Evan
H. Dillon,
MD
#{149} Ann
L. Feyock,
Pseudogestatlonal Differentiation Intrauterine
Index
terms:
Abortion,
856.823
#{149} Pregnancy,
Ultrasound
(US),
Radiology
1990;
From ogy,
the
E
Yale
RSNA
scientific
Supported ROl
School
revision received
Received
requested February National
HD24660.
1990
noassays
856.12984
Radiol-
of Medicine,
CT 06510.
333
From
the
1989
November
January 16, 1990; 26; accepted April Institutes
Address
of Health reprint
requests
is a potentially
(1).
Fortunately,
major
ad-
vances in the noninvasive diagnosis of ectopic pregnancy have also been made in recent years. The availability of sensitive and specific radioimmu-
856.12984.
of Diagnostic
pregnancy
CTOPIC
73,700
176:359-364
Haven,
J. W. Taylor,
condition that has increased dramatically in incidence in recent years. Between 1970 and 1986, the annual number of ectopic pregnancies in the United States more than quadrupled, from 17,800 to
Pregnancy. ectopic,
studies,
assembly.
by
to K.J.W.T. © RSNA,
studies,
Doppler
University
St. New
28, 1989; revision
US
#{149} Kenneth
lethal
#{149}
#{149}
Department
Cedar
grant
856.825
Pregnancy,
856.823
RDMS
MD,
PhD
Sacs: Doppler US from Normal or Abnormal Pregnancies’
Doppler ultrasound (US) evaluation of 40 empty intrauterine sac-like structures was performed to evaluate the ability of this technique to permit distinction between intrauterine pregnancy and pseudogestational sac associated with ectopic pregnancy. Proof of the location of the pregnancy was available in all cases. There were 31 intrauterine pregnancies, of which 23 were missed or incomplete abortions and eight were early normal pregnancies. With an insonating frequency of 3 MHz, the average frequency shift detected from these intrauterme pregnancies was 1.7/1.0 kHz (peak systolic/end diastolic ratio). Nine pseudogestational sacs were evaluated, of which seven demonstrated no flow and two demonstrated minimal flow that averaged 0.4/0.1 kHz. Defining intrauterine peritrophoblastic flow as a peak systolic frequency shift of 0.8 kHz or greater (equivalent to 21 cm/sec with an angle of 0#{176}) correctly classifies 26 of the 31 intrauterine pregnancies and all of the nine pseudosacs. The sensitivity of the Doppler technique for the detection of intrauterine pregnancies was 84%, and the specificity was 100%.
abnormalities,
RTR,
6.
for
the
beta
subunit
of hu-
man chonionic gonadotropin ($-hCG) now allows determination of whethen a patient presenting with signs and symptoms of ectopic gestation is pregnant. The location of the pregnancy, whether intraon extrauterme, can often be established with ultrasound (US) techniques. Imaging is most useful when an intrauterine pregnancy is seen, thus making the existence of an ectopic pregnancy highly unlikely. When an intrautenme gestation is not seen, transabdominal imaging may yield only nonspecific findings, such as the presence of an adnexab mass or an intrautenine sac-bike structure (2-7). The recent introduction of high-frequency transvaginal probes with increased resolution has allowed more detailed evaluation of these extrauterine masses as well as earlier demonstration of an intrauterine gestation (4,5,8,9). Imaging alone, however, provides only anatomic information that may not permit differentiation between an extrauterine gestational sac and some other adnexal mass or between a very early on abnormal intrauterine pregnancy and a pseudogestational sac. Doppler techniques can aid in distinguishing between gestational sacs and other structures by providing hemodynamic information to supplement the anatomic information provided by imaging alone. Doppler techniques have recently been applied transabdominally to diffementi-
ate between extrauterine sacs and other masses before a yolk sac and fetab pole are visible (10). Confusion may persist, however, when an intrauterine sac-like structure is seen that may represent a pseudogestational sac associated with an ectopic pregnancy on an early or abnormal intrauterine pregnancy. Although visualization of a double decidual sac suggests
that
the
intrauterine
structure
is
a gestational sac, abnormal intrauterme pregnancies, such as an incomplete abortion, may fail to exhibit this
sign,
and
some
pseudosacs
mimic this finding (3,11). the usefulness of Doppler in distinguishing
may
To evaluate techniques
between
pseudo-
gestational sacs and normal or abnormal intrauterine pregnancies, we performed a prospective study of patients
with
like
structures.
