expert fetal echocardiography between December 2010 and November 2016. During this period, CHD was diagnosed in 271 of the 394 fetuses assessed for cardiac disease. Among these, 222 were delivered at our hospital between January 2011 and December 2016 and of them, 15 were not admitted to the cardiac NICU and cardiac defects were not examined in detail, mainly because the parents selected only compassionate care. We therefore assessed data from 207 neonates (Figure).
Flow chart of patients. CHD, congenital heart disease; NICU, neonatal intensive care unit.
Table 2 shows the profiles of the patients: 20 neonates (10%) were delivered before 37 weeks of GA; birth weight was <2,500 g in 54 neonates (26%), among whom 7 (3%) weighed <1,500 g. Neonatal asphyxia was defined in 27 neonates (13%) and 13 neonates (6%) who had Apgar scores <7 at 1 and 5 min after delivery, respectively. ART comprised in vitro fertilization (IVF: n=15; 7%) and intracytoplasmic sperm injection (ICSI: n=5; 2%). Artificial insemination with the husband’s semen was used in the 4 other cases (2%).
All cases | Discrepancy between pre- and postnatal diagnosis | |||
---|---|---|---|---|
(−) | (+) | P value | ||
No. of patients | 207 | 183 | 24 | |
Male | 99 (47.8) | 89 (48.6) | 10 (41.7) | NS |
GA at birth (weeks) | 38 (28–41) | 39 (28–40) | 38 (34–41) | NS |
Birth weight (g) | 2,782 (980–3,936) | 2,831 (980–3,936) | 2,624 (1,316–3,722) | NS |
Cesarean delivery | 62 (30.0) | 55 (30.1) | 7 (29.2) | NS |
Apgar score (1 min) | 8 (0–9) | 8 (0–9) | 8 (2–8) | NS |
Apgar score (5 min) | 8 (1–10) | 8 (1–10) | 8.5 (6–9) | NS |
Maternal age (years) | 32 (17–43) | 32 (17–43) | 32 (22–39) | NS |
Pregnancy by ART (IVF or ICSI) | 20 (9.7) | 18 (9.8) | 2 (8.3) | NS |
GA at detection of fetal cardiac abnormality (w) | 27 (13–41) | 27 (13–41) | 27.5 (18–36) | NS |
GA at first expert fetal echocardiography (w) | 31 (19–41) | 31 (19–41) | 32 (23–37) | NS |
GA at last expert fetal echocardiography (w) | 33 (20–41) | 33 (20–41) | 33 (26–37) | NS |
≥2 assessments by expert fetal echocardiographers | 46 (22.2) | 40 (21.9) | 6 (25.0) | NS |
Data are shown as median (range) or as n (%) for categorical variables. ART, assisted reproductive technology; GA, gestational age; ICSI, intracytoplasmic sperm injection; IVF, in vitro fertilization; NS, not significant.
Cardiac abnormalities were found at <22 weeks of GA in 28 fetuses (14%), among whom 7 (3%) were <22 weeks old at the time of initial fetal echocardiography. Induced abortion is legal in Japan when the GA of the fetus is <22 weeks. The number of echocardiographic assessments was 1 for 161 fetuses (78%), 2 for 42 fetuses (20%), 3 for 3 fetuses (1%) and 4 for 1 fetus (0%).
We identified discrepancies between pre- and postnatal diagnoses in 24 neonates (12%). However, these discrepancies were not associated with sex, GA at birth, birth weight, cesarean delivery, Apgar scores at 1 and 5 min, maternal age at delivery, ART, GA at detection of cardiac abnormalities, GA at initial and final fetal echocardiographic assessments, or the proportions of mothers who underwent ≥2 fetal echocardiographic assessments.
Table 3 shows the distribution of definitive postnatal diagnoses. Complex cardiac abnormalities were classified according to the dominant heart lesion. AVSD was the most prevalent, followed by CoA or interrupted aortic arch (IAA), HLHS and others.
