The Role of Doppler Waveforms in the Fetal

Main Pulmonary Artery in the Prediction

of Neonatal Respiratory Distress Syndrome

Yong Guan, MD,1 Shengli Li, MD,1 Guoyang Luo, MD, PhD,2 Chenghong Wang, MD, PhD,3
Errol R. Norwitz, MD, PhD,4 Qian Fu, MD, PhD,1 Xingzhi Tu, MD, PhD,3 Xiaoxian Tian, MD,5
Jun Zhu, MD6

1
Department of Ultrasound, Shenzhen Maternity & Child Healthcare Hospital, Affiliated to Southern Medical
University, Shenzhen, China
2
Department of Obstetrics and Gynecology, Danbury Hospital, Danbury, CT
3
Department of Obstetrics and Gynecology, Shenzhen Maternity & Child Healthcare Hospital, Affiliated to
Southern Medical University, Shenzhen, China
4
Department of Obstetrics and Gynecology, Tufts University School of Medicine, Boston, MA
5
Department of Ultrasound, Guangxi Maternity and Child Healthcare Hospital, Nanning, China
6
National Center for Birth Defects Monitoring, West China Second University Hospital, Sichuan University,
Chengdu, Sichuan, China

Received 26 April 2014; accepted 16 July 2014

ABSTRACT: Objective. To describe changes in the (32.6%) developed RDS and 29 did not. Using less
Doppler waveforms of the fetal main pulmonary than or equal to the fifth percentile as a gestational
artery (MPA) throughout gestation and to assess their age-specific cutoff, AT alone could predict RDS with
predictive value of neonatal respiratory distress syn- a sensitivity of 78.6% and a specificity of 89.7%.
drome (RDS). The AT/ET ratio could predict RDS with 71.4% sensi-
Study Design. In the first phase of this study, we tivity and 93.1% specificity.
performed Doppler measurement of MPA accelera- Conclusions. Fetal MPA Doppler velocimetry can
tion time (AT), ejection time (ET), peak systolic veloc- reliably be obtained throughout gestation. AT and AT/
ity, end-diastolic velocity, mean velocity, pulsatility ET ratios of the fetal MPA Doppler waveform may help
index, and resistance index in 288 healthy fetuses. In identifying fetuses at risk of developing neonatal RDS.
the second phase, we carried out these measure- C 2014 Wiley Periodicals, Inc. J Clin Ultrasound 00:000–
V
ments in a prospective cohort of 52 pregnant women 000, 2014; Published online in Wiley Online Library
with impending preterm birth. (wileyonlinelibrary.com). DOI: 10.1002/jcu.22219
Results. In phase I, satisfactory fetal MPA Doppler Keywords: ultrasound; pulmonary artery; Doppler
recordings were collected in 284 of 288 (98.6%) nor- velocimetry; neonatal respiratory distress syndrome
mal fetuses. Significant and positive linear correla-
tions were found between gestational age and AT,
AT/ET ratio, peak systolic velocity, and mean velocity
(p < 0.01), with the strongest correlations concerning INTRODUCTION
AT (r 5 0.898) and AT/ET ratio (r 5 0.868). In phase II,
satisfactory fetal MPA Doppler waveforms were
obtained in 43 of 44 (97.7%) fetuses. Of these, 14 T he pulmonary system is the last fetal organ
system required for extrauterine life to
become functionally mature, and respiratory
distress syndrome (RDS), related to pulmonary
Correspondence to: S. Li and J. Zhu surfactant deficiency, remains a major cause of
This study was supported by grants from the National Natu-
neonatal morbidity and mortality.1,2 Assessment
ral Science Foundation of China (No.81270707), the National
Youth Science Funds of China (No.61101026) and the Science of fetal lung maturity is therefore one of the
and Technology Program of Shenzhen, China (No.201203082). most important goals of obstetrical manage-
ment. Determination of fetal lung maturity
C 2014 Wiley Periodicals, Inc.
V

