2. Pediatric Cardiology
1 3
Methods
Patient Selection
The Heart Center echocardiographic database was queried
for fetal studies performed at our institution between Janu-
ary 2002 and January 2017 with diagnostic codes includ-
ing the search terms mitral valve, aortic valve, stenosis,
hypoplasia, hypoplastic left heart syndrome, coarctation,
and arch. Initial fetal echocardiograms and corresponding
reports from studies interpreted by the contemporary fetal
cardiologist as COA, LHH, HLHS, mitral valve hypoplasia
(MVH) and/or stenosis, and aortic valve hypoplasia and/
or AS were included. The reported fetal diagnosis was
used for analysis. Fetuses with concern for small left-sided
structures and not committed to a diagnosis of COA or
HLHS by the diagnosing physician were coded as LHH.
Mitral valve hypoplasia was diagnosed in a fetus with a
mitral valve with Z score < − 2 but normal aortic annular
dimension and left ventricular size.
Corresponding neonatal clinical, echocardiographic,
and surgical records were reviewed independently to pre-
vent bias. Subjects were excluded if multiple fetal gesta-
tions were present; if double-inlet left ventricle, double-
outlet right ventricle, or atrioventricular septal defect was
suspected; if postnatal records including neonatal transtho-
racic echocardiogram were not available; or if the family
elected termination, comfort care, or delivery at another
institution. This project was approved by the Institutional
Review Board for Stanford University (Protocol Number
40371).
Echocardiographic Data
Standard fetal echocardiographic parameters were meas-
ured by a single investigator (LAE), blinded to postnatal
outcome at time of assessment. Cardiac measurements
included atrioventricular valve diameters, semilunar valve
diameters, main pulmonary artery diameter, ascending
aorta diameter, ductus arteriosus diameter, transverse
aortic arch diameter, aortic isthmus diameter, and ven-
tricular lengths [17–19]. A second fetal cardiologist (AA)
measured mitral valve annulus, aortic valve annulus, and
ascending aorta in a subset (n = 10) of randomly selected
patients for inter-observer variability analysis. RV and LV
length were measured in the apical four-chamber view at
end-diastole from the center of the atrioventricular valve
to the apical endocardium, and RV/LV length ratio was
calculated accordingly [20]. TV/MV ratio was calculated
by dividing the maximal diastolic annular dimension of
the tricuspid valve by that of the mitral valve [20]. Color
and spectral Doppler evaluation of the atrioventricu-
lar valves, semilunar valves, foramen ovale, pulmonary
vein, transverse aortic arch, umbilical artery, umbilical
vein, ductus venosus, and middle cerebral artery were
also recorded. Direction of flow at the foramen ovale was
classified as normal (right-to-left) or abnormal (bidirec-
tional or left-to-right). Pulmonary venous Doppler pro-
files were considered abnormal if prominent retrograde
flow during atrial systole was noted. Direction of flow in
the ascending aorta was classified as normal (prograde)
or abnormal (bidirectional or retrograde). Fetal biometry
measurements were made in accordance with practice
guidelines and included heart rate, biparietal diameter,
and cardiothoracic ratio [21]. Other calculations included
cardiovascular profile score [22], LV ejection fraction,
RV fractional area change, velocity time integrals across
semilunar valves, combined ventricular output, ventricular
stroke volume, and middle cerebral and umbilical artery
pulsatility indices. Measurements were converted to ges-
tational age-based Z scores (determined by estimated due
date from referring obstetrician) using available norma-
tive data [23–27]. A novel measurement, RV–LV length Z
score discordance, was calculated by subtracting the LV
length Z score from the RV length Z score [20]. Results
from invasive prenatal genetic testing (amniocentesis and
chorionic villus sampling) were recorded when available.
Clinical Data
Postnatal data included diagnoses, prostaglandin initiation,
neonatal intervention, hospital length of stay, survival to
discharge, mortality, and genetic testing results. Immediate
postnatal echocardiograms were retrospectively reviewed
after fetal echocardiogram assessment by a single reader
(LAE) to minimize outcome bias. Neonatal intervention
was defined as cardiac catheterization or surgery performed
in the first 30 days of life. For “intervention” versus “no
intervention” analysis, neonates who required intervention
but were not surgical candidates or died prior to interven-
tion were included in the “intervention” subgroup. For BVR
versus SVP analysis, neonates who required intervention
but died before intervention or were not surgical candidates
were not included in the analysis. Postnatal genetic testing
included fluorescent in situ hybridization, chromosomal
microarray, whole exome sequencing, and karyotype.
Statistical Analysis
All analyses were performed using SPSS version 24 (IBM;
Armonk, New York, USA) and SAS Enterprise Guide Ver-
sion 7.1 (SAS Institute Inc., Cary, NC, USA). Intraclass
correlation coefficients (ICC) were derived for fetal mitral
valve annulus, aortic valve annulus, and ascending aorta
3. Pediatric Cardiology
1 3
measurements to test for inter-observer variability. Prenatal
variables were compared by fetal diagnostic group (COA,
LHH, HLHS, AVS, MVH), neonatal intervention versus no
intervention, and BVR versus SVP. Continuous variables
were reported as mean (standard deviation) and median
(range). Categorical variables were reported as ratios (per-
centages). Patient death was dealt with statistically as fol-
lows: those who died prior to intervention were included in
the ‘required intervention’ group in analysis; those who died
prior to being committed to a SVP or BVR were excluded
from surgical outcome analysis; and patients who died prior
to discharge were included in length of stay calculations.