All cases | Discrepancy between pre- and postnatal diagnosis | |||
---|---|---|---|---|
(−) | (+) | P value | ||
No. of patients | 207 (%) | 183 (%) | 24 (%) | |
Dominant cardiac defect | ||||
AVSD | 39 (18.8) | 35 (19.1) | 4 (16.7) | NS |
CoA or IAA | 23 (11.1) | 17 (9.3) | 6 (25.0) | 0.0335 |
HLHS | 21 (10.1) | 21 (11.5) | 0 (0.0) | NS |
DORV | 20 (9.7) | 15 (8.2) | 5 (20.8) | NS |
TOF | 17 (8.2) | 15 (8.2) | 2 (8.3) | NS |
TGA | 13 (6.3) | 12 (6.6) | 1 (4.2) | NS |
PAVSD | 8 (3.9) | 6 (3.3) | 2 (8.3) | NS |
TA or TS | 8 (3.9) | 8 (4.4) | 0 (0.0) | NS |
PAIVS | 7 (3.4) | 7 (3.8) | 0 (0.0) | NS |
DILV | 7 (3.4) | 6 (3.3) | 1 (4.2) | NS |
Truncus | 6 (2.9) | 5 (2.7) | 1 (4.2) | NS |
Ebstein’s anomaly | 5 (2.4) | 5 (2.7) | 0 (0.0) | NS |
PS | 5 (2.4) | 5 (2.7) | 0 (0.0) | NS |
cTGA | 4 (1.9) | 3 (1.6) | 1 (4.2) | NS |
TAPVR | 3 (1.4) | 3 (1.6) | 0 (0.0) | NS |
VSD | 3 (1.4) | 2 (1.1) | 1 (4.2) | NS |
AS | 2 (1.0) | 2 (1.1) | 0 (0.0) | NS |
Vascular ring | 2 (1.0) | 2 (1.1) | 0 (0.0) | NS |
Other | 14 (6.8) | 14 (7.7) | 0 (0.0) | NS |
fSV | 78 (37.7) | 71 (38.8) | 7 (29.2) | NS |
fRSV | 53 (25.6) | 47 (25.7) | 6 (25.0) | NS |
fLSV | 25 (12.1) | 24 (13.1) | 1 (4.1) | NS |
Heterotaxy | 31 (15.0) | 26 (14.2) | 5 (20.8) | NS |
RI | 19 (9.2) | 17 (9.3) | 2 (8.3) | NS |
Left isomerism | 12 (5.8) | 9 (4.9) | 3 (12.5) | NS |
Chromosomal anomaly | 33 (15.9) | 30 (16.4) | 3 (12.5) | NS |
21 trisomy | 12 (5.8) | 11 (6.0) | 1 (4.2) | NS |
22q11.2 deletion | 8 (3.9) | 8 (4.4) | 0 (0.0) | NS |
13 trisomy | 3 (1.4) | 3 (1.6) | 0 (0.0) | NS |
Others | 10 (4.8) | 8 (4.4) | 2 (8.3) | NS |
Data are presented as n and (%). AS, aortic stenosis; AVSD, atrioventricular septal defect; CoA, coarctation of the aorta; cTGA, corrected TGA; DILV, double-inlet left ventricle; fLSV, functional left single ventricle; fRSV, functional right single ventricle; fSV, functional single ventricle; HLHS, hypoplastic left heart syndrome; PAIVS, pulmonary atresia and intact ventricular septum; PAVSD, pulmonary atresia and ventricular septal defect; PS, pulmonary stenosis; TA, tricuspid atresia; TOF, tetralogy of Fallot; TS, tricuspid stenosis; VSD, ventricular septal defect. Other abbreviations as in Tables 1,2.
Frequencies of the dominant diagnoses of CoA or IAA were significantly higher among neonates with discrepancies between pre- and postnatal diagnoses of CHD, compared with those without discrepancies. The frequencies of other dominant diagnosis, functional single ventricle, heterotaxy and chromosomal anomalies did not differ significantly between the 2 groups.
Table 4 gives outlines of the treatment and outcomes of neonates with prenatally diagnosed CHD. About 50% of neonates were continuously infused with prostaglandin E1 (PGE1) to maintain patency of the ductus arteriosus, and about 25% of them required mechanical ventilation before surgery. Hypoxic gas was delivered mainly via nasal cannula when heart failure was caused by pulmonary overcirculation in neonates with spontaneous respiration.
