The Expanded Amnion Sign - Journal of Ultrasound in Medicine

The Expanded Amnion Sign
Evidence of Early Embryonic Death
Abbreviations
ECA, embryonic cardiac activity; CRL, crown-rump
length
Received May 12, 2009, from the Department of
Radiology, University of California, San Francisco,
California USA. Revision requested June 2, 2009.
Revised manuscript accepted for publication June
8, 2009.
Address correspondence to Roy A. Filly, MD,
Department of Radiology, University of California,
505 Parnassus Ave, L374, San Francisco, CA 94143-
0628 USA.
E-mail: roy.filly@radiology.ucsf.edu
Videos online at www.jultrasoundmed.org
N. Tugce Yegul, MD, Roy A. Filly, MD
Article
Objective. The purpose of this study was to assess the positive predictive value for confirming early
embryonic death in the clinical scenario wherein an embryo is identified without a visible heartbeat, but
the embryonic crown-rump length (CRL) is 5 mm or less. Methods. We conducted a retrospective study
of 882 first-trimester sonograms with abnormal findings among women who were threatening to
abort. Eight hundred six met the inclusion criteria. Results. Among the cohort of 806 cases, 520
(64.5%) had an identifiable embryo, and 255 of those with an identifiable embryo had a visible amnion
(49.0%). One hundred sixteen of the 255 with a visible amnion and an identifiable embryo without a
heartbeat had a CRL that measured 5 mm or less (45.5%). The CRL of these embryos ranged from 1.7
to 5.4 mm (ie, when rounded to the nearest millimeter, these embryos would be 5 mm) with the breakdown
as follows: those measuring less than or equal to 3.4 mm (n = 28), those measuring 3.5 to 4.4
mm (n = 45), and those measuring 4.5 to 5.4 mm (n = 43). Eight of these 116 patients did not
have any documented follow-up. In the remaining 108 patients, pregnancy failure was confirmed.
Conclusions. We conclude that any embryo that is surrounded by an amnion and that also lacks a heartbeat
has unfortunately but definitively died. This is equally true for embryos measuring less than 5 mm in
CRL. Key words: amnion; early pregnancy failure; embryonic death; sonography; threatened abortion.
bstetricians have long realized that sonography
can frequently determine whether a pregnancy
in a woman who is threatening to abort
has already failed. 1 Pregnancy failure is often
confirmed after a single examination. Although the literature
strongly supports a variety of sonographic features
that unambiguously confirm failure, some features that
are suspicious require 1 or more follow-up examinations
to confirm that pregnancy failure has indeed occurred.
Among these is the circumstance wherein an embryo is
identified without a visible heartbeat, but the embryonic
crown-rump length (CRL) is 5 mm or less. This is because
there are literature references to the sonographic detection
of early embryos before the onset of cardiac contractility.
2–6 O
Therefore, it is possible that tiny embryos without
visible heartbeats are actually alive, but their heart has
simply not yet begun to beat. Thus, repeated sonograms
are commonly recommended when tiny embryos without
heartbeats are identified. Our institution recently
© 2009 by the American Institute of Ultrasound in Medicine J Ultrasound Med 2009; 28:1331–1335 0278-4297/09/$3.50

The Expanded Amnion Sign
1332
reported that in a series of 37 such embryos consecutively
seen over a 4-year period in women
who were threatening to abort, none subsequently
had a normal obstetric delivery or development
of embryonic cardiac activity (ECA) but
instead were confirmed as early pregnancy failures.
7 Our results were similar to those of Levi et
al 4 when their data were confined to women with
vaginal bleeding.