VOL. 00, NO. 00, MONTH 2014 1

 GUAN ET AL

traditionally has relied on amniocentesis and the women presenting for routine prenatal sono-
measurement of component proteins and lipid in graphic examination at Shenzhen Maternity
the amniotic fluid. However, amniocentesis is an and Child Healthcare Hospital from September
invasive procedure recommended for specific indi- 1, 2011 through December 31, 2012. Women
cations, and these tests have high sensitivity but were invited to participate in the study if they
only moderate specificity.3–5 For these reasons, had a singleton pregnancy at 22–42 weeks of
noninvasive methods using ultrasound to assess gestation. Accurate gestational age dating was
fetal lung maturity have long been sought for, but required (defined as dating by first-trimester
efforts to date (such as measurements of lung vol- ultrasound or by a certain last menstrual period
umes, gestation age, epiphysis centers, placental consistent with second-trimester ultrasound).
grading, and estimated fetal weight) have been Individual pregnant women could be recruited
unsuccessful in clinical practice.6–9 As the lungs into the study on only one occasion. Exclusion
develop throughout gestation, so does the pulmo- criteria included a known fetal chromosomal or
nary vasculature, where both the absolute num- structural abnormality, fetal growth less than
ber of pulmonary arteries rises and the total tenth or greater than the 90th percentile for ges-
amount of smooth muscular tissue increases, and tational age, any maternal pregnancy complica-
the pulmonary arterial vascular resistance tion, or prior antenatal corticosteroid treatment.
decreases slightly, leading to a gradual increase in Phase II was designed to determine whether
pulmonary blood flow.10,11 We hypothesize that fetal MPA Doppler velocimetry can be used to
noninvasive interrogation of the pulmonary vas- predict the subsequent development of neonatal
culature may provide valuable information about RDS. To this end, fetal MPA Doppler velocimetry
functional pulmonary maturity. was carried out in a separate cohort of pregnant
Doppler velocimetry provides a simple and women hospitalized at Shenzhen Maternity and
noninvasive method to assess the fetal pulmo- Child Healthcare Hospital between September 1,
nary circulation. Several investigators have 2011 and December 31, 2012 for impending pre-
used Doppler velocimetry to measure fetal pul- term birth (defined as any delivery <37 weeks).
monary blood flow in the left (or right) pulmo- This included patients with preterm labor, pre-
nary artery and their peripheral branches, but term premature rupture of membranes, hyper-
the rate of satisfactory Doppler recordings is tensive disorders, placenta previa, or fetal
moderate (77–84%), and the results are vastly growth restriction. Women could be recruited
disparate.12–14 In this study, we chose to mea- into the study on only one occasion. Exclusion
sure the Doppler waveform variables at the criteria included fetuses with a known chromo-
fetal main pulmonary artery (MPA). As preg- somal or structural abnormality, an interval from
nancy progresses, the Doppler velocity wave- Doppler examination to delivery >72 hours, or
form in the fetal MPA would change. The aim of gestational age at delivery 37 weeks. For logisti-
our study was twofold: (1) to describe changes cal reasons, antenatal corticosteroid treatment
in the Doppler waveforms of the fetal MPA was not an exclusion criterion for this cohort.
throughout gestation so as to establish gesta-
Examination Technique for Fetal MPA
tional age-specific reference ranges for each
Doppler Velocimetry
variable; and (2) to determine which Doppler
waveform variables, if any, can predict the sub- This study was performed using an Acuson
sequent development of neonatal RDS. Sequoia 512 ultrasound machine (Siemens Med-
ical Solutions, Mountain View, CA) equipped
with a 4–6-MHz convex transducer. Pregnant
women were scanned in the supine position and
PATIENTS AND METHODS
all fetuses were in sinus rhythm and in a quiet
state without fetal breathing movements. Dopp-
Study Population
ler velocity waveform measurements of the fetal
This study was approved by the Ethics Commit- MPA were performed by a single operator (Y.G.)
tee of Shenzhen Maternity and Child Health- after a routine prenatal sonographic examina-
care Hospital and, in all cases, written patient tion. Briefly, a systematic examination of the
consent was obtained. This study was carried fetal heart was first performed as previously
out in two phases. Phase I was a prospective described by Li et al15 to exclude any major
cross-sectional cohort study designed to describe structural defect. The fetal MPA was then
the normal changes in fetal MPA Doppler wave- visualized by rotating the transducer from the
form throughout gestation and included healthy four-chamber view to the short-axis view of fetal
2 JOURNAL OF CLINICAL ULTRASOUND
 DOPPLER WAVEFORMS PREDICT NEONATAL RDS

FIGURE 1. Blood flow velocity waveform in the fetal main pulmonary artery (MPA). (A) The short-axis view of the fetal heart is shown at the level
of the MPA stem and the bifurcation of the left and right pulmonary arteries (long arrow). The short arrow points to the pulmonary valves. The
asterisk indicates the sample point where the main pulmonary artery Doppler measurements are taken. Ao, aorta. (B) A representative real-time
blood flow velocity waveform in fetal main pulmonary artery. The acceleration time (AT) and ejection time (ET) can then be measured.