All neonates | Discrepancy between pre- and postnatal diagnosis |
||
---|---|---|---|
(−) | (+) | ||
No. of patients | 207 | 183 | 24 |
Medical therapy for CHD | |||
PGE1 infusion to maintain patent ductus arteriosus | 105 (50.7) | 93 (50.8) | 12 (50.0) |
Mechanical ventilation before surgical intervention | 49 (23.7) | 45 (24.6) | 4 (16.7) |
Hypoxic gas inhalation for pulmonary overcirculation | 35 (16.9) | 30 (16.4) | 5 (20.8) |
Neonates requiring catheter intervention | 33 (15.9) | 32 (17.5) | 1 (4.2) |
Age at first neonatal catheter intervention (days) | 3 (0–27) | 3 (0–27) | 11 (11–11) |
BAS | 18 (8.7) | 17 (9.3) | 1 (4.2) |
Stent placement in ductus arteriosus | 8 (3.9) | 8 (4.4) | 0 (0.0) |
Stent placement in DV of TAPVR in RI | 7 (3.3) | 7 (3.8) | 0 (0.0) |
BAV | 2 (1.0) | 2 (1.1) | 0 (0.0) |
Balloon pulmonary valvuloplasty | 2 (1.0) | 2 (1.1) | 0 (0.0) |
Neonates undergoing surgical intervention | 116 (56.0) | 102 (55.7) | 14 (58.3) |
Age at first neonatal surgery (days) | 8 (0–28) | 7 (0–28) | 9 (0–27) |
Bilateral PAB | 31 (15.0) | 30 (16.4) | 1 (4.2) |
BT | 17 (8.2) | 16 (8.7) | 1 (4.2) |
EDA | 17 (8.2) | 12 (6.6) | 5 (20.8) |
PAB | 13 (6.3) | 10 (5.5) | 3 (1.3) |
ASO | 12 (5.8) | 10 (5.5) | 2 (8.3) |
Norwood procedure | 6 (2.9) | 6 (3.3) | 0 (0.0) |
PMI | 4 (1.9) | 3 (1.6) | 1 (4.2) |
RV-PA shunt | 4 (1.9) | 3 (1.6) | 1 (4.2) |
Repair of TAPVR | 4 (1.9) | 4 (2.2) | 0 (0.0) |
Other | 14 (6.7) | 13 (7.1) | 1 (4.2) |
Length of hospitalization (days) | 49 (0–254) | 49 (0–254) | 49 (0–119) |
Neonatal death | 10 (4.8) | 9 (4.9) | 1 (4.2) |
Age at neonatal death (days) | 6 (0–28) | 7 (0–28) | 5 (5–5) |
Hospital death | 21 (10.1) | 20 (10.9) | 1 (4.2) |
Age at hospital death (days) | 34 (0–165) | 34 (0–165) | 5 (5–5) |
Data are shown as n (%) or median (range). ASO, arterial switch operation; BAS, balloon atrial septostomy; BAV, balloon aortic valvuloplasty; BT, Blalock-Taussig shunt; DV, drainage vein; EDA, extended direct anastomosis of aortic arch; PAB, pulmonary artery banding; PGE1, prostaglandin E1; PMI, pacemaker implantation; RV-PA, right ventricle-pulmonary artery. Other abbreviations as in Table 1.
The most frequent catheter intervention during the neonatal period was balloon atrial septostomy (BAS), followed by stent deployment to the ductus arteriosus and to stenotic lesions of the vertical vein in TAPVR, balloon aortic valvuloplasty and pulmonary valvuloplasty.
Over 50% of the neonates required surgical intervention. Bilateral pulmonary artery banding (PAB) was the most frequent type, followed by Blalock-Taussig shunt and aortic arch reconstruction using extended direct anastomosis (EDA). Among 37 patients who underwent neonatal surgery associated with CoA or IAA such as EDA, other type of arch reconstruction, and Norwood-like procedure for HLHS with CoA, 32 had a prenatal diagnosis of CoA or IAA including HLHS with CoA, the sensitivity of which was 86% for both. Among 170 patients who did not undergo surgery associated with CoA or IAA as newborns, 3 were prenatally diagnosed with CoA or IAA with 98% specificity.
Table 5 shows details of 24 neonates with pre- and postnatal diagnostic discrepancies in CHD diagnoses. A diagnosis of CoA was overlooked in 5 neonates, who underwent EDA that had not been planned at the time of fetal diagnosis. Another case of CoA was prenatally diagnosed with IAA. All 29 fetuses diagnosed with CHD complicated by CoA not IAA were diagnosed with CoA during the early neonatal period, and 3 did not require neonatal aortic arch reconstruction, because the aortic isthmus developed during the neonatal period. We judged that the pre- and postnatal diagnoses were not discrepant in these 3 patients.