Nonetheless, practitioners will likely still feel a
need to recommend follow-up examinations
when an embryo is identified without a visible
heartbeat, but the embryonic CRL is 5 mm or less
because this has been a long-held belief. In this
article, we report our observations of what we
have termed the “expanded amnion sign” to
describe a feature that confirms early embryonic
death in the clinical scenario wherein an
embryo is identified without a visible heartbeat,
but the embryonic CRL is 5 mm or less. In 1995,
we observed that at the earliest sonographically
visible stages of amnion identification, there was
a linear relationship of amnion diameter to CRL. 8
Thus, by the time that a relatively complete
amnion is visualized, an embryo with a CRL of at
least 7 mm (and usually larger) should be detected
within its confines (Video 1). Horrow 9 initially
suggested this observation but instead computed
the menstrual age at which the amnion was
first seen as 6.5 weeks, at which time the CRL
equals 7 mm. All embryos with a CRL of 7 mm or
greater would necessarily have a visible heartbeat
if alive. Ikegawa 10 later studied 1 group each
of normal and failed pregnancies relative to
amnion detection rates and also concurred that
the amnion was not detectable in normal pregnancies
until the embryo achieved a CRL of 7
mm or greater. Although his data produced only
10 cases meeting the criteria in our study,
Ikegawa 10 suggested that an enlarged amniotic
sac surrounding an embryo of less than 7 mm
that also failed to show cardiac activity suggested
pregnancy failure. In this study, we investigated a
consecutive series of early pregnancies in which
an embryo with a CRL of 5 mm or less and lacking
a heartbeat was identified, but surrounding
the embryo, one could also see the amnion. We
calculated the positive predictive value of this
finding to confirm early pregnancy failure in this
patient cohort.
Materials and Methods
A retrospective review of the University of
California San Francisco Medical Center ultrasound
database from January 1999 to October
2008 revealed 882 first-trimester sonograms
among women who were threatening to abort
and had abnormal sonographic findings including
the inability to show a fetal heartbeat. All of
these sonograms were reviewed for the presence
of an embryo and a visible amniotic sac. To confirm
amniotic sac visualization, both the yolk sac
and the amnion needed to be visualized, thus
excluding large yolk sacs as a possible explanation
of amnion identification. Of course, an
embryo resides within the amnion, not within
the yolk sac. Thus, it is sufficient to identify the
embryo to confirm that the surrounding membrane
is the amnion (Figure 1). However, we
used the more stringent criterion of visualization
of both the yolk sac and the amnion.
Follow-up sonograms of the same pregnancy (n
= 45) were excluded (other than for confirmation
Figure 1. All sonograms in our patient cohort were reviewed for
the presence of an embryo and a visible amniotic sac.
Visualization of both the yolk sac and the amnion excluded large
yolk sacs as a possible explanation for amnion identification. Of
course, the embryo resides within the amnion, not within the
yolk sac. Thus, it is sufficient to identify the embryo to confirm
that the surrounding membrane is the amnion.
J Ultrasound Med 2009; 28:1331–1335

of pregnancy failure or continuation). Thirtyone
of the sonograms were not retrievable at the
time of the review and were also excluded.
Therefore, 806 sonograms were included in the
study. The University of California San Francisco
Committee on Human Research approved the
study.
Each patient was scanned with both transabdominal
and endovaginal techniques (Acuson
Sequoia; Siemens Medical Solutions, Mountain
View, CA) per institutional protocol by a sonographer.
Transabdominal transducers used
included vector arrays with selectable frequencies
ranging from 1 to 4 MHz and curved arrays
with selectable frequencies ranging from 2 to 6
MHz. The endovaginal transducer used was a
tightly curved array with selectable frequencies
ranging from 4 to 8 MHz. Because the ultrasound
unit computes the CRL to the nearest tenth of a
millimeter, the CRL was rounded to the nearest
millimeter for purposes of data analysis.
The electronic medical records of these patients
were reviewed to determine the pregnancy outcome.
Pregnancy failure was confirmed if any of
the following were seen within 6 months of the
first sonogram: pathologically proven embryonic
death, a follow-up sonogram showing embryonic
death, a sonogram showing a new firsttrimester
intrauterine pregnancy, laboratory
results showing decreasing β-human chorionic
gonadotropin levels, or a clinical note indicating
spontaneous abortion.