The Shapiro–Wilk test was applied to all variables to test
for normality. For normally distributed data, ANOVA with
post hoc Tukey was used for multiple comparison groups,
while a Student’s t test was used for two comparison groups.
Non-normally distributed variables were compared with a
Kruskal–Wallis test with a post hoc Bonferroni (multiple
groups), and a Mann–Whitney U test was used for two
comparison groups. Categorical data were compared with a
Fisher’s exact test. Fetal echocardiographic parameters that
were significantly different among fetal diagnostic groups
in univariate comparison (level of significance of p < 0.05)
were selected for inclusion in stepwise multivariable logisti-
cal regression models to determine factors associated with
need for neonatal intervention and single-ventricle out-
come. All continuous and dichotomous variables deemed
significant by univariate analysis were put into the multivari-
ate model. Receiver-operating characteristic curve analyses
were performed to determine threshold values for continu-
ous variables as well as corresponding sensitivity/specificity
for assessing the need for neonatal intervention and single-
ventricle outcome.
Results
Patient Data
Of 98 studies identified with congenital left-sided lesions,
68 mothers met inclusion criteria. Thirty mothers were
excluded; of these, 13 families elected termination of preg-
nancy, seven mothers had no further records at our institu-
tion, three chose comfort care, three delivered at outside
hospitals, two had twin pregnancies, and two suffered intrau-
terine demise. Of the 68 mothers who delivered at our insti-
tution after fetal cardiac diagnosis, only three were local;
the remaining 65 relocated within a ten-mile radius to our
hospital prior to delivery.
Table 1 displays neonatal characteristics with no differ-
ence in gestational age at delivery (p = 0.90) or birth weight
(p = 0.12) among fetal diagnostic groups. Neonates with a
fetal diagnosis of COA had a higher proportion of genetic
abnormalities than neonates with a fetal diagnosis of HLHS
Table 1 Patient characteristics by fetal diagnostic group
All values expressed as median (range), mean ± standard deviation, or n/denominator (percentage)
AS aortic stenosis, COA coarctation of the aorta, HLHS hypoplastic left heart syndrome, LHH left heart hypoplasia, LOS length of stay, MVH
mitral valve hypoplasia
Cohort
N = 68
COA
N = 15
LHH
N = 9
HLHS
N = 39
AS
N = 4
MVH
N = 1
Gestational age at birth in weeks 38.1 ± 1.6 37.9 ± 2.2 38.5 ± 0.9 38.1 ± 1.6 38.2 ± 1.3 37.3
38.4 (31.4–40.1) 39.0 (32.0–40.0) 38.3 (37.0–39.7) 38.4 (31.4–40.1) 37.9 (37.0–40.1)
Birth weight in kg 3.0 ± 0.6 2.7 ± 0.7 3.0 ± 0.5 3.2 ± 0.6 3.1 ± 0.4 3.5
3.0 (1.0–4.6) 2.7 (1.0–3.7) 3.1 (2.2–4.0) 3.2 (1.5–4.6) 3.0 (2.8–3.6)
Abnormal genetics 12/58 (21%) 5/13 (33%) 2/5 (40%) 4/37 (11%) 0/2 (0%) 1/1 (100%)
Intervention
Not required 7/68 (10%) 2/15 (13%) 2/9 (22%) 1/39 (3%) 1/4 (25%) 1/1 (100%)
Not candidate 5/68 (7%) 0/15 (0%) 1/9 (11%) 3/39 (8%) 1/4 (25%) 0/1 (0%)
Arch repair 14/68 (21%) 11/15 (73%) 3/9 (33%) 0/39 (0%) 0/4 (0%) 0/1 (0%)
Biventricular 3/68 (6%) 2/15 (13%) 0/9 (0%) 0/39 (0%) 1/4 (25%) 0/1 (0%)
Norwood 39/68 (57%) 0/15 (0%) 3/9 (33%) 35/39 (90%) 1/4 (25%) 0/1 (0%)
Diagnostic accuracy 60/68 (88%) 13/15 (87%) 7/9 (78%) 38/39 (97%) 3/4 (75%) 0/1 (0%)
Catheterization 18/68 (26%) 0/15 (0%) 0/9 (0%) 15/39 (38%) 3/4 (75%) 0/1 (0%)
LOS in days 41 ± 41 26 ± 17 42 ± 57 47 ± 45 41 ± 37 4
29 (0–202) 21 (5–66) 25 (5–187) 33 (0–202) 28 (12–94)
Survival to discharge 58/68 (85%) 15/15 (100%) 9/9 (1%) 30/39 (77%) 3/4 (75%) 1/1 (100%)
Overall survival 53/68 (78%) 9/10 (93%) 8/9 (89%) 27/39 (69%) 3/4 (75%) 1/1 (100%)
4. Pediatric Cardiology
1 3
(p = 0.04); no other genetic differences were observed among
the groups. In the COA fetal diagnostic group, 5/13 tested
had genetic abnormalities (one trisomy 21, three 22q11.2
distal deletions, and one unbalanced translocation), while
only 4/37 tested in the HLHS group had genetic abnormali-
ties (two 11q deletions/Jacobsen Syndrome, one 45, X0/
Turner’s Syndrome, and one microdeletion of 5p15.31/
microduplication 15q13.3).