No. | Sex | GA at birth (weeks) |
Birth weight (g) |
Postnatal major cardiac diagnosis |
Systemic complication |
GA at last expert fetal echocardiography (weeks) |
Discrepancy at fetal echocardiography |
Unexpected additional treatment in neonatal period |
Neonatal intervention (age in days) |
Outcome at discharge (age in days) |
---|---|---|---|---|---|---|---|---|---|---|
1 | F | 40 | 2,474 | CoA; AVSD; PLSVC |
Left isomerism | 37 | CoA missed | EDA | EDA PAB (6) |
Alive (54) |
2 | F | 38 | 2,676 | CoA; HLHC; PLSVC |
32 | CoA missed | EDA | EDA (20) | Alive (43) | |
3 | M | 37 | 1,802 | CoA; DORV; subpulmonary VSD |
IUGR | 26 | CoA missed | EDA | EDA IVR ASO (9) |
Alive (113) |
4 | M | 38 | 2,564 | CoA; MS; VSD; PLSVC |
32 | CoA VSD mistaken for HLHC |
EDA VSD closure |
EDA VSD closure (12) |
Alive (58) | |
5 | M | 40 | 3,272 | CoA; VSD | 31 | CoA VSD mistaken for HLHC |
EDA PAB | EDA PAB (2) |
Alive (53) | |
6 | M | 39 | 2,676 | CoA; MS; AS; hypoplastic LV |
35 | CoA mistaken for IAA |
bltPAB (9) | Alive (45) | ||
7 | F | 38 | 2,626 | CoA; VSD | Multiple anomalies |
37 | VSD missed | Transferred due to rectal atresia (0) |
||
8 | F | 38 | 2,622 | CoA; VSD | Left isomerism |
34 | VSD missed | VSD closure | EDA VSD closure (7) |
Alive (35) |
9 | F | 41 | 2,774 | TGA; VSD | 27 | VSD missed | ASO (7) | Alive (81) | ||
10 | F | 37 | 1,726 | VSD | 13 partial trisomy |
35 | VSD missed | Alive (27) | ||
11 | F | 38 | 2,606 | DORV; multiple VSD |
34 | VSD location mistaken |
PAB instead of VSD closure |
BAS (11); PAB (15) |
Alive (50) | |
12 | F | 38 | 2,456 | DORV; subaortic VSD; PS; RAA |
SI | 33 | VSD location mistaken |
Alive (38) | ||
13 | M | 40 | 3,218 | PAVSD; MAPCA; PLSVC |
36 | PA mistaken for PS |
Alive (45) | |||
14 | F | 38 | 3,230 | AVSD; DORV; PA; TAPVR |
Right isomerism |
35 | PA RtPDA mistaken for Truncus |
PGE1 infusion |
Alive (121) | |
15 | M | 34 | 2,368 | AVSD; DORV; PA; bltPDA; PLSVC; RAA |
30 | PA bltPDA mistaken for MAPCA |
PGE1 infusion |
RV-PA shunt (10) |
Alive (84) | |
16 | M | 37 | 2,694 | AVD; DORV; PS; CAVB |
Left isomerism |
33 | fSRV mistaken for fSLV |
PMI (0) | Alive (19) | |
17 | F | 40 | 3,722 | DILV; hypoplastic RV |
30 | fSLV mistaken for fSRV |
PAB (27) | Alive (48) | ||
18 | M | 40 | 2,084 | PAVSD; PLSVC |
36 | LAA mistaken for RAA |
Alive (127) | |||
19 | M | 38 | 2,304 | TOF; RAA | 36 | RAA mistaken for LAA |
Alive (63) | |||
20 | F | 38 | 3,036 | Truncus; PAPVR; RAA |
31 | PAPVR missed | RVOTR IVR (9) |
Alive (39) | ||
21 | F | 38 | 3,340 | TOF; AVSD | Down syndrome |
29 | TOF mistaken for PS |
BT (23) | Alive (53) | |
22 | F | 37 | 1,316 | AVSD; DORV; AS; APW |
Cornelia de Lange syndrome |
34 | APW missed | Died during comfort care (5) |
||
23 | F | 41 | 2,832 | AVD; CCH; VSD; PA; MAPCA |
33 | CCH mistaken for TA |
Alive (20) | |||
24 | M | 38 | 2,558 | MA; DORV; PA; TAPVR; RAA |
Right isomerism |
33 | RI mistaken for SI |
Alive (117) |
APW, aortopulmonary window; AVD, atrioventricular discordance; AVS, aortic valve stenosis; bltPDA, bilateral patent ductus arteriosus; CAVB, complete atrioventricular block; CCH, criss-cross heart; HLHC, hypoplastic left heart complex; IUGR, intrauterine growth retardation; IVR, interventricular rerouting; LAA, left aortic arch; LV, left ventricle; MAPCA, major aortopulmonary colateral artery; MS, mitral stenosis; PA, pulmonary atresia; RtPDA, right patent ductus arteriosus; RV, right ventricle; RVOTR, right ventricular outflow tract reconstruction; SI, situs inversus. Other abbreviations as in Tables 1–4.
