32 ECTOPIC PREGNANCY
In the United States, ectopic pregnancy is the leading cause of pregnancy-related deaths during the first trimester.1 In 1992 ectopic pregnancies accounted for approximately 2% of reported pregnancies, and ectopic pregnancy-related deaths accounted for 9% of all pregnancy-related deaths.2 The incidence of ectopic pregnancy has been increasing since the 1970s (Fig. 32-1).3 This increase in ectopic pregnancy diagnosis is multifactorial, owing to both a true increase in the disease, related to increased numbers of patients at risk for the disease (Table 32-1), and earlier diagnosis of cases that otherwise would have spontaneously resolved.
|Risk Factor||Odds Ratio*|
|Previous ectopic pregnancy||8.3|
|In utero exposure to diethylstilbestrol||5.6|
|Use of intrauterine device||4.2–45.0|
|Documented tubal disease||3.8–21.0|
|Previous genital infections||2.5–3.7|
|Multiple sexual partners||2.1|
|Previous pelvic/abdominal surgery||0.9–3.8|
|Early age at first intercourse (<18 years)||1.6|
Reproduced with permission from Pisarka M, Carson SA, Buster JE: Ectopic pregnancy. Lancet 351:1115, 1998. Copyright by The Lancet Ltd.
Before sensitive pregnancy tests and sonography, ectopic pregnancy was frequently a life-threatening diagnosis. In 1970 Breen4 reviewed 654 cases of ectopic pregnancy and found that in 49% of cases, the patients were admitted in shock. Because of this, surgery was performed as soon as the diagnosis was made. Today more sensitive tests for human chorionic gonadotropin (hCG) and the use of high-resolution ultrasonography allow earlier diagnosis of previously unsuspected cases. This has led to a shift from inpatient surgical therapy of ectopic pregnancy to selective outpatient management.3
Currently, sonography and, more specifically, transvaginal sonography is used not only to diagnose ectopic pregnancy in symptomatic patients but to screen women at high risk; to guide decisions as to the most appropriate mode of therapy; and to monitor patients diagnosed with ectopic pregnancy who can be managed without surgery (either expectantly or with methotrexate). Thus, ultrasonography plays a role not only in the early diagnosis but also in our understanding of the natural history of ectopic pregnancy. This chapter illustrates the role of ultrasonography in the diagnosis and management of ectopic pregnancy.
Clinical suspicion of ectopic pregnancy is raised when a patient has a positive pregnancy test, pain, bleeding, and an adnexal mass. However, the clinical triad of pain, bleeding, and adnexal mass is present in only 45% of patients.7 Further complicating matters is that this clinical triad is not specific for ectopic pregnancy. In a study by Schwartz and DiPietro,7 only 14% of patients with clinically suspected ectopic pregnancy actually had this as their final diagnosis. Other diagnoses were symptomatic ovarian cysts (27%), pelvic inflammatory disease (PID) (15%), dysfunctional uterine bleeding (4%), and spontaneous abortions (5%). Obviously, this breakdown of patients will vary with individual hCG levels and risk factors (see Table 32-1).
Neither type of pain nor location of pain is specific for establishing the location of an ectopic pregnancy. In Breen’s4 1970 study, the location of the pain varied as follows: shoulder (11%, all of whom had ruptured ectopic pregnancies), generalized abdominal (13%), lower abdominal (74%), lower quadrant ipsilateral to ectopic (51%), lower quadrant contralateral to ectopic (9%, perhaps secondary to contralateral corpus luteum cyst), back (6%), and vaginal (0.9%).4
Patients typically present at approximately 5 to 6 weeks’ gestational age. However, because menstrual dates are often inaccurate, an early gestational age by dates should not influence the diligence with which one seeks an ectopic pregnancy.