Echocardiographic Data
Table 2 compares echocardiographic parameters among
fetuses by fetal diagnostic group. Gestational age at first
fetal echocardiogram ranged from 19 to 38 weeks (median
28 weeks). Cardiovascular profile score (p = 0.02), pres-
ence of left-to-right shunting at the patent foramen ovale
(p < 0.001), presence of an abnormal pulmonary vein Dop-
pler (p = 0.04), mitral valve Z score (p < 0.001), TV/MV
ratio (p < 0.001), left ventricular length Z score Z score
(p < 0.001), RV/LV length ratio (p < 0.001), RV–LV length Z
score discordance (p < 0.001), left ventricular ejection frac-
tion (p < 0.001), aortic valve Z score (p < 0.001), presence
of prograde aortic flow (p < 0.001), ascending aorta Z score
(p < 0.001), aorta/main pulmonary artery ratio (p < 0.001),
aortic isthmus Z score (p = 0.03), and presence of retrograde
flow in the aortic arch (p < 0.001) differed significantly
among the fetal diagnostic groups. On post hoc analysis,
mitral valve Z score, TV/MV ratio, left ventricular length Z
score, RV/LV length ratio, RV–LV length Z score discord-
ance, aortic valve Z score, presence of left-to-right shunting
at the patent foramen ovale, and presence of prograde aortic
flow differed among multiple fetal diagnostic groups.
Parameters Associated with Need for Neonatal
Cardiac Intervention and Single‑Ventricle Palliation
When neonates requiring neonatal cardiac intervention were
compared to those who were discharged without interven-
tion, lower LV length Z score (p = 0.004), aortic valve Z
score (p = 0.01), ascending aorta Z score (p = 0.02), and
aorta/main pulmonary artery ratio (p = 0.005); left-to-right
shunting at the foramen ovale (p = 0.04); and retrograde flow
in the aortic arch (p = 0.008) were associated with need for
neonatal intervention (Table 3). When we evaluated if any of
these parameters were associated with aortic arch or aortic
arch/ventricular septal defect repair versus no intervention,
none of the echocardiographic parameters were statistically
different. Lower mitral valve Z score (p < 0.001), LV length
Z score (p < 0.001), aortic valve Z score (p < 0.001), ascend-
ing aorta Z score (p < 0.001), aorta/main pulmonary artery
ratio (p < 0.001), and LV ejection fraction (p < 0.001), as
well as higher TV/MV ratio (p < 0.001) and RV/LV length
ratio (p < 0.001), RV–LV length Z score discordance
(p < 0.001), left-to-right shunting at the foramen ovale
(p < 0.001), abnormal pulmonary vein Doppler (p = 0.008),
absence of prograde aortic flow (p < 0.001), and retrograde
flow in the aortic arch (p < 0.001) were associated with SVP
(Table 4). Neonates with genetic abnormalities were more
likely to undergo biventricular repair (p = 0.04). In stepwise
logistical regression, the strongest independent variable
associated with SVP was RV/LV length ratio (p = 0.03); an
RV/LV length ratio > 1.28 predicted SVP with a sensitivity
of 76% and specificity of 96% (AUC 0.90, p < 0.001). Left-
to-right shunting at the foramen ovale (PFO) also predicted
SVP (p = 0.05).
Reproducibility
A high degree of absolute agreement among fetal echocar-
diographic parameters was observed: mitral valve annulus
ICC = 0.92, aortic valve annulus ICC = 0.74, and ascending
aorta ICC = 0.90.
Discussion
Fetal Echocardiographic Parameters Associated
with Neonatal Intervention
For fetuses at the milder end of spectrum, predicting poten-
tial neonatal intervention is challenging. We found that in
the cohort as a whole, lower LV length Z score, aortic valve
Z score, ascending aorta Z score, and aorta/main pulmonary
artery ratio; left-to-right shunting at the foramen ovale; and
retrograde flow in the aortic arch were associated with need
for neonatal intervention. Our study did not identify echo-
cardiographic differences between those not requiring inter-
vention and the small subset of our population requiring an
aortic arch or aortic arch/VSD repair. Prenatal diagnosis of
coarctation of the aorta is notoriously difficult, and fetal car-
diologists have struggled to identify fetal cardiac parameters
that consistently predict the need for neonatal arch repair.
Jowett et al. [28] found that the presence of continuous flow
in the aortic arch, aortic isthmus Z score, isthmus-to-duct
ratio, and the presence of a shelf in combination were most
sensitive and specific for the need for neonatal coarctation
repair, but none of the parameters were, by themselves, pre-
dictive of outcome. Matsui et al. [29] found the presence of
isthmal flow disturbance to be predictive of need for neona-
tal surgery. Aortic isthmus Z score, retrograde flow in the
aortic arch, and ductus arteriosus Z score were not predictive
of need for neonatal aortic arch repair in this study. While
the intention of this study was not to determine fetal echo-
cardiographic indices associated with aortic arch repair, the
finding that none of the echocardiographic parameters we
evaluated are associated with need for neonatal coarctation
7. Pediatric Cardiology
1 3
repair highlights the difficulty in prenatal diagnosis of coarc-
tation of the aorta.
Fetal Echocardiographic Parameters Associated
with SVP
When we compared fetal parameters in neonates who
underwent SVP versus those who underwent BVR, lower
mitral valve Z score, LV length Z score, aortic valve Z score,
ascending aorta Z score, aorta/main pulmonary artery ratio,
and LV ejection fraction, as well as higher TV/MV ratio
and RV/LV length ratio, RV/LV length Z score discordance,
left-to-right shunting at the foramen ovale, abnormal pul-
monary vein Doppler, absence of prograde aortic flow, and
retrograde flow in the aortic arch were associated with SVP.