Discrepant prenatal and postnatal diagnoses were associated with VSD in 8 neonates. Prenatal evaluation missed VSD in 6 of them, and 1 required additional VSD closure that had not been planned at the time of fetal diagnosis. The location of VSD was mistaken in 2, and 1 underwent PAB instead of VSD closure.
Of the 3 neonates with discrepant diagnoses of pulmonary artery atresia, 2 needed continuous infusion of PGE1 to maintain patency of the ductus arteriosus. The morphological right and left ventricles and right and left aortic arch were misdiagnosed in 2 neonates each.
The neonate with Cornelia de Lange syndrome and a complex heart malformation, in whom an aortopulmonary window had been overlooked, died on postpartum day 5 under compassionate care without aggressive treatment as decided by the parents. Another neonate with multiple anomalies and CoA complex in whom VSD had been overlooked was diagnosed with rectal atresia and transferred to another institution on postpartum day 1. The other 22 patients survived and remained alive at discharge from hospital. Fetal discrepant diagnosis of CHD did not adversely affect short-term mortality. Meanwhile, 9 neonates (4% of all cases; 38% of discrepant diagnostic cases) underwent unexpected additional treatment.
Table 6 shows details of 19 neonates (male n=12; 63%) who were treated by catheterization and/or surgical intervention within 24 h postpartum. Median GA at birth was 38 weeks (range, 32–40 weeks) and only 1 infant was delivered before 37 weeks of GA. Median birth weight was 2,694 g (range, 2,332–3,564 g) and 5 neonates had a low birth weight (<2,500 g); 7 neonates (37%) were delivered by cesarean section. Median Apgar score was 8 (range, 4–8) at 1 min and 8 (range, 6–9) at 5 min. Neonatal asphyxia defined as Apgar score <7 was evident in 4 infants at 1 min and in 2 at 5 min.
No. | Sex | At birth | Delivery | Apgar score |
Diagnosis | Intervention within 24 h after birth |
Outcome at discharge (age in days) |
|||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
GA (weeks) |
Weight (g) |
1 min |
5 min |
Postnatal major cardiac diagnosis |
fSV | Systemic complication |
Discrepancy at fetal echocardiography |
GA at last fetal echocardiography (weeks) |
Catheter intervention |
Surgical intervention |
||||
1 | M | 38 | 3,266 | Vaginal | 8 | 9 | AVSD; DORV; PA; TAPVR |
fRSV | Right isomerism |
None | 34 | Stenting for DV stenosis |
Alive (61) |
|
2 | F | 39 | 2,858 | Vaginal | 8 | 7 | AVSD; PS; TAPVR |
fLSV | Right isomerism |
None | 33 | Stenting for DV stenosis |
Alive (41) |
|
3 | M | 39 | 3,030 | Vaginal | 8 | 8 | AVSD; DORV; PS; TAPVR |
fRSV | Right isomerism |
None | 36 | Stenting for DV stenosis |
Alive (47) |
|
4 | M | 40 | 2,590 | Vaginal | 6 | 7 | AVSD; PA; TAPVR |
fRSV | Right isomerism |
None | 38 | Stenting for DV stenosis |
Died (102) |
|
5 | F | 38 | 3,460 | Vaginal | 8 | 8 | AVSD; DORV; PA; TAPVR |
fRSV | Right isomerism |
None | 35 | Stenting for DV stenosis |
Died (58) |
|
6 | M | 39 | 3,006 | Vaginal | 8 | 8 | AVSD; DORV; PS; TAPVR |
fRSV | Right isomerism |
None | 35 | Stenting for DV stenosis |
TAPVR repair |
Died (83) |
7 | M | 37 | 2,888 | Cesarian | 8 | 8 | HLHS | fRSV | None | 34 | BAS for restrictive FO |
Alive (58) |
||
8 | F | 38 | 2,664 | Cesarian | 8 | 8 | HLHS | fRSV | None | 35 | BAS for restrictive FO |
ASD creation, bltPAB |
Alive (172) |
|
