Prediction of Abortion Using Three-
Dimensional Ultrasound Volumetry of the
Gestational Sac and the Amniotic Sac in
Threatened Abortion
Marwan Odeh, MD,1,2* Ella Ophir, MD,1,2 Vitaly Grinin, MD,1 Rene Tendler, MD,1
Mohamad Kais,1 Jacob Bornstein, MD1,2
1
Department of Obstetrics and Gynecology, Western Galilee Hospital, Nahariya, Israel
2
Bar Ilan University—Galilee Faculty of Medicine, Zefat, Israel
Received 28 May 2011; accepted 7 May 2012
ABSTRACT: Purpose. To determine whether gesta-
tional sac volume (GSV) or amniotic sac volume
(ASV) and/or the difference between them can predict
abortion in women with first-trimester threatened
abortion.
Methods. Ninety patients between 6 and 12 weeks
of gestation presenting with vaginal bleeding were
studied. Seventy-six delivered after 24 weeks of ges-
tation (group A) and 14 aborted before 20 weeks of
gestation (group B). All patients had a singleton via-
ble pregnancy demonstrated by transvaginal ultra-
sound. Gestational sac and amniotic sac volumes
were measured in all the patients using three-dimen-
sional transvaginal ultrasound with Virtual Organ
Computer-aided Analysis software, and the gesta-
tional sac volume 2 amniotic sac volume (GSV 2
ASV) was calculated.
Results. The groups did not differ in terms of age,
parity, number of previous abortions, or term deliv-
eries. The GSV (group A: mean 32.0 6 27.7 cm3;
group B: 26.7 6 29.1 cm3) and the ASV (group A:
21.1 6 25.5 cm3; group B: 20.6 6 26.0 cm3) were not
statistically different, while the GSV 2 ASV was sig-
nificantly smaller in group B (aborting before week
20) (group A: 10.9 6 10.9 cm3; group B: 6.1 6 8.6
cm3; p < 0.05). Using receiver operator curves, the
area under the curve for predicting normal preg-
nancy outcome of the GSV 2 ASV measurement was
0.654. When the GSV 2 ASV was 1.8 cm3 or less,
abortion was predicted with 84% sensitivity and 43%
specificity.
Conclusions. The measurement of the GSV and the
ASV are not good predictors of abortion in patients
Correspondence to: M. Odeh
' 2012 Wiley Periodicals, Inc.
VOL. 00, NO. 0, MONTH 2012
with first-trimester vaginal bleeding, whereas the use
of the GSV 2 ASV may be helpful in predicting the
outcome of pregnancy. V C 2012 Wiley Periodicals, Inc.
J Clin Ultrasound 00:000–000, 2012; Published online
in Wiley Online Library (wileyonlinelibrary.com). DOI:
10.1002/jcu.21957
Keywords: gestational sac volume; amniotic sac
volume; threatened abortion; VOCAL
iscarriage occurs in about 10–15% of clini-
M cally recognized conceptions.1 Several sono-
graphic (US) parameters are used to predict a
normal pregnancy outcome in the first trimester.
These include the size and location of the gesta-
tional sac,2 the growth rate of the gestational
sac,3 the gestational sac size in relation to the
presence or absence of fetal pole and yolk sac,4–7
the size and shape of the yolk sac,8,9 the fetal
heart rate,10,11 and the relationship between the
gestational sac mean size and the crown rump
length (CRL).12–15
Threatened abortion is a risk factor for poor
pregnancy outcome,16 including spontaneous
pregnancy loss before 24 weeks of gestation, pree-
clampsia, preterm delivery, placental abruption,
cesarean delivery, and intrauterine growth
restriction.16
We have reported that gestational sac volume
(GSV) is smaller in cases of missed abortion
beginning as early as 7 weeks of gestation.17
However, to the best of our knowledge, three-
dimensional (3D) measurements of the GSV or
the amniotic sac volume (ASV) have not been uti-
1
 ODEH ET AL
lized in the prediction of abortion in the first tri-
mester.
The purpose of this study was to test the hy-
pothesis that in women with threatened abortion,
pregnancies with a small GSV or small ASV, as
measured by 3D US, are at greater risk of abor-
tion.
MATERIALS AND METHODS
Over a 12-month period, women who presented
to the emergency ward with vaginal bleeding and
a proved viable fetus between 6 and 12 weeks of
gestation were included in the study. Patients