The strongest independent variable associated with SVP by
multivariable analysis was RV/LV length ratio; an RV/LV
length ratio > 1.28 was associated with SVP with a sensitiv-
ity of 76% and specificity of 96%. Any left-to-right shunting
at the PFO was also associated with SVP by multivariate
analysis.
Few prenatal models have been developed to predict SVP
in fetuses with borderline left-sided cardiac structures. Bolin
et al. [30] reported that a prenatal aortic valve Z score plus
mitral valve Z score of − 6.4 or less predicted SVP with 83%
sensitivity and specificity, while retrograde flow in the aortic
arch was specific (95%) but not sensitive (60%) for SVP. McEl-
hinney et al. found that in fetuses with valvar AS with high
likelihood of progression to HLHS and who underwent fetal
aortic valvuloplasty, a total threshold score (derived from pre-
intervention fetal echocardiogram Z score s for LV long-axis
dimension, LV short-axis dimension, aortic annulus diameter,
and mitral valve annulus diameter, and mitral regurgitation or
AS maximum systolic gradient) of four predicted BVR with
Table 3 No intervention versus
neonatal intervention
All values expressed as median (range), mean ± standard deviation, or n/denominator (percentage). Bolded
values represent statistical significance
LV left ventricle, MV mitral valve, PFO patent foramen ovale, RV right ventricle, TV tricuspid valve
†
Patients who died prior to intervention were included in the ‘required intervention’ group
Parameter No Intervention
(n = 7)
Required Intervention†
(n = 61) p
Cardiovascular profile score 8 ± 2 8 ± 1 0.42
8 (6–10) 8 (5–10)
Left-to-right shunting at PFO 2/7 (29) 41/59 (69) 0.04
Abnormal pulmonary vein Doppler 0/7 (0) 12/59 (20) 0.33
Mitral valve Z score − 1.35 ± 1.95 − 4.24 ± 3.96 0.06
− 0.94 (− 4.04–1.00) − 4.01 (− 12.06–3.41)
TV/MV ratio 1.44 ± 0.27 2.52 ± 1.51 0.07
1.42 (1.14–1.99) 2.08 (0.70–6.81)
LV length Z score − 1.76 ± 1.23 − 3.79 ± 3.16 0.004
− 2.00 (− 4.04–0.22) − 3.32 (− 11.62–2.03)
RV/LV length ratio 1.08 ± 0.12 1.77 ± 1.02 0.05
1.06 (0.92–1.28) 1.32 (0.79–6.73)
RV–LV length Z score discordance 1.39 ± 1.03 3.84 ± 3.33 0.10
1.33 (0.12–3.07) 2.49 (− 1.48–11.6)
LV ejection fraction 0.59 ± 0.14 0.36 ± 0.28 0.07
0.60 (0.34–0.76) 0.40 (0.00–0.78)
Aortic valve Z score − 1.30 ± 3.49 − 4.83 ± 3.49 0.01
− 0.84 (− 6.66–1.72) − 4.69 (− 12.56–2.38)
Prograde aortic flow 7/7 (100) 36/59 (61) 0.09
Ascending aorta Z score − 1.74 ± 2.25 − 4.61 ± 3.17 0.02
− 1.69 (− 5.27–0.39) − 4.61 (− 12.57–2.26)
Aorta/main pulmonary artery ratio 0.70 ± 0.20 0.44 ± 0.24 0.005
0.72 (0.41–0.98) 0.35 (0.11–1.26)
Aortic isthmus Z score − 1.60 ± 2.54 − 3.52 ± 2.10 0.07
− 0.97 (− 4.94–1.84) − 3.29 (− 7.55–0.59)
Retrograde flow in aortic arch 2/7 (29) 47/58 (81) 0.008
Abnormal genetics 2/4 (50) 10/54 (19) 0.19
8. Pediatric Cardiology
1 3
sensitivity 100%, negative predictive value 100%, specificity
55%, and positive predictive value 43% [11]. In a study of
patients with atrioventricular septal defect and double-outlet
right ventricle with borderline LVs, the presence of an apex-
forming LV was the only fetal echocardiographic parameter
predictive of BVR [31].
Our finding that RV/LV length ratio > 1.28 associates with
SVP with reasonable sensitivity and excellent specificity in
a broader population of fetuses with left-sided obstructive
lesions adds to this literature and may be a valuable tool for
fetal cardiologists to understand which fetuses will require
SVP versus BVR and counsel expectant families accordingly.