Patients with a history of ectopic pregnancy have a 12.6% chance of having a repeat ectopic pregnancy because risk factors contributing to the first episode are still present, along with further scarring of the fallopian tube as a result of the prior ectopic pregnancy. For patients with a history of PID and ectopic pregnancy, this risk is as high as 27%.8 Other important risk factors are tubal ligation and use of an intrauterine contraceptive device. Both of these methods of contraception effectively prevent pregnancy. However, if pregnancy does occur, an ectopic location should be strongly suspected (Fig. 32-2).9 A risk factor that is increasing in prevalence is that of women undergoing in vitro fertilization. The incidence of tubal pregnancy after oocyte retrieval/embryo transfer is as high as 4.5% (Fig. 32-3).10 This increased rate is the result of risk factors for ectopic pregnancy overlapping those for infertility, use of superovulating agents, which may alter tubal motility, and embryo transfer with retrograde embryo migration into diseased tubes.11,12 Even though identification of women at increased risk for ectopic pregnancy is helpful, because ectopic pregnancy is so common, it should be remembered that “a woman who is capable of conceiving is capable of having a pregnancy in a location other than the uterine cavity.”4
FIGURE 32-2 Ectopic pregnancy in a patient with intrauterine device (IUD). The patient presented with positive human chorionic gonadotropin (hCG) and abdominal pain. A. Transvaginal sagittal view of the uterus shows the IUD in appropriate position. No intrauterine pregnancy is visualized. B. Transvaginal transverse view shows the IUD in the uterine cavity and normal left ovary (calipers) with two cysts. C. Transvaginal sagittal view of the adnexa shows a tubal ring (arrows) with yolk sac. Because IUDs prevent implantation in the uterus, patients who become pregnant with an IUD in place have a high likelihood of ectopic pregnancy. The patient was treated surgically.
(From Dialani V, Levine D: Ectopic pregnancy: A review. Ultrasound Q 20:105, 2004.)
FIGURE 32-3 Transabdominal sonogram in a patient after ovarian hyperstimulation. A. Transabdominal scan shows an enlarged (11 cm) hyperstimulated ovary (calipers) with large cysts. B. Transabdominal image shows the uterus (UT) with a ring-like mass in the left adnexa (arrow). C. Transvaginal scan shows the relationship between the ectopic pregnancy (arrow) with yolk sac and the ovary (O).
The role of quantification of serum beta-hCG correlation with the clinical scenario and sonographic findings cannot be overemphasized in the diagnosis of ectopic pregnancy. A negative hCG essentially excludes the diagnosis of a live pregnancy, although a chronic ectopic pregnancy may be present. Sensitive radioimmunoassays are widely available, which become positive at approximately 23 menstrual days (9 days after conception). This is before the first missed menstrual period and before a gestational sac can be seen with transvaginal sonography.
When reviewing the literature or evaluating guidelines in clinical practice, it is important to know which standard of hCG is being used in the referring clinician’s laboratory. In the 1960s, the Second International Standard (IS) was introduced. Later, a purer standard, the International Reference Preparation (IRP) was introduced and is the main standard in use today. Correlation between the second IS and IRP can be performed by multiplying by a factor of 1.8 (second IS × 1.8 = IRP). Throughout the rest of this chapter, the hCG levels are corrected to IRP.
Early studies on gestational sac visualization evaluated normal early pregnancy and documented the presence of a sac when the hCG was 1000 to 2000 mIU/mL (IRP).13–19 It must be emphasized that the majority of these studies evaluated normal early pregnancies and described an intrauterine gestational sac as any collection of fluid in the endometrial cavity. Small fluid collections of 2 mm (without a decidual reaction) were considered sufficient to describe an early gestational sac. As discussed later in this chapter, this type of fluid collection can be due to the pseudogestational sac of an ectopic pregnancy or a decidual cyst and may not represent an intrauterine pregnancy (IUP). Therefore, a fluid collection without a surrounding hyperechogenic rim is not sufficient to exclude the diagnosis of an ectopic pregnancy. However, these values of hCG are helpful in triaging patients. When hCG is below the discriminatory zone (2000 mIU/mL, IRP) and no intrauterine gestation is visualized, the diagnosis could be an early IUP, a miscarriage, or an ectopic pregnancy, and therefore, close follow-up is indicated, rather than treatment.20 When the hCG value is above the discriminatory zone, one can expect to see an intrauterine gestational sac; however, even without visualization of a sac, there could still be a very early normal IUP.
