1. Outcome of Prenatally Diagnosed Congenital Heart Disease: An Update
D.H. Brick, L.D. Allan
Department of Pediatric Cardiology, Babies Hospital 2N, New York Presbyterian Hospital, 3959 Broadway, New York, NY 10032, USA
Abstract. The objective of this study was to compare
the outcome of prenatally diagnosed congenital heart
disease (CHD) in a recent time period with previously
reported experience. All cases of fetal CHD during
the time period 1993 to 1999 were analyzed in terms
of their outcome. During this time frame, 408 cases of
fetal CHD were detected. Of these, 84% of mothers
were referred because of suspicion of a cardiac
anomaly during an obstetric scan. The mean gesta-
tional age at diagnosis was 26 weeks. Termination of
pregnancy occurred in 98 cases. There were 92 deaths
in the continuing pregnancies and a survival rate of
60%. The detection rate of CHD prenatally continues
to increase as obstetric screening for cardiac malfor-
mations becomes more widespread. Diagnostic cate-
gories continue to be skewed toward more complex
forms of malformation, although comparisons with
previous studies show some trends. In particular,
there are proportionately fewer pregnancies with ex-
tracardiac malformations. In addition, the outcome
in continuing pregnancies is substantially improved
from previous reports.
Key words: Congenital heart disease Ð Fetus Ð
Prenatal diagnosis
In a previous publication, a series of 1006 prenatally
diagnosed cases of congenital heart disease (CHD)
extending over the period 1980 to 1992 was described
from Guy's Hospital in London. Upon relocation to
New York in 1993, a fetal cardiac referral unit was
established at New York Presbyterian Hospital. It
has been shown that the fetal spectrum of disease is
di€erent from that seen postnatally [1] and that the
outcome of fetally diagnosed malformations is poorer
than that expected from experience after birth [2, 6,
10], despite optimizing perinatal management in
those fetally diagnosed cases in order to avoid early
neonatal decompensation [5, 8]. However, during this
time period, there have been improvements in neo-
natal survival for all lesions, especially for the hypo-
plastic left heart syndrome, which in the earlier
period was left untreated. We therefore analyzed re-
cent results for comparison with our previous expe-
rience and with other reports.
Material and Methods
Of 2500 fetuses referred for fetal echocardiography during the 7-
year period, 408 were found to have congenital heart malforma-
tions. There were approximately 400 referrals each year; this rate of
referral remained fairly stable. There was an overall positivity rate
in referred pregnancies of 16% but the rate of CHD increased from
10% in 1994 to 25% in 1999. Fetuses found to have rhythm dis-
orders were not included in this group unless there was also a
structural anomaly. The gestational age range, the diagnosis, the
reason for referral, and the outcome of each pregnancy were ana-
lyzed.
Results
The gestational age at diagnosis ranged from 15 to 42
weeks, with a mean of 26 weeks. The diagnostic
categories are shown in Table 1. The reason for re-
ferral was suspicion of CHD during an obstetric scan
in 343 cases (84%). Other indications included ma-
ternal diabetes in 4, a family history of CHD in 13, an
extracardiac fetal anomaly in 29, a fetal arrhythmia
in 3, fetal hydrops in 4, and advanced maternal age in
1. The reason for referral was unknown in the re-
maining 11. Chromosomal analysis was obtained in
all ongoing pregnancies and in more than half of the
pregnancies in which termination took place. Of 408
cases, 63 had documented extracardiac malforma-
tions, of which 39 were proven to have chromosomal
anomalies. Of these, there were 16 cases of trisomy
21, 12 of trisomy 18, and 4 of Turner's syndrome.
There were 187 cardiac anomalies (46%) diagnosed
prior to 24 weeks of gestation, the legal gestational
age limit for termination of pregnancy in New YorkCorrespondence to: L.D. Allan
Pediatr Cardiol 23:449±453, 2002
DOI: 10.1007/s00246-002-1330-x

2. State. Of these 187, 93 sets of parents elected ter-
mination of pregnancy (49.7%). An additional
5 mothers traveled out of state to achieve interrup-
tion of pregnancy at later gestations. There were 19
intrauterine deaths, 54 neonatal deaths, and 19
deaths after the neonatal period. There are 185
known survivors, with follow-up ranging from 1 week
to 6 years. There are 13 continuing pregnancies, with
20 lost to follow-up. Therefore, there was at least a
33% mortality in the continuing pregnancies. Those
cases lost to follow-up were evenly distributed across
the diagnostic spectrum.
Discussion
In a referral center, it is impossible to know the de-
nominator of the population screened for CHD and
the resulting detection rate during routine scanning.