Results
Among the cohort of 806 cases, 520 (64.5%) had
an identifiable embryo, and 255 of those with an
identifiable embryo had a visible amnion (49.0%;
Table 1). As stated above, none of the embryos
had a visible heartbeat. One hundred sixteen of
the 255 cases with a visible amnion and an identifiable
embryo without a heartbeat had a CRL
that measured 5 mm or less (45.5%; Figure 2 and
Video 2). The CRL of these embryos ranged from
1.7 to 5.4 mm (ie, when rounded to the nearest
millimeter, these embryos would be 5 mm) with
the breakdown as follows: those measuring less
than or equal to 3.4 mm (n = 28), those measuring
3.5 to 4.4 mm (n = 45), and those measuring
4.5 to 5.4 mm (n = 43).
Table 1. Data Analysis Related to Embryo Size
Eight of these 116 patients did not have any
documented follow-up. In the remaining 108
patients, pregnancy failure was confirmed.
Therefore, on the basis of the results of our study,
the positive predictive value of the expanded
amnion sign in determining early pregnancy failure
was 100%.
Discussion
Sonography’s track record for determining
whether an early pregnancy has already failed
has made this examination pivotal in the management
of pregnancies with bleeding in the first
Figure 2. Endovaginal sonogram of a tiny embryo surrounded
by an amniotic sac, a sac too large for the observed CRL. As
noted in the introductory text, for very early embryos, the CRL is
linearly related to the amnion diameter. Thus, this embryo should
have had a CRL of approximately 10 mm, much larger than in
this case. Note also that the entirety of the amnion is visible as a
complete membrane encircling the embryo. Compare with
Video 1.
Yegul and Filly
Embryos Amnion Seen Amnion Not Seen Total
≤5.4 mm a 116 b 43 159
>5.4 mm a 139 222 361
Total 255 265 520
a None of the embryos had a heartbeat.
b All of these cases could be confidently diagnosed as pregnancy failures
even though the embryos had a CRL of 5 mm or less. Thus, 116 of 159 small
embryos (73%) could be diagnosed with confidence.
J Ultrasound Med 2009; 28:1331–1335 1333

The Expanded Amnion Sign
1334
trimester. 1–8 This is because pregnancy failure
often can be confirmed with a single examination
performed in less than 30 minutes. As noted
above, the literature supports that a variety of
sonographic features unambiguously confirm
early pregnancy failure.
Previous studies have cautioned that there are
embryos with a CRL of 5 mm or less and without
sonographically detectable ECA that may be
alive because the embryo’s heart has not yet
begun to beat. 2,3,5,6,8 More recent studies have
suggested that 3 mm, not 5 mm, should be the
new cutoff 4,7 for diagnosis of death if ECA is not
seen. Yet another study went to the opposite
extreme and recommended using a CRL of 10
mm as the cutoff before diagnosing embryonic
death. 11 These recommendations usually necessitate
a follow-up sonographic examination, thus
incurring additional medical costs and furthering
maternal uncertainty regarding the status of
the pregnancy. We recently published outcomes
of embryos with a CRL of 5 mm or less without
ECA among 546 women with vaginal bleeding.
None of these embryos later proved to be alive.
Although embryos without ECA and a CRL of
less than 5 mm have progressed to normal
pregnancies in asymptomatic women, our
results and those of Levi et al 4 in bleeding
women showed that there were no successful
pregnancies in either cohort, and follow-up
sonography may be superfluous. 7 The data from
this study show that if an embryo with a CRL of 5
mm or less and without sonographically
detectable ECA is also surrounded by an amnion,
a virtually certain diagnosis of embryonic death
can be made. Horrow 9 also suspected as early as
1992 that this observation identified failed pregnancies
but thought that her sample size was too
small to draw definitive conclusions.