heart, thereby showing the pulmonary valves Neonatal Outcome

and the bifurcation of the right and left
The clinical outcome and data of all fetuses
branches of the pulmonary artery (Figure 1A).
enrolled in phase II of the study were
The pulsed Doppler sample gate was placed at
abstracted from the medical records, including
the middle of the fetal MPA (between the pul-
gestational age at delivery, route of delivery,
monary valves and the pulmonary artery bifur-
Apgar scores, birth weight, fetal gender, antena-
cation) and away from the arterial walls. After
tal steroid administration, neonatal intensive
enlarging the image as much as possible, the
care unit (NICU) admission, and presence or
sample gate was adjusted to 3 mm and the
absence of RDS. Neonatal RDS was diagnosed
angle of insonation (between the sound beam
by pediatricians who were blinded to the fetal
and the direction of blood flow) was maintained
MPA Doppler waveform measurements, accord-
at <20 . Doppler gain and scale were adjusted
ing to the infant’s clinical history, the presence
for optimal velocity waveform display clearly
or absence of respiratory compromise on clinical
showing the peak systolic velocity (PSV) and
examination, an increased oxygen requirement,
early diastolic notch (Figure 1B). The high-pass
response to surfactant therapy, and/or evidence
filter was set at 100 Hz to record diastolic blood
of hyaline membrane disease on chest radio-
flow. The blood flow waveform was then dis-
graph. In all cases, the diagnosis was confirmed
played with a velocity range of 100 cm/s and a
by medical record review.
sweep speed of 200 mm/s. The shortest time
interval that could be measured was 1 ms.
Statistical Analysis
The Doppler velocity waveform in the fetal
MPA produced a specific “spike and dome” pat- Statistical analysis was performed using SPSS
tern.14 After the optimal fetal MPA waveform for Windows version 18.0 (SPSS, Inc., Chicago,
was obtained, relevant Doppler velocity varia- IL). In phase I, fetuses were grouped into ten 2-
bles were measured three times using manual- week intervals (22–24, 24–26, 26–28, 28–30, 30–
trace, and measurements were averaged. The 32, 32–34, 34–36, 36–38, 38–40, and 40–42
Doppler variables included acceleration time weeks). The data were expressed as mean 6 SD.
(AT; ie, the time interval from the foot to the Regression analysis and Spearman’s correlation
top of PSV), ejection time (ET; ie, from the coefficient calculations were used to investigate
beginning to the end of ventricular systole), the relationship between the individual fetal
PSV (the maximum blood flow velocity reached MPA Doppler waveform variables and gesta-
during systole), end-diastolic maximum velocity tional age. The normal reference range was
(EDV), mean velocity (MV; ie, the time- deducted from 90% confidence intervals (CI) for
averaged maximum velocity), pulsatility index each MPA Doppler waveform variable. Intraob-
(PI 5 [PSV 2 EDV]/MV), and resistance index server reproducibility observations were eval-
(RI 5 [PSV 2 EDV]/PSV) (Figure 1B). From uated by repeating fetal MPA Doppler
these measurements, the AT/ET ratio was measurements three times every 5 minutes in
calculated. 10 fetuses by a single operator (Y.G.) and were
VOL. 00, NO. 00, MONTH 2014 3
 GUAN ET AL

90.26 612.09
8% for AT, 6.8% for ET, 8.4% for AT/ET, 12% for

47.70 6 2.58
175.20 6 7.29
0.27 6 0.02

7.61 6 1.70
2.52 6 0.19
0.91 6 0.02
32.32 6 5.27
(n 5 10)
40–42
PSV, 16% for EDV, 10.2% for PI, 4.4% for RI,
and 13% for MV.
In phase II, the demographic, clinical, and
sonographic characteristics between neonates
that did and did not develop RDS were com-

47.69 6 2.87
177.07 6 8.20
0.27 6 0.02
88.95 6 9.31
7.45 6 1.66
2.58 6 0.23
0.91 6 0.02
31.41 6 3.86
(n 5 44) pared by independent sample t test. p < 0.05
38–40
Changes in Doppler Velocity Variables of the Fetal Main Pulmonary Artery Waveform with Gestational Age in Normal Singleton Fetuses

was considered statistically significant, and all

tests were two-tailed. Logistic regression analy-

Abbreviations: AT, acceleration time; ET, ejection time; EDV, end diastolic velocity; MV, mean velocity; PI, pulsatility index; PSV, peak systolic velocity; RI, resistancee index.
sis was used to explore the relationship between
84.61 611.07 Doppler waveform variables and the develop-
45.22 6 2.80
177.40 6 9.13
0.26 6 0.02

7.79 6 2.35
2.53 6 0.20
0.91 6 0.02
30.85 6 4.23
ment of RDS after controlling for gestational
(n 5 32)
36–38

age and corticosteroid administration.