9 | M | 39 | 2,892 | Vaginal | 8 | 9 | HLHS | fRSV | None | 34 | BAS for restrictive FO |
bltPAB | Alive (183) |
|
10 | M | 39 | 3,006 | Vaginal | 7 | 8 | TGA | None | 31 | BAS for restrictive FO |
Alive (85) |
|||
11 | F | 38 | 2,440 | Cesarian | 7 | 7 | AVSD; DORV; PS; CAVB |
fRSV | Left isomerism |
None | 25 | PMI for CAVB |
Alive (74) |
|
12 | M | 37 | 2,694 | Cesarian | 6 | 7 | AVD; DORV; PS; CAVB |
fRSV | Left isomerism |
fRSV mistaken for fLSV |
33 | PMI for CAVB |
Alive (19) |
|
13 | M | 32 | 2,382 | Cesarian | 4 | 6 | AVSD; AS; CoA; CAVB |
Left isomerism |
None | 29 | PMI for CAVB |
Died (15) |
||
14 | F | 39 | 2,332 | Vaginal | 8 | 9 | AS | None | 31 | BAV | Alive (49) |
|||
15 | M | 37 | 2,604 | Vaginal | 8 | 8 | PAIVS; AS |
fLSV | None | 34 | BAV | Died (7) |
||
16 | M | 36 | 2,416 | Vaginal | 7 | 9 | TAPVR | None | 31 | TAPVR repair |
Alive (134) |
|||
17 | F | 37 | 2,470 | Vaginal | 8 | 8 | TAPVR | None | 30 | TAPVR repair | Alive (30) |
|||
18 | M | 40 | 3,564 | Cesarian | 6 | 6 | CAF | None | 39 | CAF closure |
Alive (74) |
|||
19 | F | 37 | 2,626 | Cesarian | 8 | 8 | Ebstein | fLSV | None | 34 | Starnes operation |
Died (15) |
bltPAB, bilateral pulmonary artery banding; CAF, coronary artery fistula; Ebstein, Ebstein anomaly. Other abbreviations as in Tables 1–5.
Dominant diagnoses of CHD were AVSD, HLHS and TAPVR in 8, 3 and 2 neonates, respectively; 13 neonates (68%) had a functional single ventricle: 10 and 3 had a functional right and left ventricle, respectively; 9 (47%) neonates were complicated with heterotaxy syndrome, with right isomerism in 6 and left isomerism in 3. Median GA at last fetal echocardiography was 34 weeks (range, 25–39 weeks). Pre- and postnatal diagnoses were discrepant in only 1 neonate with left isomerism in whom a functional right single ventricle was mistaken for a functional left single ventricle.
Postnatal predictions of the hemodynamics and treatments based on fetal diagnoses were correct in all neonates who were treated by catheter and/or surgical intervention within 24 h postpartum. Draining vein stenosis and stenotic lesions in 6 fetuses with TAPVR were correctly diagnosed prenatally. Restrictive foramina ovalia in 3 fetuses with HLHS and 1 fetus with TGA were also correctly diagnosed.
A total of 12 neonates required catheter intervention within 24 h postpartum. Stents were most frequently deployed to treat a stenotic drainage vein in TAPVR in 6 neonates who had severe cyanosis immediately after birth because of pulmonary venous obstruction, followed by BAS for a restrictive foramen ovale in 4 neonates who had severe cyanosis (because of pulmonary congestion in 3 with HLHS and insufficient blood communication between the left and right atria in the other 1 with TGA), and balloon valvuloplasty for critical aortic valve stenosis in 2.
A total of 10 neonates underwent the following surgical procedures within 24 h of delivery: pacemaker implantation for complete atrioventricular block complicated with left isomerism in which low output syndrome developed (n=3), repair of simple TAPVR with severe pulmonary congestion (n=2), repair of complex TAPVR after stent implantation failed (n=1), creation of an ASD after BAS failed (n=1), bilateral PAB after interatrial communication was established in HLHS (n=2), closure of a giant coronary artery fistula (n=1)23 and the Starnes operation for Ebstein anomaly with circular shunt (n=1).