with known or suspected fetal malformations or
with thrombophilia, and patients with missed
abortion or blighted ovum or known uterine mal-
formation were not included. The study was
approved by the Institutional Review Board (Hel-
sinki Committee) and written informed consent
was obtained from every subject. All patients
were studied using transvaginal sonography,
with an endocavitary 5–9 MHz transducer (Vol-
uson 730 Expert; GE Medical Systems, Milwau-
kee, WI), by a single sonographer (M.O.) experi-
enced in performing these examinations. Viabil-
ity of the fetus was documented by visualization
of fetal heart motion. The CRL was measured
and then the Virtual Organ Computer-Aided
Analysis (VOCAL) software was applied using
maximal sweep angle to ensure that the entire
gestational sac volume was applied (Voluson 730
Expert Operation Manual; GE Medical Systems).
Volume measurements were performed after
patient discharge from our ultrasound clinic,
using 308 rotations of the volume of interest;
thus, a sequence of six sections of the volume was
measured. The GSV and the ASV were measured
manually by drawing the contour of the area of
interest of each section (Figures 1 and 2) and the
GSV 2 ASV calculated.
Statistical analysis
Qualitative data were described by frequencies
and percentages; quantitative data were
expressed as mean, standard deviation, median,
and range.
The Wilcoxon rank sum test (Mann--Whitney
test) was used where appropriate to compare the
different parameters. Receiver operator curves
were used to calculate the sensitivity and speci-
ficity of the GSV, ASV, and the GSV 2 ASV in pre-
dicting normal pregnancy outcome. The data
were analyzed using the statistical software
2 JOURNAL OF CLINICAL ULTRASOUND
SPSS 11.5 (Chicago, IL), with p < 0.05 being con-
sidered statistically significant.
RESULTS
Ninety patients were included in the study; 76
(84.4%) delivered after 24 weeks of gestation
(group A), while 14 (15.6%) aborted before 20
weeks of gestation (group B).
The patient data are presented in Table 1. The
groups did not differ in age, parity, number of
previous abortions, or term deliveries. The mean
values of GSV (32.0 6 27.7 cm3 in group A and
26.7 6 29.1 cm3 in group B) and of ASV (21.1 6
25.5 cm3 in group A and 20.6 6 26.0 cm3 in group
B) were not statistically different, while the GSV
2 ASV was significantly different (10.9 6 10.9
cm3 in group A and 6.1 6 8.6 cm3 in group B; p <
0.05). Using receiver operator curves (ROC), the
area under the curve for predicting normal preg-
nancy outcome of the GSV 2 ASV measurement
was 0.654; CI 95% (0.499–0.809) (Figure 3).
When the GSV 2 ASV difference was 1.8 cm3 or
less, abortion was predicted with 84% sensitivity
and 43% specificity.
DISCUSSION
Nazari et al14 reported that a small gestational
sac diameter (MSD) or small CRL or MSD 2
CRL difference of less than 10 mm was a good
predictor of abortion. Dickey et al13 also
reported 80% fetal death when the MSD 2 CRL
was less than 5 mm, 26.5% when the difference
was 5–7.9 mm, and 10.65% when the difference
was 8 mm or more. Bromley et al12 reported a
94% abortion rate when the MSD 2 CRL was
less than 5 mm compared with 8% when it was
greater than 5 mm. Choong et al18 assessed the
MSD, CRL, embryonic heart rate, maternal age,
and gestational age in a multivariate model for
predicting spontaneous miscarriage while the
MSD, CRL, and the MSD 2 CRL were assessed
in univariate logistic regression analysis and
found that the multivariate model demonstrated
the best ROC curve for predicting miscarriage.
This multivariate model did not include volume
measurement.
In our study, the GSV and the ASV did not
differ between the groups, while the GSV 2 ASV
difference was significantly smaller in the abor-
tion group. We have previously reported the nor-
mal values of the GSV and the ASV in normal
pregnancies.19 When all normal cases in our pre-
vious study19 (142 cases) were also incorporated
 GSV AND ASV AND ABORTION PREDICTION