Thereafter, the MPA sonographic characteris-
tics of the fetuses with and without RDS in
phase II were compared with the normal refer-
43.85 6 2.46
178.45 6 6.80
0.24 6 0.01
80.8 6 8.87
7.49 6 2.94
2.61 6 0.21
0.90 6 0.03
29.10 6 3.55

ence range (90% CI) generated in phase I after

(n 5 28)
34–36

controlling for gestational age. p < 0.05 was con-

sidered statistically significant. The sensitivity
and specificity of each MPA Doppler waveform
variable to predict the subsequent development
of neonatal RDS were then calculated using less
81.00 610.04
7.30 6 2.61
41.00 6 2.79
179.00 6 7.76
0.23 6 0.02

2.61 6 0.28
0.91 6 0.03
28.40 6 3.87
Gestational Age (weeks)

(n 5 27)

than or equal to the fifth percentile for each

32–34

Doppler variable as cutoff.

TABLE 1

38.00 6 3.12
178.00 6 7.70
0.22 6 0.02
77.40 6 8.22
7.10 6 2.14
2.59 6 0.33
0.91 6 0.03
27.50 6 3.45

RESULTS
(n 5 46)
30–32

Reference Ranges for Fetal MPA Doppler

Waveform Parameters Throughout
Gestation
7.30 6 2.01
36.00 6 2.46
178.23 6 7.10
0.20 6 0.02
79.41 6 8.68

2.66 6 0.19
0.91 6 0.02
27.50 6 3.32

A total of 288 women were included in phase I

(n 5 27)
28–30

of the study. Adequate fetal MPA Doppler wave-

forms were obtained in 284 (98.6%) cases; four
subjects were excluded from the final analysis
due to an inability to get a satisfactory MPA
32.78 6 2.04
177.52 6 6.37
0.19 6 0.01
75.30 6 7.31
7.22 6 2.17
2.65 6 0.29
0.91 6 0.02

Doppler waveform. In all four cases (gestational

26.26 6 3.9
(n 5 24)
26–28

ages 30, 34, 37, and 38 weeks), the fetus was

orientated with its spine toward the maternal
spine. Of the 284 fetuses included in the final
analysis, 22 were imaged at 22–24 weeks, 24 at
24–26 weeks, 24 at 26–28 weeks, 27 at 28–30
6.90 6 1.62
30.70 6 1.78
180.56 6 5.80
0.17 6 0.01
74.86 6 8.24

2.73 6 0.26
0.92 6 0.02
25.90 6 4.10
(n 5 24)
24–26

weeks, 46 at 30–32 weeks, 27 at 32–34 weeks,

28 at 34–36 weeks, 32 at 36–38 weeks, 44 at
38–40 weeks, and 10 at 40–42 weeks. The mean
Data are presented as mean 6 SD.

maternal age was 29 6 4.4 years. MPA Doppler

variables at each 2-week gestational age inter-
7.33 6 1.49
27.45 6 1.34
180.15 6 4.93
0.15 6 0.01
73.63 6 6.65

2.60 6 0.20
0.89 6 0.03
24.70 6 3.08
(n 5 22)

vals are presented in Table 1 and Figure 2.

22–24

Spearman’s correlation coefficients and regres-

sion analysis are shown in Table 2. AT
(r 5 0.898, p < 0.01) and AT/ET ratio (r 5 0.868,
p < 0.01) increased significantly and dramati-
Parameter

AT/ET ratio

EDV (cm/s)
PSV (cm/s)

MV (cm/s)

cally with increasing gestational age; PSV, EDV,

Doppler

AT (ms)
ET (ms)

and MV also showed a significant but less

marked increase with increasing gestational age
RI
PI

4 JOURNAL OF CLINICAL ULTRASOUND

 DOPPLER WAVEFORMS PREDICT NEONATAL RDS

FIGURE 2. Association between blood flow velocity waveform parameters in the fetal main pulmonary artery and gestational age. Individual val-
ues and reference range (90% CI) of Doppler variables in the fetal main pulmonary artery are plotted against gestational age (GA) in weeks (w).
The Doppler variables shown include acceleration time (AT) (A), ejection time (ET) (B), AT/ET ratio (C), peak systolic velocity (PSV) (D), end-
diastolic velocity (EDV) (E), resistance index (RI) (F), pulsatility index (PI) (G), and time-averaged mean velocity (MV) (H). In each graph, the open
blue circles (reference group) represent normal singleton fetuses, which were used to establish the 90% CI (phase I). The red triangles (respiratory
distress syndrome [RDS] group) and yellow squares (no RDS group) represent the study subjects in the prospective cohort (phase II) who did and
did not develop neonatal respiratory distress syndrome, respectively.