Limitations
Our study population was limited to mothers diagnosed
with left-sided obstructive lesions who ultimately deliv-
ered at our institution. Our results do not reflect fetuses
with false-negative diagnoses or those that died in utero,
were terminated, underwent comfort care as neonates, or
were delivered at outside institutions, possibly yielding
a selection bias. In addition, the fetal diagnostic group
was identified as per the diagnosis of the contemporary
fetal cardiologist; the diagnostic criteria used may have
Table 4 Biventricular versus
single-ventricle outcome
Patients who were not surgical candidates or died prior to surgery were excluded from analysis. All values
expressed as median (range), mean ± standard deviation, or n/denominator (percentage). Bolded values rep-
resent statistical significance
BVR biventricular repair, LV left ventricle, MV mitral valve; PFO, patent foramen ovale; RV, right ventri-
cle; TV, tricuspid valve; SVP, single-ventricle palliation
Parameter BVR (n = 25) SVP (n = 38) p
Cardiovascular profile score 8 ± 1 8 ± 1 0.34
8 (6–10) 8 (5–10)
Left-to-right shunting at PFO 5/25 (20) 2/37 (5) < 0.001
Abnormal Pulmonary vein Doppler 0/25 (0) 9/36 (25) 0.008
Mitral valve Z score − 1.30 ± 1.92 − 6.00 ± 3.56 < 0.001
− 1.18 (− 5.90–1.44) − 6.33 (− 12.06–1.14)
TV/MV ratio 1.39 ± 0.42 3.16 ± 1.48 < 0.001
1.27 (0.91–2.49) 2.84 (1.12–6.81)
LV length Z score − 1.49 ± 1.21 − 4.98 ± 3.13 < 0.001
− 1.36 (− 4.08–0.59) − 5.43 (− 11.62–2.03)
RV/LV length ratio 1.06 ± 0.19 2.10 ± 1.10 < 0.001
1.02 (0.79–1.75) 2.01 (0.97–6.73)
RV–LV length Z score discordance 0.94 ± 1.12 5.32 ± 2.93 < 0.001
0.95 (− 1.48–3.36) 5.97 (0.33–11.60)
LV ejection fraction 0.60 ± 0.14 0.22 ± 0.23 < 0.001
0.62 (0.15–0.78) 0.14 (0.00–0.67)
Aortic valve Z score − 1.82 ± 2.12 − 6.18 ± 2.92 < 0.001
− 1.40 (− 6.66–1.72) − 6.16 (− 12.56–0.97)
Prograde aortic flow 25/25 (100) 17/36 (47) < 0.001
Ascending aorta Z score − 2.42 ± 2.01 − 5.51 ± 3.23 < 0.001
− 2.09 (− 6.70–1.13) − 5.62 (− 12.57–2.26)
Aorta/main pulmonary artery ratio 0.59 ± 0.18 0.31 ± 0.25 < 0.001
0.59 (0.31–0.98) 0.34 (0.11–1.26)
Aortic isthmus Z score − 2.47 ± 2.27 − 3.74 ± 2.10 0.09
− 2.21 (− 7.55–1.84) − 3.87 (− 7.44–0.59)
Retrograde flow in aortic arch 11/24 (46) 33/36 (92) 0.0002
Abnormal genetics 7/19 (37) 5/39 (13) 0.04
9. Pediatric Cardiology
1 3
differed by fetal cardiologist and over time. Hence, there
was not a uniform set of echocardiographic criteria for
diagnostic subgroups. Finally, our results reflect a single
institution’s experience and may not be generalizable. Spe-
cifically, the decision to pursue SVP versus BVR is highly
variable among institutions, and our decision to commit
a neonate to a SVP or BVR may not align with other’s
decision-making processes.
Compliance with Ethical Standards
Conflict of interest The authors declare no conflicts of interest.
References
1. Friedberg MK, Silverman NH, Moon-Grady AJ, Tong E, Nourse
J, Sorenson B, Lee J, Hornberger LK (2009) Prenatal detection
of congenital heart disease. J Pediatr 155:26–31. https://doi.
org/10.1016/j.jpeds.2009.01.050
2. Quartermain MD, Pasquali SK, Hill KD, Goldberg DJ, Huhta JC,
Jacobs JP, Jacobs ML, Kim S, Ungerleider RM (2015) Variation in
prenatal diagnosis of congenital heart disease in infants. Pediatrics
136:e378–e385. https://doi.org/10.1542/peds.2014-3783
3. Freud LR, Moon-Grady A, Escobar-Diaz MC, Gotteiner NL,
Young LT, McElhinney DB, Tworetzky W (2015) Low rate of
prenatal diagnosis among neonates with critical aortic steno-
sis: insight into the natural history in utero. Ultrasound Obstet
Gynecol 45:326–332. https://doi.org/10.1002/uog.14667
4. Marek J, Tomek V, Skovránek J, Povysilová V, Samánek M
(2011) Prenatal ultrasound screening of congenital heart disease
in an unselected national population: a 21-year experience. Heart
97:124–130. https://doi.org/10.1136/hrt.2010.206623
5. Khoshnood B, De Vigan C, Vodovar V, Goujard J, Lhomme A,
Bonnet D, Goffinet F (2005) Trends in prenatal diagnosis, preg-
nancy termination, and perinatal mortality of newborns with
congenital heart disease in France, 1983–2000: a population-
based evaluation. Pediatrics 115:95–101. https://doi.org/10.1542/
peds.2004-0516
6. Chew C, Halliday JL, Riley MM, Penny DJ (2007) Population-
based study of antenatal detection of congenital heart disease by
ultrasound examination. Ultrasound Obstet Gynecol 29:619–624.
https://doi.org/10.1002/uog.4023
7. Khoo NS, Van Essen P, Richardson M, Robertson T (2008)
Effectiveness of prenatal diagnosis of congenital heart defects
in South Australia: a population analysis 1999–2003. Aust N Z
J Obstet Gynaecol 48:559–563. https://doi.org/10.1111/j.1479-
828X.2008.00915.x
8. Kipps AK, Feuille C, Azakie A, Hoffman JIE, Tabbutt S, Brook
MM, Moon-Grady AJ (2011) Prenatal diagnosis of hypoplastic
left heart syndrome in current era. Am J Cardiol 108:421–427.