In the only case of discrepant diagnosis (no. 16 in Table 5 and no.12 in Table 6) the neonate received prenatally planned pacemaker implantation in the first day of life because of low cardiac output from congenital heart block. He was discharged from hospital alive and the discrepant diagnosis did not adversely affect his outcome.
Of the 19 neonates who were seriously ill, 3 (16%) died as newborns and 6 (32%) died in hospital.
The reason relatively few fetuses with CHD were delivered by cesarean section was that hospital’s policy is to encourage transvaginal childbirth as much as possible, even when a fetus has been diagnosed with CHD.20
The initial screening for fetal heart abnormalities was late, at a median of 27 weeks of GA. Initial fetal echocardiography was also late, at a median 31 weeks of GA, and 88% received expert inspection only once. Fetal echocardiography screening and multiple expert inspections should proceed earlier, and this is recommended in various guidelines.1–3
Only 3 neonates were diagnosed with isolated VSD as the most prevalent type of CHD. A small VSD might have been overlooked at prenatal diagnosis, or fetuses that required neonatal intervention might have been selectively referred. A few cases of TAPVR might have been missed by fetal echocardiography screening, as reported previously.24–26 The relatively high frequency of complications with heterotaxy syndrome seems to be due to racial and ethnic characteristics.27 Chromosomal anomalies detected in this study were associated with CHD.28
Almost 50% of neonates received continuous infusion of PGE1. One major advantage of prenatal diagnosis is that a life crisis caused by neonatal closure of the ductus arteriosus with ductus arteriosus-dependent CHD can be avoided.9
Neonates with ductus arteriosus-dependent systemic circulation underwent ductal stenting with bilateral PAB according to hospital policy.29 Stents were deployed for stenotic lesions in the drainage vein of TAPVR when neonates were complicated with right isomerism, because the outcomes of neonatal surgery for this anomaly are poor.30
In this study, both neonatal and hospital mortality rates were high. Neonates admitted to the cardiac NICU with prenatal diagnoses of CHD might include relatively more critical cases. A comparative study with neonates not diagnosed with CHD until after birth is necessary.
Discrepant Diagnosis of CHD and Its ImpactThe rate of diagnostic discrepancy was considerably lower than previously reported,10–14 regardless of the relatively strict definition. Our expert fetal echocardiography provided by pediatric cardiologists who are specialists in fetal and neonatal heart disease functioned well.
Final diagnoses of CoA or IAA and the need for surgery to treat EDA were prevalent among neonates with discrepant diagnoses. CoA can progress with a postnatal ductus arteriosus restriction, whereas we often encounter aortic isthmus enlargement with postnatal increased vascular flow. Definitive diagnosis of CoA during the fetal period is difficult, and it is important to convey this to parents when explaining a prenatal diagnosis. In the present study 3 neonates with a prenatal diagnosis of hypoplastic left heart complex developed CoA after delivery. We now explain to parents that CoA can progress after delivery, particularly in neonates with a small left heart.
We overlooked VSD during prenatal diagnosis in 6 fetuses. When a VSD is morphologically small to medium-sized, blood flow shunts can be difficult to identify in the fetus because of the isotonic pressure between the right and left ventricles. We mistook the position of the VSD in 2 fetuses with DORV. The neonatal physiology might be completely different if the VSD is subaortic or subpulmonary, and the planned therapeutic interventions could be completely modified. A more accurate diagnosis is necessary, and new imaging techniques such as 3D echocardiography and fetal cardiac magnetic resonance (CMR) may help to precisely define this cardiac segment.13
We mistook pulmonary artery blood flow supply from the ductus arteriosus as truncus arteriosus and major aortopulmonary collateral arteries in 2 neonates who had pulmonary artery atresia. Unlike during the postpartum period, the pulmonary artery can be assessed without disturbing air in the lungs using fetal echocardiography. Fetal test precision may be improved using 3D echocardiography with color flow mapping.
There were discrepant prenatal diagnoses of morphological left ventricle or right ventricle. Morphological discrimination of a postpartum left or right ventricle by echocardiography can be difficult. Bilateral misdiagnoses of a right and left aortic arch might be caused by the right and left sides of fetuses becoming uncertain during tests. A basic operation should be established to determine the right and left sides of fetuses. Fetal CMR may resolve this problem.