(r 5 0.446, 0.203, 0.507; p < 0.01 for each). PI tion (n 5 6). The mean gestational age at the
was inversely linearly correlated with gesta- time of sonographic examination was 31.9 6 1.8
tional age (r 5 20.184, p < 0.01), while ET and weeks. Technically acceptable Doppler wave-
RI did not change significantly throughout ges- forms were obtained in 43 of 44 eligible fetuses
tation (r 5 20.103, 0.011; p > 0.05). (97.7%). Of these, 14 (32.6%) developed RDS
and 29 (67.4%) did not. A comparison of the
maternal and neonatal demographic, clinical,
Use of Fetal MPA Doppler Velocimetry
and ultrasound data between fetuses that did
Waveforms to Predict Neonatal RDS
and those that did not develop RDS is shown in
In phase II, 52 fetuses at high risk of develop- Table 3. Fetuses that developed RDS were born
ing neonatal RDS by virtue of preterm birth at an earlier gestational age (29.50 6 1.54 ver-
<37 weeks were scanned, and fetal MPA Dopp- sus 34.47 6 1.63 weeks; p < 0.001) with a lower
ler velocimetry waveforms were analyzed as birth weight (1,474 6 369 versus 2,501 6 442 g;
described. Of these, eight fetuses were excluded p < 0.001), were more likely to have received
because they delivered after 37 weeks (n 5 2) or antenatal corticosteroids (100% versus 34.5%;
more than 72 hours after sonographic examina- p < 0.001), more likely to have a 5-minute Apgar
VOL. 00, NO. 00, MONTH 2014 5
 GUAN ET AL

FIGURE 2. (Continued)

TABLE 2
Regression Equations for Doppler Velocimetry Variables of the Fetal Main Pulmonary Artery Waveform with Gestational Age

Doppler Parameters Regression Equations* R2 F p Value

Acceleration time (AT) y 5 1.1871x 1 1.6597 0.81 1,085.82 <0.01

AT/ET ratio y 5 0.0069x 1 0.0018 0.78 785.47 <0.01
Peak systolic velocity (PSV) y 5 1.0389x 1 47.108 0.19 68.72 <0.01
End diastolic velocity (EDV) y 5 0.069x 1 4.647 0.03 7.04 0.02
Pulsatility index (PI) y 5 20.0122x 1 2.999 0.05 10.08 <0.01
Mean velocity (MV) y 5 0.4888x 1 12.882 0.23 75.63 <0.01

*In the regression equations: x 5 gestational age; y 5 fetal main pulmonary artery Doppler parameter. Ejection time (ET) and resistance index
had no significant correlation with gestational age.

score <7 (28.6% versus 0%; p < 0.001), and Fetuses from phase II that developed RDS
more likely to be admitted to the NICU (100% (n 5 14, range 27.14–32.14 weeks) demonstrated
versus 58.6%; p < 0.001) than fetuses that did different ranges of AT, AT/ET ratio, and PSV
not develop RDS (Table 3). There were signifi- than those of the reference population (phase I)
cant differences in AT, AT/ET ratio, PSV, and of the same gestational age (n 5 93, range
MV (p < 0.001 for each) between fetuses that 26.00–32.86 weeks) (Table 4), whereas there was
did and those that did not develop RDS, no significant difference in MPA sonographic var-
whereas the other Doppler waveform variables iables between fetuses from phase II that did not
were not different between the two groups develop RDS (n 5 29, range 30.57–36.57 weeks)
(Table 3). However, after controlling for gesta- and a gestational age-matched control cohort
tional age and administration of corticoste- from phase I (n 5 99, range 30.00–37.14 weeks).
roids, none of the Doppler waveform variables AT was below the phase I fifth percentile
had statistical differences between the two used as a gestational age-specific cutoff (ie, less
groups (p > 0.05 for each, range 0.227–0.850, than the lower limit of the 90% CI) in 11 of 14
lowest: PSV, highest: AT). fetuses (78.6%) that developed RDS, and in only
6 JOURNAL OF CLINICAL ULTRASOUND
 DOPPLER WAVEFORMS PREDICT NEONATAL RDS

TABLE 3
Demographic, Clinical, and Sonographic Characteristics of the Study Population Samples with and Without Neonatal Respira-
tory Distress Syndrome (RDS)