empty
PATIENTS Patients pregnancy
intrauterine
AND
sac-
METHODS
suspected
of having
at clinical
examination
an ectopic were
referred for US localization of the pregnancy. Most patients initially underwent transabdominal imaging. If this image demonstrated an early intrauterine pregnancy on probable early intrauterine pregnancy,
no
performed.
A probable
pregnancy
further
was
defined
uterine
sac-like
ameten
less than
shape, tient
and with
transabdominal performed.
tune
was
like
structure
(especially
as an
empty
with
decidual level equal
mIU/mL Reference
Doppler abnormal
defined
chorionic
di-
or oval sac in a pato on great-
(6,000 IU/L) Preparation
(First
scan dem-
an empty intrauterine with an abnormal An
intra-
a mean
2.0 cm, round
a double a $-hCG
sac-like
appearance,
evaluation sac-like
was struc-
as an intrauterine that
had
a crenated
Abbreviations: ternational
intrauterine
If the transabdominal
onstnated structure
man
was
early
structure
en than 6,000 International
[FIRP]).
evaluation
an
appearance),
fl-hCG
beta
gonadotropin, Reference
sac-
abnormal
subunit FIRP
shape
lacked
of huFirst
In-
Preparation.
359
Figure 1. Peak systolic frequency shifts plotted against end diastolic frequency shifts for 31 intrauterine pregnancies (JUP) and nine pseudogestational sacs. A 3-MHz transducer frequency was used. A cutoff value of 0.8 kHz for peak systolic frequency shift resulted in the correct classification of 26 of the 31 intrauterine pregnancies and all of the
PEAK
SYSTOLIC
(kHz)
4
3
pseudosacs.
a double at a mean
decidual diameter
sac, on remained empty greaten than 2.0 cm
2
(11,12).
Other indications of abnormality used to justify Doppler examination included visualization of a sac-like structure at low hCG levels, or an inadequate rise or an actual decline in the hCG level (2,7,11-13). If the initial tnansabdominal imaging findings were equivocal, a transvaginal scan
was
obtained
tnautenine tients
to characterize
structure
the
further.
underwent
only
A few
I
in-
pa-
a transvaginal
ex-
0
0.5
I
amination. If the transvaginal scan demonstrated an abnormal intrauterine saclike structure on if the other criteria for abnormality were met, a tnansvaginal Doppler examination was performed. Finally, eight patients with normal-appearing empty intrauterine sac-like structures were examined before elective abortion was performed. We used three different US machines in
this
study.
An
ATL
600
unit
END
2
1.5
2.5
DIASTOLIC(kHz)
(Advanced
Technology Laboratories, Bothell, Wash) with an insonating frequency of 3 MHz was used for most of the transabdominal Doppler examinations. A Toshiba SSA 270
machine
Tustin, Doppler
(Toshiba
Medical
Calif) with capabilities
frequency
pulsed and
of 3 MHz
transabdominal 128 unit (Acuson,
was used A 3-MHz
was
examination. Mountain
for the remaining probe with pulsed
Doppler
capabilities
was
examinations. and colon
used
dominal examinations, and probe, initially with pulsed and later with color Doppler, for tnansvaginal examinations. output the
was
set
calculated
peak
as possible intensity
tions,
the
a small
sample
so that (spatial
was
pulsed
Figure 2. Patient 1. Transabdominal pulsed Doppler examination of an incomplete abortion showing penitrophoblastic flow. An intrauterine sac-like structure with a poorly echogenic rind was seen in this patient with vaginal spotting and a decreasing 3-hCG level. Doppler US demonstrated high-velocity, low-impedance flow (0.5-kHz scale: 1.8/1.0 [3MHz probe]). Dilation and curettage yielded products of conception.
transab-
a 5-MHz capabilities was used The power
as low
average)
In
for
in situ
temporal
mW/cm2.