https://doi.org/10.1016/j.amjcard.2011.03.065
9. Levy DJ, Pretorius DH, Rothman A, Gonzales M, Rao C, Nunes
ME, Bendelstein J, Mehalek K, Thomas A, Nehlsen C, Ehr J,
Burchette RJ, Sklansky MS (2013) Improved prenatal detec-
tion of congenital heart disease in an integrated health care sys-
tem. Pediatr Cardiol 34:670–679. https://doi.org/10.1007/s0024
6-012-0526-y
10. Gardiner HM, Kovacevic A, van der Heijden LB, Pfeiffer PW,
Franklin RC, Gibbs JL, Averiss IE, Larovere JM (2014) Pre-
natal screening for major congenital heart disease: assessing
performance by combining national cardiac audit with maternity
data. Heart 100:375–382. https://doi.org/10.1136/heartjnl-2013-
304640
11. Morris SA, Ethen MK, Penny DJ, Canfield MA, Minard CG,
Fixler DE, Nembhard WN (2014) Prenatal diagnosis, birth
location, surgical center, and neonatal mortality in infants with
hypoplastic left heart syndrome. Circulation 129:285–292. https
://doi.org/10.1161/CIRCULATIONAHA.113.003711
12. McElhinney DB, Marshall AC, Wilkins-Haug LE, Brown DW,
Benson CB, Silva V, Marx GR, Mizrahi-Arnaud A, Lock JE,
Tworetzky W (2009) Predictors of technical success and post-
natal biventricular outcome after in utero aortic VALVULO-
PLASTY for aortic stenosis with evolving hypoplastic left heart
syndrome. Circulation 120:1482–1490. https://doi.org/10.1161/
CIRCULATIONAHA.109.848994
13. Blaufox AD, Lai WW, Lopez L, Nguyen K, Griepp RB, Par-
ness IA (1998) Survival in neonatal biventricular repair of left-
sided cardiac obstructive lesions associated with hypoplastic
left ventricle. Am J Cardiol 82:1138–1140-A10. https://doi.
org/10.1016/S0002-9149(98)00576-1
14. Rhodes LA, Colan SD, Perry SB, Jonas RA, Sanders SP (1991)
Predictors of survival in neonates with critical aortic steno-
sis. Circulation 84:2325–2335. https://doi.org/10.1161/01.
CIR.84.6.2325
15. Hickey EJ, Caldarone CA, Blackstone EH, Lofland GK, Yeh T,
Pizarro C, Tchervenkov CI, Pigula F, Overman DM, Jacobs ML,
McCrindle BW, Congenital Heart Surgeons’ Society (2007) Criti-
cal left ventricular outflow tract obstruction: the disproportion-
ate impact of biventricular repair in borderline cases. J Thorac
Cardiovasc Surg 134:1429–1436. https://doi.org/10.1016/j.jtcvs
.2007.07.052
16. Tani LY, Minich L, Pagotto LT, Shaddy RE (1999) Left heart
hypoplasia and neonatal aortic arch obstruction: is the Rhodes left
ventricular adequacy score applicable? J Thorac Cardiovasc Surg
118:81–86. https://doi.org/10.1016/S0022-5223(99)70144-3
17. Rychik J, Ayres N, Cuneo B, Gotteiner N, Hornberger L, Spe-
vak PJ, Van Der Veld M (2004) American Society of Echocar-
diography guidelines and standards for performance of the fetal
echocardiogram. J Am Soc Echocardiogr 17:803–810. https://doi.
org/10.1016/j.echo.2004.04.011
18. Tan J, Silverman NH, Hoffman JI, Villegas M, Schmidt KG (1992)
Cardiac dimensions determined by cross-sectional echocardiogra-
phy in the normal human fetus from 18 weeks to term. Am J Car-
diol 70:1459–1467. https://doi.org/10.1016/0002-9149(92)90300
-N
19. Sharland GK, Allan LD (1992) Normal fetal cardiac measure-
ments derived by cross-sectional echocardiography. Ultra-
sound Obstet Gynecol 2:175–81. https://doi.org/10.104
6/j.1469-0705.1992.02030175.x
20. Roman KS, Fouron J-C, Nii M, Smallhorn JF, Chaturvedi R, Jae-
ggi ET (2007) Determinants of outcome in fetal pulmonary valve
stenosis or atresia with intact ventricular septum. Am J Cardiol
99:699–703. https://doi.org/10.1016/j.amjcard.2006.09.120
21. Salomon LJ, Alfirevic Z, Berghella V, Bilardo C, Hernandez-
Andrade E, Johnsen SL, Kalache K, Leung KY, Malinger G,
Munoz H, Prefumo F, Toi A, Lee W (2011) Practice guidelines
for performance of the routine mid-trimester fetal ultrasound scan.
Ultrasound Obstet Gynecol 37:116–126. https://doi.org/10.1002/
uog.8831
22. Hofstaetter C, Hansmann M, Eik-Nes SH, Huhta JC, Luther SL
(2006) A cardiovascular profile score in the surveillance of fetal
hydrops. J Matern Fetal Neonatal Med 19:407–413. https://doi.
org/10.1080/14767050600682446
23. Ebbing C, Rasmussen S, Kiserud T (2007) Middle cerebral artery
blood flow velocities and pulsatility index and the cerebroplacen-
tal pulsatility ratio: longitudinal reference ranges and terms for
10. Pediatric Cardiology
1 3
serial measurements. Ultrasound Obstet Gynecol 30:287–296.