Characteristic RDS (n 5 14) No RDS (n 5 29) p Value

Maternal age (years) 27.21 6 7.19 29.97 6 5.02 0.150

Parity 0.010
0 8 (57.1%) 15 (51.7%)
1 0 10 (34.5%)
2 6 (42.9%) 4 (13.8%)
Indications for hospitalization 0.714
Hypertensive disorders 6 (42.9%) 8 (27.6%)
Preterm premature rupture of membranes 5 (35.8%) 9 (31.0%)
Placenta previa 1 (7.1%) 7 (24.1%)
Fetal growth restriction 1 (7.1%) 3 (10.3%)
Preterm labor 1 (7.1%) 2 (6.9%)
Gestational age at sonographic examination (weeks) 29.31 6 1.43 34.22 6 1.59 <0.001
Gestational age at delivery (weeks) 29.50 6 1.54 34.47 6 1.63 <0.001
Antenatal corticosteroid administration 14 (100%) 10 (34.5%) <0.001
Cesarean delivery 12 (85.7%) 22 (75.9%) >0.99
Gender (male) 7 (50%) 13 (44.8%) >0.99
Birth weight (g) 1,474 6 369 2,501 6 442 <0.001
5-minute Apgar score <7 4 (28.6%) 0 <0.001
NICU admission 14 (100%) 17 (58.6%) <0.001
Fetal MPA Doppler parameters
Acceleration time (AT) (ms) 31.23 6 3.09 42.46 6 2.63 <0.001
Ejection time (ET) (ms) 179.92 6 5.88 175.96 6 6.04 0.118
Acceleration time/ejection time ratio 0.17 6 0.02 0.24 6 0.01 <0.001
Peak systolic velocity (PSV) (cm/s) 69.91 6 14.87 85.68 6 9.01 <0.001
End-diastolic velocity (EDV) (cm/s) 6.19 6 2.22 7.36 6 2.25 0.131
Pulsatility index (PI) 2.64 6 0.26 2.60 6 0.19 0.567
Resistive index (RI) 0.92 6 0.03 0.92 6 0.02 0.735
Mean velocity (MV) (cm/s) 24.66 6 5.45 30.22 6 3.64 <0.001

Data are presented as mean 6 SD or n (%).

Abbreviations: MPA, main pulmonary artery; NICU, neonatal intensive care unit.

TABLE 4
Sonographic Characteristics of the Fetuses with and Without Neonatal Respiratory Distress Syndrome (RDS) Compared with
the Reference Population of the Same Gestational Age

Reference No RDS Reference

Characteristics RDS (n 5 14) (n 5 93) p Value (n 5 29) (n 5 99) p Value

Maternal age (years) 27.21 6 7.19 28.92 6 4.05 0.194 29.97 6 5.02 29.98 6 4.57 0.989
Gestational age at sonographic 29.31 6 1.43 29.39 6 2.01 0.836 34.22 6 1.59 34.29 6 2.16 0.685
examination (weeks)
Fetal MPA Doppler parameters
Acceleration time (AT) (ms) 31.23 6 3.09 36.23 6 3.98 <0.001 42.46 6 2.63 42.56 6 3.75 0.996
Ejection time (ET) (ms) 179.92 6 5.88 178.45 6 7.71 0.352 175.96 6 6.04 177.62 6 8.06 0.429
Acceleration time/ejection time ratio 0.17 6 0.02 0.20 6 0.02 <0.001 0.24 6 0.01 0.24 6 0.02 0.787
Peak systolic velocity (PSV) (cm/s) 69.91 6 14.87 76.92 6 9.16 0.023 85.68 6 9.01 81.41 6 10.77 0.080
End diastolic velocity (EDV) (cm/s) 6.19 6 2.22 6.84 6 2.29 0.326 7.36 6 2.25 7.48 6 2.58 0.730
Pulsatility index (PI) 2.64 6 0.26 2.63 6 0.31 0.931 2.60 6 0.19 2.53 6 0.23 0.135
Resistive index (RI) 0.92 6 0.03 0.91 6 0.03 0.261 0.92 6 0.02 0.91 6 0.03 0.124
Mean velocity (MV) (cm/s) 24.66 6 5.45 26.98 6 3.85 0.055 30.22 6 3.64 29.53 6 4.09 0.406

Data are presented as mean 6 SD. Reference ranges were calculated from the 90% CI obtained in phase I, and compared with the RDS group and
no RDS group of the similar range of gestational age.

3 of 29 (10.3%) that did not (Figure 2A) (sensi- for PSV (Figure 2D) (sensitivity 35.7%, specific-
tivity 78.6%, specificity 89.7%). These figures ity 100%); and 3 of 14 fetuses (21.4%) versus 1
were 10 of 14 fetuses (71.4%) versus 2 of 29 of 29 fetuses (3.4%) for MV (Figure 2H) (sensi-
fetuses (6.9%) for AT/ET ratio (Figure 2C) (sen- tivity 21.4%, specificity 96.6%), whereas ET,
sitivity 71.4%, specificity 93.1%); 5 of 14 EDV, PI, and RI showed no relation with
fetuses (35.7%) versus in 0 of 29 fetuses (0%) neonatal RDS.
VOL. 00, NO. 00, MONTH 2014 7
 GUAN ET AL