Systems,
and color an insonating used in one An Acuson View, Calif)
less
than
Doppler
94
examina-
volume
(typically
3
mm) was used to prevent the detection of signals from nearby structures. Signals were sought from the echogenic rind sunrounding
the
ticular
sac-like
attention
area of the immediately structure,
given
and
from
The
signals
quency
shift
and
were The
any
obtained observed
thickened
visualized
the
location were signals
largest
in each
preference, tion; findings in the case
360
case,
the
of
in
in vitro
of an earlier US of the eight normal
Radiology
#{149}
including,
date
fre-
from recorded were
of the pregnancy of either pathologic tra- or extrauterine quent US.
was
required, evaluation contents
or
by means of the insubse-
missed of
which on correlat-
age, for was order
of
fertiliza-
study, or, pregnan-
RESULTS Doppler examination was performed in 40 patients who had empty intrauterine sac-like structures. Proof of the location of the pregnancy was subsequently available in all cases. Intrauterine pregnancies were diagnosed by means of dilation and Curettage or prior or subsequent US in 31 of these 40 cases, 23 of which were
or
eight
of
abortions and early normal inThe average
incomplete
which
tmautenine
were
pregnancies.
gestational
age
me
pregnancies
range
of
inal
Doppler
34
the
for was
to
109
days.
31 intrauter62 days, with
a
Transabdom-
examination was per25 of these 31 cases. The ATL 600 unit with a 3-MHz pulsed Doppler probe was used in 24 cases. The Acuson 128 unit with a 3-MHz probe with both pulsed and color capabilities was used in two cases, one of which was also examined with the ATL 600 machine. Six patients underwent evaluation with the Acuson unit with a tmansvaginal 5-MHz probe with both pulsed and color capabilities. Because the Acuson unit formed
adnex-
ed with the estimated gestational which the best available indicator used
par-
myometnium the sac-like
with the
with
to any
rind, from the surrounding
al mass. they film.
structure,
cies, the current mean sac diameter; an earlier fl-hCG level; or the date of onset the last menstrual period. To be included in the study group, proof of the location
in
August
1990
Figure
3. Patient 2. Transvaginal colon Doppler flow image of an incomplete abortion showing penitrophoblastic flow. Obvious flow is seen around this abnormally shaped sac-like structure. Dilation and Cu-
nettage
yielded
reports
products
pulsed
Doppler
bocity
in centimeters
these
values
frequency
of conception.
were shifts
shifts per
as yeback
use
to
Data
Analysis
of the Most
of the
examinations
were
per-
Doppler equation to provide companable data. The highest observed frequency shifts in these 31 intrauterine pregnancies averaged 1.7/1.0 kHz (peak systolic/end diastolic ratio), with a range from 0.3/0 to 4.0/2.4 kHz. The highest frequency shift and largest amplitude were generally detected from only one portion of the
formed with a 3-MHz transducer. The few values obtained with the 5MHz transducer were adjusted, with use of the Doppler equation, to be comparable with the 3-MHz data. The peak systolic and end diastolic frequency shifts observed around
myometnium,
are plotted
typically
the
pant
adja-
these
intrauterine
cent to the thickened portion of the echogenic rind, which corresponds to the site of the developing placenta. These signals could be detected with both the ATL and Acuson units. The transvaginal probe also enabled detection of such flow in the cases in which it was used. Color flow imaging,
ficity
applied
either
transabdominably
showed
penitrophoblastic
flow
in seven
flow,
flow was noted with Doppler technique. Nine
patients
had
of
only
the
minimal
pulsed
176
Number
#{149}
flow
a pathologically
2
like
structure.