We misdiagnosed 1 case of criss-cross heart as tricuspid atresia. The major error was in understanding the hemodynamics, and the treatment strategy was changed. We might have reached the correct diagnosis had we conducted a careful sweep using the 4-chamber view along the direction of the long axis of the fetus. 3D reconstruction by echocardiography or CMR may be useful.
Among neonates with discrepancies between pre- and postnatal diagnoses, fetal discrepant diagnosis of CHD did not adversely affect the neonatal clinical course, because these neonates received treatments according to the diagnosis immediately after birth. However, a change in the therapeutic plan to unexpected surgery places a mental burden on the parents. Improvements in the accuracy of diagnosis are expected. It is necessary to improve fetal heart screening level at primary and secondary facilities in order to improve the prognosis of the fetus in the future.
Neonates Requiring Immediate Intervention After BirthAmong the neonates who required invasive interventions within 24 h after birth, prenatal diagnoses were accurate. Pediatric cardiologists in the cardiac NICU participated in diagnoses based on fetal echocardiography. They are familiar with CHD in which symptoms of conditions such as HLHS or TGA with a restrictive foramen ovale and TAPVR with a stenotic drainage vein worsen immediately after delivery. Prenatal diagnostic accuracy was improved due to signs of these anomalies.
Although our judgements of draining vein stenosis and stenotic sites in TAPVR were correct, they were subjective and empirical. A prospective study in fetuses with TAPVR is necessary.
These neonates were seriously ill and there was a high mortality rate. However, those who left the hospital alive might have died without accurate prenatal diagnoses and appropriate immediate postnatal treatment.
Some reports have indicated that the benefit of antenatal diagnosis of duct-dependent CHD in terms of patient outcomes is due to early intravenous administration of PGE1 to maintain arterial duct patency.9,14 However, PGE1 can be immediately administered in any NICU that does not specialize in CHD treatment, if duct-dependent CHD is suspected postpartum. We believe that prenatal diagnosis confers an advantage on neonates who need invasive catheter intervention and/or surgery immediately postpartum. A comparative study between neonates with prenatal diagnosis of CHD and neonates diagnosed with CHD only after birth is necessary.
If precise methods and optimal timing for neonatal intervention can be established under a prenatal diagnosis of CHD, the limited special medical facilities will be used more effectively and treatment outcomes as well as cost-effectiveness should improve.
Study LimitationsThe main limitation was that only patients referred to the cardiac NICU under a prenatal diagnosis of CHD were included in the present study. Future multi-institutional studies should assess diagnoses based on expert fetal echocardiography.
We only investigated length of stay and mortality as markers of short-term prognosis during the first hospitalization. Further studies should assess the effects of prenatal incorrect diagnoses on intrauterine fetal death or artificial termination of pregnancy, as well as medium- and long-term mortality and postnatal growth and development of infants with CHD.
We detected discrepancies between the pre- and postnatal diagnoses of CHD in 12% of neonates admitted to a cardiac NICU. The most frequent discrepancy was a diagnosis associated with CoA, followed by VSD. Unexpected treatments were added to 38% of discrepant diagnostic cases. However, discrepant diagnoses did not adversely affect clinical course. Among the neonates in the present study, 9% required invasive intervention within 24 h after birth, and all were accurately diagnosed prenatally.
This retrospective study was approved by the Ethics Committee of the National Cerebral and Cardiovascular Center and performed in accordance with the Declaration of Helsinki.
We thank Seiko Araki, Aiko Kouda, Yuka Nakamura, Fujiko Nakamoto, Mitsuyo Saegusa, Emi Yamashita and Hiromi Ikado for their role in performing fetal echocardiography. This study was supported by the Intramural Research Fund (29-4-1) for Cardiovascular Disease of the National Cerebral and Cardiovascular Center.
I.S. is an Associate Editor of Circulation Journal. S.Y. is a member of Circulation Journal ’s Editorial Board. This study was supported by the Intramural Research Fund (29-4-1) for Cardiovascular Disease of the National Cerebral and Cardiovascular Center.
This retrospective study was approved by the Ethics Committee of the National Cerebral and Cardiovascular Center (M29-068) and performed in accordance with the Declaration of Helsinki.