DISCUSSION age was ET, probably because pulmonary artery

In the last 10 years, Doppler velocimetry has vasoconstriction stabilizes the resistance in
these vessels throughout gestation.
been increasingly used to interrogate the
One of the major findings of this study was
maternal-fetal circulation, including the fetal
the significant and dramatic increase in AT and
pulmonary circulation. Fetal oxygen exchange
AT/ET ratio in the fetal MPA with increasing
occurs almost exclusively in the placenta and
gestational age (Table 1 and Figure 2A and 2C).
not in the lungs. For this reason, the pulmonary
This suggests that pulmonary vascular compli-
circulation in the fetus is maintained in a state
ance increases and mean pulmonary arterial
of high resistance, high pressure, and low
pressure decreases as pregnancy advances,
flow.16 Doppler ultrasonography provides a sim-
resulting in a gradual increase in pulmonary
ple, noninvasive, and reproducible way of blood flow, which is consistent with results of
directly evaluating the status of the fetal pul- Chaoui et al.14 Given that AT and AT/ET ratio
monary circulation. In this study, we specifically are relatively easily and reproducibly measura-
sampled blood flow in the center of fetal MPA, ble, and are strongly correlated with gestational
because it was relatively easy to identify and to age, we think that they could be used as reli-
record the pulmonary artery waveform. The able indices to evaluate fetal pulmonary matu-
MPA waveform could best be seen using the ration. Azpurua et al17 found that the AT/ET
right ventricular outflow tract view and/or the ratio correlated negatively with the amniotic
three-vessel view. Overall, in this study (includ- fluid lecithin/sphingomyelin ratio, but they did
ing both phases I and II), the MPA waveform not investigate the association between the AT/
was adequately visualized in 327 of 332 single- ET ratio and clinical RDS because they had too
ton fetuses (98.5%), which is higher than previ- few cases of RDS (only one infant). These Dopp-
ously reported in either the pulmonary artery ler indices may also help identifying fetuses at
branches or the intrapulmonary arteries.13,14 As risk of pulmonary immaturity.18 Fuke et al19
pregnancy progresses, the vascular beds of the showed that this same fetal MPA AT/ET ratio
fetal lung expand, leading to a progressive may predict pulmonary hypoplasia.
decrease in pulmonary arterial vascular resist- In phase II of this study, the AT, AT/ET ratio,
ance and a slow increase in pulmonary blood PSV, and MV (but not ET, EDV, PI, or RI) were
flow. It is therefore not surprising that the significantly lower in the RDS group than in
velocity waveform and Doppler variables would the non-RDS group (Table 3). Unexpectedly, all
change to some extent with increasing gesta- these variables showed no significant differen-
tional age. ces between the RDS and non-RDS group after
Phase I of this study investigated the effect of controlling for gestational age and antenatal
increasing gestational age on the Doppler wave- administration of corticosteroids. This might be
form in the MPA. Of the eight Doppler variables due to the quite small number of cases and
analyzed, five increased significantly with almost no overlap of gestational age between
increasing gestational age (most dramatically the two groups. On the other hand, the AT, AT/
AT and AT/ET ratio, but also PSV, EDV, and ET ratio, and PSV were significantly lower in
MV); PI decreased with increasing gestational the RDS group than in the normal reference
age, and ET and RI did not change significantly group from phase I after controlling for gesta-
throughout gestation (Table 1 and Figure 2). tional age (Table 4). Taken together, these data
The effects of gestational age on MPA PI and RI suggest that fetuses that go on to develop RDS
are consistent with prior publications by Rasa- have higher pulmonary vascular resistance and
nen et al12 and Chaoui et al.14 The less impres- pressure, and lower pulmonary blood flow than
sive (but still statistically significant) increase those that do not. A fetal MPA AT/ET ratio
in PSV and MV with increasing gestational age prior to delivery could predict the subsequent
likely reflects the gradual increase in pulmo- development of neonatal RDS with a sensitivity
nary blood flow velocity in late pregnancy. The of 71.4% and a specificity of 93.1%, while AT
absolute EDV in the fetal pulmonary circulation could predict the development of RDS with a
is very low, so that its accurate measurement is sensitivity of 78.6% and a specificity of 89.7%.
difficult and subject to significant variation. As mentioned previously, all amniocentesis tests
This is probably why EDV measurements for fetal lung maturity have high sensitivity
showed the highest intraobserver coefficient of and relatively moderate specificity. In compari-
variation (16%) in this study. The other Doppler son, AT and the AT/ET ratio had both high
variable that was not affected by gestational sensitivity and specificity to assess fetal lung
8 JOURNAL OF CLINICAL ULTRASOUND
 DOPPLER WAVEFORMS PREDICT NEONATAL RDS