FIGURE 1. Gestational sac volume in an 8-week pregnancy. The volume was measured using the VOCAL software by drawing the outline of the
gestational sac. The volume is displayed in the lower right corner.

FIGURE 2. Amniotic sac volume in the same patient as in Figure 1. The volume was measured using the VOCAL software by drawing the outline
of the amniotic sac. The volume is displayed in the lower right corner.

VOL. 00, NO. 0, MONTH 2012 3

 ODEH ET AL

TABLE 1
Patient Characteristics: Group A: Delivery After 24 weeks; Group B: Abortion Before 20 weeks

Group A Group B
Mean (SD) Mean (SD)
Median (range) Median (range) p

Age (years) 28.2 (5.8) 28.9 (5.6) NS

27.6 (18–45) 28.9 (18–39)
Gestational age (days) 64.6 (14.5) 64.1 (15.2) NS
66.0 (42–90) 64.0 (44–86)
Number of pregnancies (mean) 3.38 (2.4) 2.8 (1.3) NS
3.0 (1–12) 2.5 (1–6)
Previous abortions 0.9 (1.6) 0.8 (1.1) NS
0.0 (0–8) 0.5 (0–4)
Previous term deliveries 1.5 (1.4) 1.0 (0.7) NS
1.0 (0–6) 1.0 (0–2)
CRL (mm) 28.5 (18.4) 26.5 (21.4) NS
25.1 (0.4–72.0) 19.1 (3.7–62)
GSV (cm3) 32.0 (27.7) 26.7 (29.1) NS
24.3 (1.2–92.1) 11.9 (0.5–78.1)
ASV(cm3) 21.1 (25.5) 20.6 (26.1) NS
8.4 (0.01–92.1) 2.4 (0.02–70.3)
GSV-ASV (cm3) 10.9 (10.9) 6.1 (8.6) <0.05
6.8 (0–44.9) 2.6 (0–32.1)

Abbreviations: ASV: amniotic sac volume; CRL: crown rump length; GSV: gestational sac volume; NS: not significant.