Another problem solved by the expanded
amnion sign described herein is the so-called
chorionic bump. 12 This feature may be mistaken
for an embryo without ECA. However, if the
embryo is surrounded by amnion, then the
examiner can be certain that it is not a chorionic
bump. Recall that among the 806 sonograms
included in the study, 326 (40.4%) had a visible
amnion. Thus, with modern endovaginal transducers,
amnion visualization is common.
Because we used endovaginal transducers with
selectable frequencies from 4 to 8 MHz, we were
not using the current state-of-the-art transducers
available that image at frequencies of greater
than 8 MHz. Therefore, it is probable that future
studies will show even higher proportions of
early pregnancy failure wherein the amnion is
visualized.
Although it would be preferable to confirm our
results prospectively, we think that this finding,
the expanded amnion sign, can be effectively
used with confidence on the basis of the results
of this retrospective study alone. We conclude
that any embryo that is surrounded by an
amnion and that also lacks a heartbeat has
unfortunately but definitively died. This is equally
true for embryos measuring less than 5 mm in
CRL. Furthermore, the ability to show the amnion
in this circumstance is common. Among embryos
lacking a heartbeat and having a CRL of less than
5 mm, 116 of 159 cases (73%) had a visible
amnion.
References
1. Laing FC, Frates MC, Benson CB. Ultrasound evaluation
during the first trimester of pregnancy. In: Callen PW (ed).
Ultrasonography in Obstetrics and Gynecology. 5th ed.
Philadelphia, PA: WB Saunders Co; 2008:181–224.
2. Brown DL, Emerson DS, Felker RE, Cartier MS, Smith WC.
Diagnosis of early embryonic demise by endovaginal
sonography. J Ultrasound Med 1990; 9:631–636.
3. Goldstein RB. Endovaginal sonography in very early pregnancy:
new observations. Radiology 1990; 176:7–8.
4. Levi CS, Lyons EA, Zheng XH, Lindsay DJ, Holt SC.
Endovaginal US: demonstration of cardiac activity in
embryos of less than 5.0 mm in crown-rump length.
Radiology 1990; 176:71–74.
5. Pennell RG, Needleman L, Pajak T, et al. Prospective comparison
of vaginal and abdominal sonography in normal
early pregnancy. J Ultrasound Med 1991; 10:63–67.
6. Schouwink MH, Fong BF, Mol BW, van der Veen F.
Ultrasonographic criteria for non-viability of first trimester
intra-uterine pregnancy. Early Pregnancy 2000; 4:203–213.
7. Aziz S, Cho RC, Baker DB, Chhor C, Filly RA. Five-millimeter
and smaller embryos without embryonic cardiac activity:
outcomes in women with vaginal bleeding. J Ultrasound
Med 2008; 27:1559–1561.
8. McKenna KM, Feldstein VA, Goldstein RB, Filly RA. The
empty amnion: a sign of early pregnancy failure.
J Ultrasound Med 1995; 14:117–121.
9. Horrow MM. Enlarged amniotic cavity: a new sonographic
sign of early embryonic death. AJR Am J Roentgenol 1992;
158:359–362.
J Ultrasound Med 2009; 28:1331–1335

10. Ikegawa A. First-trimester detection of amniotic sac in relation
to miscarriage. J Obstet Gynaecol Res 1997; 23:283–
288.
11. Hately W, Case J, Campbell S. Establishing the death of an
embryo by ultrasound: report of a public inquiry with recommendations.
Ultrasound Obstet Gynecol 1995;
5:353–357.
12. Harris RD, Couto C, Karpovsky C, Porter MM, Ouhilal S.
The chorionic bump: a first-trimester pregnancy sonographic
finding associated with a guarded prognosis.
J Ultrasound Med 2006; 25:757–763.
Yegul and Filly
J Ultrasound Med 2009; 28:1331–1335 1335