sys-
was
58/28
Patient
Cases 1.-A
17-year-old
female
adolescent presented with vaginal spotting of 1 weeks duration after 11 weeks of amenorrhea. Her serum 3hCG level had decreased during the past 2 days to 34,800 mIU/mL (34,800 IU/L) (FIRP). Transabdominal US demonstrated mass
an
structure (Fig
empty
and
2). Transabdominal
intrauterine
no adnexal
flow
cm/sec
(converted
quency
shift
of 2.3/1.1
justing
the
value
imaging
to a fre-
kHz
to make
after
ad-
it compara-
ble with the 3-MHz data). Dilation and curettage yielded products of conception. Patient 3.-An 18-year-old woman with a prior left-sided ectopic pregnancy presented with spotting of 3 days
duration
after
44 days
of amen-
omrhea. Her serum 3-hCG level on the day of presentation was 8,130 mIU/mL (8,130 IU/L) (FIRP). TransUS
imaging
demonstrated
a complex mass in the right adnexa and a sac-like structure within the uterus (Fig 4). Tmansabdominal duplex Doppler evaluation demonstrated high-velocity, low-impedance flow around the might adnexal mass (2.3/1.0
kHz)
but
no
detectable
flow
around the intrauterine structure (0/ 0 kHz). A right-sided tubal ectopic pregnancy was found at lapanoscopy, while dilation and curettage yielded products
of conception,
thus
con-
firming that the intrauterine structune was a pseudogestational sac. Patient 4.-A 22-year-old woman presented with abdominal pain after 49 days of amenorrhea. Tmansvaginal imaging demonstrated an irregularly shaped
intrauterine
containing
an
5). A complex
100%.
Illustrative
Color
with use of the transvaginal 5-MHz probe demonstrated obvious flow around the sac (Fig 3). Pulsed Dopplen examination yielded a peak systolic/end diastolic velocity ratio of
no
as a peak
tobic frequency of 0.8 kHz or greater correctly classifies 26 of 31 intrauterme pregnancies and all of the nine pseudosacs. With use of this cnitenion, the sensitivity of this technique for the detection of intrauterine pregnancies was 84% and the speci-
sac-like
proved ectopic pregnancy, indicating that the visualized intrauterine structune was a pseudogestational sac. Three of these nine also underwent dilation and curettage, which failed to show intrauterine products of conception. The average gestational age for these patients was 47 days, with a range of 35 to 72 days. Of the nine pseudosacs, five were evaluated Volume
structures
1. Defining
obvious
the eight cases in which it was used. Because color flow imaging does not provide quantitative information, pulsed Doppler evaluation was also performed in each of these eight cases. In the one case in which color flow imaging failed to demonstrate penitrophoblastic
sac-like
in Figure
penitrophoblastic
or tnansvaginally,
presented with vaginal spotting after 39 days of amenorrhea. Tmansvaginal imaging demonstrated an irregular and crenated empty intrauterine sac-
abdominal
second,
converted with
transabdominally with the ATL pulsed Doppler probe; two were examined transabdominally with the Acuson unit, including one that was also examined with the ATL unit; and one was examined transabdominally with the Toshiba machine. Three of the nine patients with pseudosacs underwent tnansvaginal evaluation with the Acuson 5-MHz pulsed Doppler probe before colon capability was available, including one who also underwent examination with the ATL unit. Of the nine pseudosacs, seven demonstrated no detectable flow around the structure (0/0 kHz), whereas two demonstrated minimal flow that averaged 0.4/ 0.1 kHz. Transabdominal color flow imaging demonstrated no evidence of penitrophoblastic flow around the pseudosac in the three cases in which it was used.
sac-like apparent yolk mass was seen
structure sac (Fig in the
night adnexa. Tnansvaginal pulsed Doppler evaluation of the mass demonstrated typical penitrophoblastic flow, indicating the presence of an ectopic pregnancy. Evaluation of the intrautenine sac-like structure demonstrated no evidence of penitrophoblastic flow,
indicating
that
the
structure
represented
a pseudogestational sac despite the apparent yolk sac (Fig 5). Findings at laparoscopy confirmed the presence of an ectopic pregnancy.