the concentrations of sodium, potassium, chlo-

FIGURE 3. Receiver operator curve (ROC) for the difference between
the gestational sac and amniotic sac volumes (GSV 2 ASV).
into the statistical analysis, the results were sim-
ilar to those of the present study: the GSV and
the ASV were not significantly different, while
the GSV 2 ASV was significantly different. Using
ROC, the sensitivity was 89.5% and specificity
was 42.9% for predicting abortion when the GSV
2 ASV was 1.8 cm3 or less.
The composition of the amniotic fluid differs
from that of the extra-amniotic (coelomic) fluid.
Cambell et al20 found significant differences in
4 JOURNAL OF CLINICAL ULTRASOUND
ride, urea, bicarbonate, creatinine, calcium, phos-
phate, bilirubin, protein, albumin, and glucose in
the amniotic fluid and coelomic fluid between 7
and 12 weeks of pregnancy. The osmolality of
both fluids was similar. This difference in fluid
composition may be attributed to their different
origins. The coelomic fluid may be a transudate
produced and secreted by the trophoblast and the
chorion from maternal serum, while amniotic
fluid volume and composition will be determined
by the fetal cord and amnion.20 Thus one might
presume that causes of fetal origin for abortion
might cause a smaller amniotic fluid volume,
while maternal causes for abortion will be associ-
ated with smaller gestational sac volume. How-
ever, our findings do not support this hypothesis,
possibly because the process of abortion after pro-
ven fetal viability is complicated and may be
related to several factors in any given patient.
To investigate further the relation between
GSV, ASV, and abortions, a large number of
patients are needed, preferably with known
causes of the abortion, such as chromosomal
aberrations or thrombophilia. For this purpose, a
large multicenter study is needed.
REFERENCES
1. Simpson JL, Mills JL, Holmes LB, et al. Low fetal
loss rates after ultrasound-proved viability in
early pregnancy. JAMA 1987;258:2555.
 GSV AND ASV AND ABORTION PREDICTION
2. Nyberg DA, Laing FC, Filly RA. Threatened abor-
tion: sonographic detection of normal and abnor-
mal gestation sac. Radiology 1986;158:397.
3. Nyberg DA, Mack LA, Laing FG, et al. Distin-
guishing normal from abnormal gestational sac
growth in early pregnancy. J Ultrasound Med
1987;6:23.
4. Bernard KG, Cooperberg PL. Sonographic differ-
entiation between blighted ovum and early viable
pregnancy. AJR 1985;144:597.
5. Hurwitz SR. Yolk sac sign: sonographic appear-
ance of the fetal yolk sac in missed abortion.
J Ultrasound Med 1986;5:435.
6. Kurtz AB, Needleman L, Pennell RG, et al. Can
detection of the yolk sac in the first trimester be
used to predict the outcome of pregnancy: A pro-
spective sonographic study. AJR 1992;158:843.
7. Levi CS, Lyons EA, Lindsay DJ. Early diagnosis of
nonviable pregnancy with endovaginal ultrasound.
Radiology 1988;167:383.
8. Reece EA, Scioscia AL, Pinter E, et al. Prognostic sig-
nificance of the human yolk sac assessed by ultrasonog-
raphy. Am J Obstet Gynecol 1988;159:1191.
9. Lindsay DJ, Lovett IS, Lyons EA, et al. Yolk sac di-
ameter and shape at endovaginal US: predictors of
pregnancy outcome in the first trimester. Radiol-
ogy 1992;183:115.
10. Laboda LA, Estroff JA, Benacerraf BR. First tri-
mester bradycardia: A sign of impending fetal loss.
J Ultrasound Med 1989;8:561.
11. Achiron R, Tadmor O, Mashiach S. Heart rate as a
predictor of first-trimester spontaneous abortion
after ultrasound-proven viability. Obstet Gynecol
1991;78:330.
VOL. 00, NO. 0, MONTH 2012
12. Bromley B, Hatlow BL, Laboda LA, et al. Small
sac size in the first trimester: a predictor of poor
fetal outcome. Radiology 1991;178:375.
13. Dickey RP, Olar TT, Taylor SN, et al. Relationship
of small gestational sac-crown-rump length differ-
ences to abortion and abortus karyotypes. Obstet
Gynecol 1992;79:554.
14. Nazari A, Check JH, Epstein RH, et al. Relation-
ship of small-for-dates sac size to crown-rump
length and spontaneous abortion in patients with
a known date of ovulation. Obstet Gynecol
1991;78:369.
15. Dighe M, Cuevas C, Moshiri M, et al. Sonography
in first trimester bleeding. J Clin Ultrasound
2008;36:352.
16. Weiss J, Malone FD, Vidaver J, et al. Threatened
abortion: a risk factor for poor pregnancy outcome,
a population–based screening study. Am J Obstet
Gynecol 2004;190:745.
17. Odeh M, Tendler R, Kais M, et al. Gestational sac
volume in missed abortion compared to normal
pregnancy. J Clin Ultrasound 2010;38:367.
18. Choong S, Rombauts L, Ugon A, et al. Ultrasound
prediction of risk of spontaneous miscarriage in
live embryos from assisted conceptions. Ultra-
sound Obstet Gynecol 2003;22:571.
19. Odeh M, Hirsh Y, Degani S, et al. Three-dimen-
sional sonographic volumetry of the gestational
sac and the amniotic sac in the first trimester.
J Ultrasound Med 2008;27:373.
20. Campell J, Wathen N, Macintosh M, et al. Bio-
chemical composition of amniotic fluid and extra-
embryonic coelomic fluid in the first trimester of
pregnancy. Br J Obstet Gynaecol 1992;99:563.
5