https://doi.org/10.1002/uog.4088
24. Hadlock FP, Deter RL, Harrist RB, Park SK (1984) Estimating
fetal age: computer-assisted analysis of multiple fetal growth
parameters. Radiology 152:497–501. https://doi.org/10.1148/
radiology.152.2.6739822
25. Krishnan A, Pike JI, McCarter R, Fulgium AL, Wilson E, Dono-
frio MT, Sable CA (2016) Predictive models for normal fetal car-
diac structures. J Am Soc Echocardiogr 29:1197–1206. https://
doi.org/10.1016/j.echo.2016.08.019
26. Colan SD (2016) Normal echocardiographic values for cardio-
vascular structures. In: Lai WW, Mertens LL, Cohen MS, Geva T
(eds) Echocardiography in pediatric and congenital heart disease:
from fetus to adult. Wiley: Oxford, 2016; 883–901
27. Sluysmans T, Colan SD (2016) Structural measurements and
adjustments for growth. In: Lai WW, Mertens LL, Cohen MS,
Geva T (eds) Echocardiography in pediatric and congenital heart
disease: from fetus to adult. Wiley, Oxford, pp 61–72
28. Jowett V, Aparicio P, Santhakumaran S, Seale A, Jicinska H,
Gardiner HM (2012) Sonographic predictors of surgery in fetal
coarctation of the aorta. Ultrasound Obstet Gynecol 40:47–54.
https://doi.org/10.1002/uog.11161
29. Matsui H, Mellander M, Roughton M, Jicinska H, Gardiner HM
(2008) Morphological and physiological predictors of fetal aortic
coarctation. Circulation 118:1793–1801. https://doi.org/10.1161/
CIRCULATIONAHA.108.787598
30. Bolin E, Watrin C, Garuba O, Altman C, Ayres N, Morris S (2013)
Abstract 15552: Risk factors for postnatal surgery in the fetus with
borderline small left heart. Circulation 128:A15552–A15562
31. Pitkänen OM, Hornberger LK, Miner SES, Mondal T, Smallhorn
JF, Jaeggi E, Nield LE (2006) Borderline left ventricles in prena-
tally diagnosed atrioventricular septal defect or double outlet right
ventricle: echocardiographic predictors of biventricular repair. Am
Heart J 152:163.e1–7. https://doi.org/10.1016/j.ahj.2006.04.018
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
![Vol.:(0123456789)1 3
Pediatric Cardiology
https://doi.org/10.1007/s00246-019-02155-7
ORIGINAL ARTICLE
Fetal Echocardiographic Parameters and Surgical Outcomes
in Congenital Left‑Sided Cardiac Lesions
Lindsay A. Edwards1
· Alisa Arunamata1
· Shiraz A. Maskatia1
· Amy Quirin1
· Shazia Bhombal2
· Katsuhide Maeda3
·
Theresa A. Tacy1
· Rajesh Punn1
Received: 27 March 2019 / Accepted: 10 July 2019
© Springer Science+Business Media, LLC, part of Springer Nature 2019
Abstract
This study aimed to evaluate fetal echocardiographic parameters associated with neonatal intervention and single-ventricle
palliation (SVP) in fetuses with suspected left-sided cardiac lesions. Initial fetal echocardiograms (1/2002–1/2017) were
interpreted by the contemporary fetal cardiologist as coarctation of the aorta (COA), left heart hypoplasia (LHH), hypo-
plastic left heart syndrome (HLHS), mitral valve hypoplasia (MVH) ± stenosis, and aortic valve hypoplasia ± stenosis (AS).
The cohort comprised 68 fetuses with suspected left-sided cardiac lesions (COA n = 15, LHH n = 9, HLHS n = 39, MVH
n = 1, and AS n = 4). Smaller left ventricular (LV) length Z score, aortic valve Z score, ascending aorta Z score, and aorta/
pulmonary artery ratio; left-to-right shunting at the foramen ovale; and retrograde flow in the aortic arch were associated
with the need for neonatal intervention (p = 0.005–0.04). Smaller mitral valve (MV) Z score, LV length Z score, aortic valve
Z score, ascending aorta Z score, aorta/pulmonary artery ratio, and LV ejection fraction, as well as higher tricuspid valve-
to-MV (TV/MV) ratio, right ventricular-to-LV (RV/LV) length ratio, left-to-right shunting at the foramen ovale, abnormal
pulmonary vein Doppler, absence of prograde aortic flow, and retrograde flow in the aortic arch were associated with SVP
(p < 0.001–0.008). The strongest independent variable associated with SVP was RV/LV length ratio (stepwise logistical
regression, p = 0.03); an RV/LV length ratio > 1.28 was associated with SVP with a sensitivity of 76% and specificity of 96%
(AUC 0.90, p < 0.001). A fetal RV/LV length ratio of > 1.28 may be a useful threshold for identifying fetuses requiring SVP.
Keywords Left heart hypoplasia · Single-ventricle palliation · Fetal cardiology · Fetal echocardiography
Introduction
Congenital left-sided cardiac obstructive lesions encompass
a broad range of diagnoses, ranging from mild coarctation of
the aorta (COA) to hypoplastic left heart syndrome (HLHS).
While the prenatal diagnosis rate of HLHS is among the
highest for any form of congenital heart disease, prenatal
detection rates of other left-sided obstructive lesions includ-
ing COA and aortic valve stenosis (AS) remain poor [1–11].
The diagnostic challenge combined with the critical nature
of these conditions often leads to conservative management
on the part of the fetal cardiologist, with recommendations
for patient relocation for delivery at a tertiary care center
with a pediatric cardiology service. Moreover, for fetuses
with AS or left heart hypoplasia (LHH) who do not meet a
standard diagnosis of HLHS, the decision to pursue biven-
tricular repair (BVR) or single-ventricle palliation (SVP)
remains unclear [12–16].
In this study, we sought to determine fetal echocardio-
graphic predictors of need for neonatal intervention and
surgical outcome (BVR versus SVP).