All that can be said is that of a constant number of
cases referred for fetal echocardiography (approxi-
mately 400 cases per year), an increasing proportion
were found to have CHD. The four-chamber view of
the heart is a recommended part of obstetric ultra-
sound evaluation [11]. As a result, malformations
detectable in this view predominate in all fetal series,
with atrioventricular septal defects and the hypo-
plastic left heart syndrome being the most commonly
detected in this and other series (Table 2). The simi-
larity of the diagnostic categories between each of the
fetal series previously reported is remarkable. There
are 408 cases in this series, 847 cases in the multi-
center series from Italy [6], 170 in the series from Yale
[10], and 1006 in the series from London [2], repre-
senting a total of 2431 cases. There is a fair degree of
consistency in the ®ndings in three di€erent countries
spanning a 20-year period. As noted previously, the
spectrum of disease is di€erent from that seen in
postnatal life, with hypoplastic left heart syndrome
and atrioventricular septal defect occurring two or
three times more frequently as in postnatal life [7, 9].
In contrast, transposition of the great arteries is
consistently underrepresented in fetal series; it is less
than half as common as found postnatally, although
the rate of detection has more than doubled between
our previous and our current series. The rate of de-
tection of tetralogy of Fallot, perhaps because of a
high association of this lesion with extracardiac
malformations, now approaches the expected post-
natal rate. The rate of detection of ventricular septal
defects has increased over time but is still about half
the rate expected in infants. The increased detection
of ventricular septal defects appears to be due to a
combination of improvements in ultrasound image
quality, including the development of higher fre-
quency transducers, the introduction of harmonic
imaging, improved ``zoom'' capabilities, and better
color penetration, which have been available to us for
the past 2 years. It is of interest that in the Italian
study, secundum atrial septal defects comprised 3.6%
2Table 1. Diagnostic categories
Outcome
Diagnosis No. of cases TOP IUD NND Inf/ChD A
Atrioventricular septal defect 67 18 4 5 5 31
Hypoplastic left heart syndrome 64 17 1 18 7 21
Tetralogy of Fallot 40 10 2 6 2 18
Ventricular septal defect 33 4 4 4 Ð 16
Mitral atresia 23 14 1 1 1 4
Transposition 21 3 1 Ð 1 16
Double-outlet right ventricle 18 6 Ð 2 1 9
Ebstein's/tricuspid dysplasia 18 6 Ð Ð Ð 8
Coarctation 17 4 Ð 2 1 10
Pulmonary stenosis/atresia 16 3 Ð 3 Ð 8
Aortic stenosis 12 2 1 3 Ð 4
Tumors 13 1 1 1 Ð 10
Tricuspid atresia 11 4 1 1 Ð 5
Cardiomyopathy 11 5 2 Ð 1 3
Common arterial trunk 7 1 Ð 1 Ð 5
Corrected TGA 7 1 Ð Ð Ð 6
Double-inlet ventricle 5 1 Ð Ð Ð 6
Atrial septal defect, secundum 1 Ð Ð Ð Ð 1
Miscellaneous 22
Total 406
Where numbers do not add up, patients were lost to follow-up (22 patients). A; Inf/ChD, infant death/death in childhood; IUD, intrauterine
death; NND, neonatal death; TOP, termination of pregnancy.
450 Pediatric Cardiology Vol. 23, No. 4, 2002

3. of cases prenatally diagnosed, although this diagnosis
was rarely made in the other fetal series. Indeed, this
rate of diagnosis of atrial septal defects is higher than
that reported in infants [7, 9]. It is dicult to imagine
how a secundum atrial septal defect can be diagnosed
in the presence of a normal foramen ovale prenatally,
unless there is virtually a common atrium as occurred
in one of our cases. Even after birth, a secundum
atrial defect is not reliably diagnosable in the ®rst 6
months of life because during this time atrial com-
munications commonly close spontaneously.
The gestational age range was the same in this, the
Italian, and our previous series, with malformations
being detected for the ®rst time at 15 weeks and at all
gestational ages up to term. Gestational age was not
stated in the Yale series. Of note, 43% of malforma-
tions were detected prior to 24 weeks of gestation in the
Italian series, 46% in the series, and 68% in the London
series. This higher rate of early detection in the United
Kingdom re¯ects an almost universal routine scanning
policy in that country between 18 and 20 weeks of
gestation, resulting in the fetal detection of serious
congenital heart in 23% of a€ected pregnancies, albeit
with marked geographical di€erences in the detection
rate [4]. In the United States, there is no such policy
and the preference is for scanning on ``indication
only.'' In practice, almost all pregnancies in the United
States are scanned, but at later gestational ages, and
scanning is focused, rather than comprehensive, and
therefore does not always include cardiac evaluation.