duplex
Doppler evaluation demonstrated high-velocity, low-impedance flow (1 .8/1 .0 kHz) around the intrauterine structure, and flow consistent with a corpus luteum (1.3/0.7 kHz) was observed in the left ovary. Dilation and curettage yielded products of conception, which confirmed the presence of an incomplete abortion. Patient 2.-A 34-year-old woman
DISCUSSION Ectopic pregnancy now accounts for 1.4% of all reported pregnancies (1). The possibility of an ectopic gestation must therefore be considered in any woman of reproductive age who presents with suggestive signs and symptoms (14). However, the classic clinical triad of abnormal utemRadiology
361
#{149}
me bleeding, pelvic pain, and an adnexal mass is present in less than 50% of patients (15). Thus, diagnosis rests primarily with the recently developed noninvasive techniques that allow the existence and, generally, the location of a pregnancy to be determined. Pregnancy can be established by means of the serum fl-hCG level, and the location can usually be established by means of US imaging. The recent addition of tnansvaginal imaging has improved the diagnostic capabilities of US in cases of suspected ectopic pregnancy (4,5). This technique usually provides additional information that is not available with a transabdominal
approach.
Nyberg
et
al (4) found that transvaginal sonognaphy yielded additional information in 60% of cases. Among the 25 ectopic pregnancies in their study, the transvaginal approach demonstrated an extrauterine sac-like structure in 10 cases. While the visualization of a thick-walled adnexal ring is highly suggestive of an extrauterine gestation, this appearance alone is not completely specific. Other structures may mimic this appearance, resulting in a false-positive diagnosis of ectopic pregnancy. The lack of complete specificity of adnexal findings may be compounded by the occasional visualization of an intrauterine sac-like structure. Such pseudogestational sacs may be seen in up to 20% of ectopic pregnancies (16). US imaging alone may not provide sufficient information to differentiate these pseudogestational sacs from early or abnormal intrauterine gestations. Doppler evaluation, which has mecently been used to supplement US imaging in cases of suspected ectopic pregnancy (10,17), adds physiologic information to the anatomic detail provided by imaging. Taylor et al (10) identified a characteristic pattern of flow around extrauterine gestational sacs. High-velocity (Doppler shift of up to 4 kHz at an insonating frequency of 3 MHz) and bow-impedance (Pourcelot index of 0.385 ± 0.2) flow was seen in 38 (54%) of 70 ectopic pregnancies. The corpus luteum had a similar but lower-velocity (less than 2. 1 kHz), statistically highem-impedance (Pouncebot index of 0.504 ± 0.2) pattern of flow. Extrauterine penitnophoblastic flow, defined as a peak systolic frequency shift of 2.1 kHz or greater, was noted by Taylor et al (10) in 29 (41%) of 70 ectopic pregnancies. Only nine (15%) of the 59 corpora lutea demonstrated such flow. Thus, Doppler evaluation was often able to add useful physiologic 362
Radiology
#{149}
a.
b. Figure 4. Patient 3. Transabdominal Doppler US evaluation of ectopic pregnancy with pseudogestational sac. (a) Absence of flow around the pseudogestational sac. Sagittal scan of the pelvis demonstrated an intrauterine sac-like structure in this patient with pain and vaginal spotting, and Doppler examination demonstrated no detectable flow. Dilation and curettage yielded no products of conception. (b) Peritrophoblastic flow around ectopic gestation. Transverse scan demonstrated the pseudogestational sac and a right-sided adnexal mass. Doppler examination demonstrated high-velocity, low-impedance flow around this mass (0.5-kHz scale: 2.3/1.0 [3-MHz probe]). Findings at laparoscopy confirmed a right-sided tubal ectopic
pregnancy.