* Lindsay A. Edwards
lindsayaedwards@gmail.com
1
Division of Cardiology, Department of Pediatrics , Stanford
University School of Medicine , 750 Welch Road, Suite 305,
Palo Alto, CA 94304, USA
2
Division of Neonatology, Department of Pediatrics , Stanford
University School of Medicine , Palo Alto, CA, USA
3
Department of Cardiothoracic Surgery , Stanford University
School of Medicine , Palo Alto, CA, USA](https://image.slidesharecdn.com/fetalechocardiographicparametersandsurgicaloutcomesincongenitalleftsidedcardiaclesions-191017161638/85/fetal-echocardiographic-parameters-and-surgical-outcomes-in-congenital-left-sided-cardiac-lesions-1-320.jpg)
![Pediatric Cardiology
1 3
Methods
Patient Selection
The Heart Center echocardiographic database was queried
for fetal studies performed at our institution between Janu-
ary 2002 and January 2017 with diagnostic codes includ-
ing the search terms mitral valve, aortic valve, stenosis,
hypoplasia, hypoplastic left heart syndrome, coarctation,
and arch. Initial fetal echocardiograms and corresponding
reports from studies interpreted by the contemporary fetal
cardiologist as COA, LHH, HLHS, mitral valve hypoplasia
(MVH) and/or stenosis, and aortic valve hypoplasia and/
or AS were included. The reported fetal diagnosis was
used for analysis. Fetuses with concern for small left-sided
structures and not committed to a diagnosis of COA or
HLHS by the diagnosing physician were coded as LHH.
Mitral valve hypoplasia was diagnosed in a fetus with a
mitral valve with Z score < − 2 but normal aortic annular
dimension and left ventricular size.
Corresponding neonatal clinical, echocardiographic,
and surgical records were reviewed independently to pre-
vent bias. Subjects were excluded if multiple fetal gesta-
tions were present; if double-inlet left ventricle, double-
outlet right ventricle, or atrioventricular septal defect was
suspected; if postnatal records including neonatal transtho-
racic echocardiogram were not available; or if the family
elected termination, comfort care, or delivery at another
institution. This project was approved by the Institutional
Review Board for Stanford University (Protocol Number
40371).
Echocardiographic Data
Standard fetal echocardiographic parameters were meas-
ured by a single investigator (LAE), blinded to postnatal
outcome at time of assessment. Cardiac measurements
included atrioventricular valve diameters, semilunar valve
diameters, main pulmonary artery diameter, ascending
aorta diameter, ductus arteriosus diameter, transverse
aortic arch diameter, aortic isthmus diameter, and ven-
tricular lengths [17–19]. A second fetal cardiologist (AA)
measured mitral valve annulus, aortic valve annulus, and
ascending aorta in a subset (n = 10) of randomly selected
patients for inter-observer variability analysis. RV and LV
length were measured in the apical four-chamber view at
end-diastole from the center of the atrioventricular valve
to the apical endocardium, and RV/LV length ratio was
calculated accordingly [20]. TV/MV ratio was calculated
by dividing the maximal diastolic annular dimension of
the tricuspid valve by that of the mitral valve [20]. Color
and spectral Doppler evaluation of the atrioventricu-
lar valves, semilunar valves, foramen ovale, pulmonary
vein, transverse aortic arch, umbilical artery, umbilical
vein, ductus venosus, and middle cerebral artery were
also recorded. Direction of flow at the foramen ovale was
classified as normal (right-to-left) or abnormal (bidirec-
tional or left-to-right). Pulmonary venous Doppler pro-
files were considered abnormal if prominent retrograde
flow during atrial systole was noted. Direction of flow in
the ascending aorta was classified as normal (prograde)
or abnormal (bidirectional or retrograde). Fetal biometry
measurements were made in accordance with practice
guidelines and included heart rate, biparietal diameter,
and cardiothoracic ratio [21]. Other calculations included
cardiovascular profile score [22], LV ejection fraction,
RV fractional area change, velocity time integrals across
semilunar valves, combined ventricular output, ventricular
stroke volume, and middle cerebral and umbilical artery
pulsatility indices. Measurements were converted to ges-
tational age-based Z scores (determined by estimated due
date from referring obstetrician) using available norma-
tive data [23–27]. A novel measurement, RV–LV length Z
score discordance, was calculated by subtracting the LV
length Z score from the RV length Z score [20]. Results
from invasive prenatal genetic testing (amniocentesis and
chorionic villus sampling) were recorded when available.
Clinical Data
Postnatal data included diagnoses, prostaglandin initiation,
neonatal intervention, hospital length of stay, survival to
discharge, mortality, and genetic testing results. Immediate
postnatal echocardiograms were retrospectively reviewed
after fetal echocardiogram assessment by a single reader
(LAE) to minimize outcome bias. Neonatal intervention
was defined as cardiac catheterization or surgery performed
in the first 30 days of life. For “intervention” versus “no
intervention” analysis, neonates who required intervention
but were not surgical candidates or died prior to interven-
tion were included in the “intervention” subgroup. For BVR
versus SVP analysis, neonates who required intervention
but died before intervention or were not surgical candidates
were not included in the analysis. Postnatal genetic testing
included fluorescent in situ hybridization, chromosomal
microarray, whole exome sequencing, and karyotype.
Statistical Analysis
All analyses were performed using SPSS version 24 (IBM;
Armonk, New York, USA) and SAS Enterprise Guide Ver-
sion 7.1 (SAS Institute Inc., Cary, NC, USA). Intraclass
correlation coefficients (ICC) were derived for fetal mitral
valve annulus, aortic valve annulus, and ascending aorta](data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7)