However, the number of cases detected as a result of a
suspicion of cardiac anomaly during an obstetric scan
in the United States have increased over time. Such
cases comprised 84% of this series as opposed to 33%
of the earlier Yale series.
As1 fetal echocardiography has been employed in
more cases because of abnormal screening scans, the
rate of extracardiac and chromosomal anomalies has
decreased as a proportion of the cases of CHD de-
tected prenatally. In addition, it may re¯ect a change
in the practice of the perinatologist in interrupting a
pregnancy with multiple malformations and a
chromosomal anomaly, without referral to the pedi-
atric cardiologist. In the Yale series, 32% had extra-
cardiac anomalies and 28% chromosomal anomalies,
whereas the rates in the current series were 16% and
9%, respectively. The rate of chromosomal anomalies
was 17% in our previous series and 17.5% in the
Italian series, and an additional 17% in that series
had extracardiac malformations. As a result of the
decrease in the proportion of multiple anomalies,
there has been a reduction in the proportion of
spontaneous intrauterine deaths from 11.9%, 11%,
and 9% in the Italian, London, and Yale series, re-
spectively, to 5% in the current series. Although there
was a high rate of chromosomal analysis, some spe-
ci®c diagnostic groups were not routinely karyotyped
because of the rarity of association with chromoso-
mal anomalies, for example, the heterotaxy syn-
dromes or simple transposition of the great arteries.
In some of the interrupted pregnancies, the decision
was based on the severity of the heart disease and
chromosomal analysis was not always performed. A
small number of cases of chromosomal abnormality
3Table 2. Percentages of malformations
Diagnosis %
Brick Fesslova et al. [6] Smythe et al. [10] Allan et al. [2]
Atrioventicular septal defect 16 16 22 17.5
Hypoplastic left heart syndrome 15 14 13 16
Tetralogy of Fallot 10 7.8 8.2 3
Ventricular septal defect 8.0 12 8.8 5
Mitral atresia 5.6 5 6.4 5
Transposition 5.6 4.6 3.5 2
Double-outlet right ventricle 4.4 3.6 8.8 3
Coarctation 4.2 5.6 2.9 11
Ebstein's/tricuspid dysplasia 4.4 16 4.1 7
Pulmonary stenosis/atresia 3.9 4 5.2 5
Aortic stenosis 2.9 2.7 6.4 4
Tumors 2.9 1.8 Ð 1
Tricuspid atresia 2.7 4 2.3 4
Cardiomyopathy 2.7 6.5 Ð 4
Common arterial trunk 1.7 1.6 1.2 1.5
Corrected TGA 1.7 2 Ð 0.3
Double-inlet ventricle 1.2 Ð Ð 2
Atrial septal defect 0.2 3.6 1.2 Ð
Others 7.1 1.3 5.8 0.5
Brick and Allan: Outcome of Fetal Congenital Heart Disease 451

4. may therefore have been overlooked. Similarly, there
may have been extracardiac malformations in addi-
tion to the cardiac lesion in some cases in which
termination, without subsequent autopsy, occurred.
Termination of pregnancy occurred in 29% in the
Italian series, 55% in the London series, 45% in the
Yale series, and 24% in the current series. This decision
is strongly related to gestational age at diagnosis,
which was late in the current series with a mean of 26
weeks. Termination was chosen in nearly 50% of those
cases diagnosed prior to 24 weeks of gestation. Deci-
Fig. 1. The overall outcome of 408 cases of fetal CHD in all di-
agnostic categories. TOP, termination of pregnancy; NND, neo-
natal death; LTFU, lost to follow-up; IUD, intrauterine death;
InfD, infant death; ChD, death in childhood.
Fig. 2. The percentage survival is shown broken down into three
time periodsÐ1993 to 1996, 1997 and 1998, and 1999. There are
three diagnostic groupsÐthe total case of CHD, the cases of the
hypoplastic left heart syndrome (HLHS), and those of atrioven-
tricular4 canal defect (AVC).
Fig. 3. The percentage of patients alive at last follow-up in con-
tinuing pregnancies, for each main diagnostic category, is shown.
``Simple'' lesions, such as transposition of the great arteries, are
associated with a good outcome, whereas the outcome for the severe
left heart lesions, such as the hypoplastic left heart syndrome and
mitral atresia, remains poor despite improving surgical results. Corr
TGA, corrected transposition; D-TGA, transposition of the great
arteries; ASD, atrial septal defect; VSD, ventricular septal defect;
AVC, atrioventricular canal defect; DORV, double-outlet right
ventricle; Tat, tricuspid atresia; TOF, tetralogy of Fallot; PA/PS,
pulmonary atresia/pulmonary stenosis; Coa, coarctation; HLHS,
hypoplastic left heart syndrome; MAt, mitral atresia.