maturity compared with amniocentesis test, 6. Osada H, Iitsuka Y, Masuda K, et al. Application
which is invasive. However, there were too few of lung volume measurement by three-
cases of RDS in our study to fully examine the dimensional ultrasonography for clinical assess-
relationship between fetal MPA Doppler wave- ment of fetal lung development. J Ultrasound
Med 2002;21:841.
forms and RDS at each 2-week gestational age
7. Deter RL, Hadlock FP, Carollo BR. Ultrasonically
interval, and longitudinal assessment of MPA determined menstrual age as an indicator of fetal
Doppler measurements within the same fetuses lung maturity. J Clin Ultrasound 1982;10:153.
should be further investigated. 8. Tabsh KM. Correlation of ultrasonic epiphyseal
In summary, fetal MPA Doppler waveforms centers and the lecithin:sphingomyelin ratio.
can be measured throughout gestation and can Obstet Gynecol 1984;64:92.
be used as a safe and reproducible technique for 9. Harman CR, Manning FA, Stearns E. The correla-
the assessment of the fetal pulmonary circula- tion of ultrasonic placental grading and fetal pul-
tion. We established reference ranges and mean monary maturation in five hundred sixty-three
values for a series of Doppler velocity variables pregnancies. Am J Obstet Gynecol 1982;143:941.
in the fetal MPA from 22 through 42 weeks of 10. Laudy JA, Wladimiroff JW. The fetal lung. 1:
Developmental aspects. Ultrasound Obstet Gyne-
gestation. We have further shown that selected
col 2000;16:284.
variables of the fetal MPA Doppler waveform--- 11. Rasanen J, Wood DC, Weiner S, et al. Role of the
specifically, AT and AT/ET ratios less than or pulmonary circulation in the distribution of human
equal to the fifth percentile for gestational fetal cardiac output during the second half of preg-
age---may provide a reliable noninvasive test to nancy. Circulation 1996;94:1068.
identify fetuses at risk of developing neonatal 12. Rasanen J, Hubta JC, Weiner S, et al. Fetal branch
RDS. Before this test can be introduced into pulmonary arterial vascular impedance during the
clinical practice, additional large prospective second half of pregnancy. Am J Obstet Gynecol
clinical trials are needed to better examine the 1996;174:1441.
relationship between fetal MPA Doppler wave- 13. Laudy JA, de Ridder MA, Wladimiroff JW. Human
forms and neonatal RDS, to investigate how fetal pulmonary artery velocimetry: repeatability
and normal values with emphasis on middle and
this relationship is affected by gestational age,
distal pulmonary vessels. Ultrasound Obstet
and to determine whether this is true also of Gynecol 2000;15:479.
fetuses with threatening delivery in the late 14. Chaoui R, Taddei F, Rizzo G, et al. Doppler echo-
preterm (34–37 weeks) and early term period cardiography of the main stems of the pulmonary
(37–39 weeks). arteries in the normal human fetus. Ultrasound
Obstet Gynecol 1998;11:173.
15. Li S, Luo G, Norwitz ER, et al. Prenatal diagnosis
of congenital vascular rings and slings: sono-
REFERENCES
graphic features and perinatal outcome in 81 con-
1. Kamath BD, Macguire ER, McClure EM, et al. secutive cases. Prenat Diagn 2011;33:334.
Neonatal mortality from respiratory distress syn- 16. Copel JA, Reed KL. Doppler Ultrasound in Obstet-
drome: lessons for low-resource countries. Pedia- rics and Gynecology. New York: Raven Press, Ltd;
trics 2011;127:1139. 1995, p 307.
2. Angus DC, Linde-Zwirble WT, Clermont G, et al. 17. Azpurua H, Norwitz ER, Campbell KH, et al.
Epidemiology of neonatal respiratory failure in the Acceleration/ejection time ratio in the fetal pulmo-
United States: projections from California and nary artery predicts fetal lung maturity. Am J
New York. Am J Respir Crit Care Med 2001;164: Obstet Gynecol 2010;203:e1.
1154. 18. Kim SM, Park JS, Norwitz ER, et al. Acceleration
3. Spong CY, Mercer BM, D’alton M, et al. Timing of time-to-ejection time ratio in fetal pulmonary
indicated late-preterm and early-term birth. artery predicts the development of neonatal respi-
Obstet Gynecol 2011;118:323. ratory distress syndrome: a prospective cohort
4. Ventolini G, Neiger R, Hood D, et al. Update on study. Am J Perinatol 2013;30:805.
assessment of fetal lung maturity. J Obstet Gynae- 19. Fuke S, Kanzaki T, Mu J, et al. Antenatal predic-
col 2005;25:535. tion of pulmonary hypoplasia by acceleration time/
5. Luo G, Norwitz ER. Revisiting amniocentesis for ejection time ratio of fetal pulmonary arteries by
fetal lung maturity after 36 weeks’ gestation. Rev Doppler blood flow velocimetry. Am J Obstet Gyne-
Obstet Gynecol 2008;1:61. col 2003;188:228.

VOL. 00, NO. 00, MONTH 2014 9