Figure 5. Patient 4. Transvaginal pulsed Doppler examination of a pseudogestational sac. US imaging demonstrates an irregularly shaped intrauterine sac-like structure with an apparent
yolk
sac. However,
Doppler
amination demonstrates no evidence penitrophoblastic flow, suggesting rect diagnosis of a pseudogestational Findings at laparoscopy confirmed ic pregnancy.
exof
the
corsac. an ectop-
information that helped in distinguishing extrauterine gestational sacs from other adnexal masses. We used a similar approach to differentiate between pseudogestational sacs and intrauterine pregnancies by noting the presence or absence of penitrophobbastic flow. Our results indicate that Doppler evaluation of the uterine contents usually aids in differentiating between these structunes. While a high cutoff value (2.1 kHz with use of a 3-MHz transducer) is necessary penitmophobbastic
to define extrauterine flow to prevent
confusion with luteab flow, lower cutoff value (0.8 kHz
a much with use
of a 3-MHz
transducer)
may
be used
to define intrauterine blastic flow, because
penitrophono normal intrauterine structure produces signals that exceed this level. This cutoff value is valid only when a 3-MHz transducer is used; if a 5-MHz transducer is used, the cutoff value would be 1.3
kHz. Alternatively, a velocity of 21 cm /sec may be used as the cutoff value
with
when frequency
any
the
transducer
angle shift
frequency
is 0#{176}. A peak of 0.8
kHz
systolic
or greater
August
1990
(assuming a 3-MHz transducer) was noted in 26 of 31 intrauterine pregnancies but in none of the nine pseudogestational sacs. Similar low-impedance penitrophoblastic signals were detected transvaginally in a large series of normal and abnormal intrauterine pregnancies by Schaaps and Soyeur (18). However, no specific numerical values were reported for the observed flow, and no pseudogestational sacs were evaluated. Some intrauterine pregnancies may fail to demonstrate peritrophoblastic flow, as occurred in five of 31 cases in our study. Four of the five false-negative diagnoses occurred among the earlier pregnancies exammed. One was an early normal pregnancy examined tnansvaginally at 35 days, while two were early normal pregnancies examined transabdominally at 34 and 43 days. The other false-negative diagnosis among the early pregnancies occurred in an incomplete abortion examined transabdominally at 40 days. The transvaginal approach demonstrated peritrophoblastic flow earlier than did the transabdominal
approach.
While
the
tnansvaginal Doppler results were negative in one normal pregnancy 35 days,
they
were
positive
at
in three
pregnancies at 36 days (two normal pregnancies, one incomplete abontion). Transabdominal Doppler findings were negative in three pregnandies at 34, 40, and 43 days (two normal pregnancies, one abortion) but were positive in three other early pregnancies at 40, 42, and 43 days (two normal pregnancies, one abontion). Penitrophoblastic flow was consistently detected transvaginally aften 36 days and transabdominally aften 43 days, except in one incomplete abortion examined at 48 days. This case may have represented a virtually complete abortion to which blood flow had essentially ceased. No
false-positive
diagnoses
of in-
trautenine pregnancies were made in our study. Myometnial and endometrial flow, however, have occasionalby been noted to produce peak systolic frequency shifts as high as 0.6 kHz with use of a 3-MHz transducer. Thus, a cutoff value of 0.8 kHz was chosen to avoid this source of error. Other potential sources of high systolic velocity within the uterus are molar pregnancies and myometrial fibroids (19). In the application of this technique to the evaluation of an intrauterine sac-like structure, only the echogenic rind and the adjacent myometrium should be interrogated. Care should be taken to avoid the inVolume
176
Number
#{149}
2
ternogation of fibroids to avoid possible false-positive results. In addition, when the periphery of the uterus is examined, flow within the uterine artery may be detected. In the first trimester, this signal is characterized by a high systolic velocity with nelatively little diastolic flow. Typically, a small “hump” occurs in the early diastolic portion of the waveform. This signal has an entirely different appearance than the peritrophoblastic waveform and should not be confused with it. The signal characteristics of pentrophoblastic flow reflect the hemodynamics of early placentation and correlate with the histologic examination of products of conception (10). Lacunar spaces, which are the precursons of the intervilbous space, are present within the developing placenta 10 days after conception. The lacunar and intervillous spaces offer little resistance to blood flow and thus produce signals with a large diastolic component. The high systolic velocity reflects the large pressure gradient that exists between the matennal circulation and the developing placenta. At less than 22 days after conception, only low systolic velocities can be detected tnansvaginally (vs 26 days transabdominally). We hypothesize that as the placenta develops, larger and higher-pressure maternal blood vessels are invaded, and a larger volume of blood enters the interviblous space. This development produces signals with higher velocity and banger amplitude. In our study, such signals were first detected transvaginally at 22 days after conception (at 36 days maternal gestational age). With use of the transabdominal approach, such signals were detected as early as 26 days after conception (at 40 days maternal gestational age) and consistently after 29 days post conception (at 43 days matennal gestational age). After initial detection, the peak systolic velocity rises
rapidly
and
plateaus
at a peak
systolic frequency shift of 2-4 kHz (with a 3-MHz probe). Finally, pentrophoblastic signals are often charactenized by such marked spectral broadening that the upper border of the waveform may be difficult to define. This marked disorganization would appear to reflect wide vaniations in velocity between the mdividual blood cells within the intervilbous space. Any use of Doppler US during earby pregnancy
of any possible veloping fetus
requires
bioeffects (20-23).