452 Pediatric Cardiology Vol. 23, No. 4, 2002

5. sions concerning termination are also in¯uenced by
diagnostic category, which is becoming more benign,
and surgical options, which are improving. The overall
outcome is illustrated in Fig. 1. There was a 45% sur-
vival in continuing pregnancies in the Italian series,
41% in the London series, 38% in the Yale series, and
60% in the current series. Grouping all diagnostic
categories, there has been an improvement in survival
for each calendar year (Fig. 2). Although length of
follow-up is clearly shorter in the later patients, these
data appear to demonstrate improving results for the
management of critically ill infants with CHD. If our
two largest groups are examined by time of diagnosis,
survival improved in the hypoplastic left heart syn-
drome and in the group with atrioventricular septal
defect (Fig. 3). However, the outcome still appears to
be worse than expected from pediatric practice. This is
because of the bias in fetal series, compared to post-
natal series, toward more complex disease as shown in
the diagnostic categories seen in Table 2. The outcome
for each diagnostic category is shown in Fig. 4. Pre-
natal diagnosis has been shown to result in a better
homodynamic status in neonates than in those cases
diagnosed postnatally [5, 8]. This can result in im-
provements in mortality in those fetally diagnosed
when sucient numbers of the same diagnostic cate-
gory are compared to those diagnosed after birth [3].
In summary, the detection rate of CHD prena-
tally continues to increase as obstetric screening for
cardiac malformations becomes more widespread. In
conjunction with this, the gestational age at diagnosis
is likely to decrease in the near future. As a result,
interruption of pregnancy may become more fre-
quent, particularly in cases of complex disease and in
those cases associated with extracardiac or chrom-
osomal anomalies. On the other hand, improved
cardiac scanning skills in obstetric practice may lead
to the increased detection of more benign or suc-
cessfully repaired forms of heart disease, resulting in
a fetal spectrum of disease more similar to that seen
in infants. In continuing pregnancies, a better out-
come can be expected in those diagnosed prior to
birth as a result of planned delivery of the neonate
within the cardiac center and optimization of imme-
diate postnatal care for each diagnostic category.
References
1. Allan LD, Crawford DC, Anderson RH (1985) Spectrum of
congenital heart disease detected echocardiographically in
prenatal life. Br Heart J 54:523±526
2. Allan LD, Sharland GK, Milburn A, et al (1994) Prospective
diagnosis of 1006 consecutive cases of congenital heart disease
in the fetus. J Am Coll Cardiol 23:1452±1458
3. Bonnet D, Coltri A, Butera G, et al (1999) Fetal detection of
transposition of the great arteries reduces morbidity and
mortality in newborn infants. Circulation 99:916±918
4. Bull C (1999) Current and potential impact of fetal diagnosis
on prevalence and spectrum of serious congenital heart disease
at term in the UK. Lancet 354:1242±1247
5. Copel JA, Tan SA, Kleinman CS (1997) Does prenatal diag-
nosis of congenital heart disease alter short-term outcome?
Ultrasound Obstet Gynecol 10:237±241
6. Fesslova V, Nava S, Villa L, and the Fetal Cardiology Study
Group of the Italian Society of Pediatric cardiology (1999)
Evolution and long-term outcome in cases with fetal diagnosis
of congenital heart disease: Italian multicentre study. Heart
82:594±599
7. Fyler DC, Buckley LP, Hellenbrand WE, Cohn HE (1980)
Report of the New England Regional Infant Cardiac Care
Program. Pediatrics 65(Suppl):376±446
8. Kumar RK, Newburger JW, Gauvreau K, Kamenir SA,
Hornberger LK (1999) Comparison of outcome when hypo-
plastic left heart syndrome and transposition of the great ar-
teries are diagnosed prenatally versus when diagnoses of these
two conditions are made only postnatally. Am J Cardiol 83:
1649±1653
9. Scott DJ, Rigby ML, Miller GAH, Shinebourne EA (1984)
The presentation of symptomatic heart disease in infancy
based on 10 years experience (1973±1982). Implications for the
provisions of services. Br Heart J 52:248±257
10. Smythe JF, Copel JA, Kleinman CS (1992) Outcome of pre-
natally detected cardiac malformations. Am J Cardiol 69:
1471±1474
11. Ultrasonography in pregnancy (1993) ACOG Tech Bull 187:
1±8
Brick and Allan: Outcome of Fetal Congenital Heart Disease 453