consideration
on the Diagnostic
deUS
techniques, including Doppler US, have the potential to produce bioeffects through either of two mechanisms: heating on cavitation. The most recent official American Institute
statement of Ultrasound
by
the in
Medicine states that there is “no dinect evidence at the present time that cavitation causes any biologic effects in human
subjects
under
diagnostic
conditions” (24). However, at very high intensity levels, US can clearly induce bioeffects through cavitation, as evidenced by its use in shock-wave bithotnipsy. Furthermore, prolonged exposure to Doppler ultrasound could theoretically produce an increase in fetal temperature sufficient to cause teratogenic effects. Such effects are highly unlikely under typical diagnostic conditions, but any potential risk can be minimized by limiting fetal exposure to Doppler ultrasound (20,23). This goal can be accomplished by limiting the dumation of the examination and by using the lowest acceptable power output setting with compensatory increases in the receiver gain and time gain curve
settings.
Furthermore,
Doppler
examination of the intrauterine contents in early pregnancy should be limited to cases in which there is evidence that the pregnancy, whether intraor extrauterine, is abnormal or cases in which there is strong suspicion that the intrauterine structure is a pseudogestational sac. In such cases, the theoretical risks of possible harm to a normal pregnancy by a few seconds of exposure to Doppler ultrasound are far outweighed by the very real risks of an undiagnosed ectopic pregnancy.
Continued
advances
in
equipment, particularly the development of more sensitive Doppler detection devices and the commercial availability of transvaginal color flow imaging, will further decrease these minimal risks. In summary, our results suggest that differentiation between pseudogestationab sacs and early normal or abnormal intrauterine pregnancies is possible with use of Doppler techniques. Because of the theoretical hazards of exposing normal early pregnancies to Doppler ultrasound, this technique should be used only when there is a strong suspicion of ectopic pregnancy or when specific criteria for an abnormal gestation, whether intraor extrauterine, are met. This technique may provide supplemental information in patients with suspected ectopic pregnancy whose US image demonstrates nonspecific adnexal findings or an intraRadiology
#{149} 363
uterine sac-like structure. The recent development of commercially available transvaginal probes with both pulsed and colon Doppler capabilities should allow further applications of this technique at acceptable intensity levels. U
cniminatory zone. Obstet 8.
9.
2.
Centers for Disease Control. Ectopic pregnancy-United States, 1986. MMWR 1989; 38:481-484. Kadar N, Taylor KJW, Rosenfield AT, Romero R. Combined use of serum hCG and sonography in the diagnosis of ectopic pregnancy.
3.
4.
5.
1987;
Rempen
A.
Ultrasound
7.
364
1983;
Romero Hobbins of adnexa! agnosis Gyneco! Rornero nosis of
10.
11.
141:609-615.
12.
Edwards
M,
Bohm-Valez
17.
M,
Beyler S. Roberts J. Mendelson EB. Transvaginal sonography in the evaluation of early norma! pregnancy: correlation with HCG level. AJR 1989; 153:75-79. deCrespigny LC, Cooper D, McKenna M. Early detection of intrauterine pregnancy with ultrasound. J Ultrasound Med 1988;
Vaginal Med
1988;
in ectopevaluation.
J
#{149} Radiology
14.
7:381-387.
R, Kadar N, Castro D, Jeanty P. JC, DeCherney AH. The value sonographic findings in the diof ectopic pregnancy. Am J Obstet 1988; 158:52-55. R, Kadar N, Jeanty P. eta!. Diagectopic pregnancy: value of the dis-
levels.
Radiology
15.
R. Abnormal by US and
gonadotropin
levels
Radiol-
ogy 1986; 161:854. Dorfman SF. Epidemiology of ectopic pregnancy. Clin Obstet Cynecol 1987; 30:173-180. Loffer FD. The increasing problem of ectopic pregnancies and its impact on patients and physicians. J Reprod Med 1986; 31:74-77.
16.
Marks
WM,
Filly
RA,
Callen
PW,
FC. The decidual cast of ectopic cy: a confusing ultrasonographic ance. Radiology 1979; 133:451-454.
Laing
mester
of pregnancy.
Vol.
18.
19.
Taylor
DE,
26.
KJW.
female
clinics
Doppler
pelvis In:
eds.
Duplex
York:
first
Taylor
tn-
KJW,
Doppler
in diagnostic
New
ultra-
and
ultra-
ultrasound.
Churchill
Livingstone,
1989:93-116. Schaaps JP. Soyeur D. Pulsed Doppler a vaginal probe: necessity, convenience luxury? J Ultrasound Med 1989; 8:315320. Taylor
KJW,
Schwartz
Gestationa! nosis with
PE,
Kohorn
on or
El.
trophoblastic Doppler US.
neoplasia: Radiology
Stewart
Leo
diag1987;
165:445-448. 20.
Harris
CR,
RC. tic 21.
ultrasound US
317. Hill
devices.
CR.
the
CRB. issues
sound 1987;
24.
172:313-
safe
is safe
62:B149.
safety: Radiology
Nicolaides
quality
in diagnos-
1989;
1989;
Ultrasound
(editorial)?
173:304-306. Soothill PW,
Campbell sues are
level
how
Br J Radio!
Sanders
image
Radiology
Ultrasound:
Merritt
FP,
between exposure
(editorial)?
23.
HF,
Relationship
and
pregnanserum cho-
(letter).
EH, in the
sound:
22.
Kadar N, Romero cy: early diagnosis
Dillon sound Strandness
1986;
158:393-396.
13.
rionic
sonography
a prospective
Taylor KJW, Ramos IM, Feyock AL, et a!. Ectopic pregnancy: duplex Doppler evaluation. Radiology 1989; 173:93-97. Nyberg DA, Laing FC, Filly RA. Threatened abortion: sonographic distinction of normal and abnormal sacs. Radiology 1986; 158:397-400. Nyberg DA, Filly RA, Filho DLD, Laing FC, Mahony BS. Abnormal pregnancy: early diagnosis by US and serum chonionic gonadotropin
149:1181-1186.
ic pregnancy:
6.
AJR
Nyberg DA, Mack LA, Harvey D, Wang K. Value of the yolk sac in evaluating early pregnancies. J Ultrasound Med 1988; 7:129-135. Nyberg DA, Mack LA, Jeffrey RB, Laing FC. Endovaginal sonographic evaluation of ectopic pregnancy: a prospective study. AJR
RL,
chorionic gonadotropin 1985; 66:357-360.
7:7-10.
References 1.
Bree
human Gynecol
what
are
1989;
KH,
Rodeck
S. Amniotic fluid and fetal not heated by obstetric ultrascanning. Br J Obstet Gynaecol
CH,
tis-
94:675-677.
American Institute of Ultrasound in Medicine, Bioeffects Committee. Bioeffects considerations ultrasound. 7(suppl):54.
for
the
J Ultrasound
safety
of diagnostic
Med
1988;
pregnanappear-
August
1990