In a study by Mehta,20 676 patients with symptoms suggestive of ectopic pregnancy were evaluated. Of 128 patients without a definite IUP on sonography, 51 had hCG values of greater than 2000, and 17 of these (33%) had an IUP on follow-up. Technical quality, presence of fibroids, intrauterine contraceptive devices, large hemorrhage, and multiple gestations have been described as factors in nonvisualization of an early gestational sac with respect to a discriminatory level of hCG.14,18,20 However, in Mehta’s20 study, none of these factors could be identified in 9 of 17 cases. The rate of indeterminate scans in this study (19%) was similar to that seen by Braffman et al,15 who used an hCG threshold of 1500 mIU/mL. Braffman et al found fluid collections without an intradecidual sign in 40 of 90 patients. This emphasizes that if an hCG level of 2000 is used to treat patients for ectopic pregnancy, then it is possible those women with early intrauterine pregnancies will be inappropriately treated and will lose an otherwise normal pregnancy. Therefore, it is recommended that if a patient presents with any symptoms suggestive of ectopic pregnancy, for example, bleeding, pain, and a pelvic sonogram that appears normal (without a fluid collection or gestational sac) the report should indicate that “although the pelvic sonogram appears normal, an ectopic pregnancy or a normal IUP too early to be seen sonographically cannot be excluded.”
In the past, most patients suspected to have ectopic pregnancy were treated with laparotomy or laparoscopy. Currently, watchful waiting or treatment with methotrexate can be used for stable patients. However, it should be noted that, in many of these patients, the diagnosis of ectopic pregnancy is not confirmed. If patients are treated merely because of a single elevated hCG value, it is possible that normal early pregnancies will be improperly treated. In clinical practice, the discriminatory zone level of hCG should be taken as a guideline rather than an absolute. This raises the issue of whether the “discriminatory zone” for diagnosis of ectopic pregnancy should be raised. This is problematic in that any increase in the discriminatory zone, while increasing sensitivity for normal early pregnancy, will decrease the already low sensitivity for early ectopic pregnancy. An hCG of 2000 is the level at which a sac will be seen in most, but not all, normal early pregnancies.
The important concept to understand is that there is a gray zone of hCG levels in which sonography is frequently indeterminate. When a patient is in an indeterminate group and is stable and reliable, we recommend serial hCG levels or repeat sonography, or both. Unstable patients should be treated as clinically indicated.
A normal IUP demonstrates an hCG doubling time of 2 days (range, 1.2 to 2.2 days).21 In ectopic pregnancy, this doubling time is increased. If the hCG levels are rising abnormally (<60% increase over at least 48 hours and not steadily declining), the patient is presumed to have an ectopic pregnancy. Even small gestational sacs should grow 0.8 mm per day; therefore, a repeat exam in 2 to 3 days will show a definite change.
Several studies demonstrated that ectopic and abnormal IUPs have a lower mean level of serum progesterone compared with normal pregnancies.23–28 A low progesterone level of less than 5 ng/mL can exclude a normal pregnancy with 99.8% accuracy.22 It has been suggested that this test is most helpful in patients who have an hCG below a discriminatory threshold when sonographic findings are nonspecific.23 However, a prospective study by Condous et al29 showed that a single visit strategy with ultrasound and progesterone is not sufficient to exclude ectopic pregnancy.
Diagnostic laparoscopy is considered the gold standard for the diagnosis of ectopic pregnancy, although this invasive approach has a false-negative rate of 3% to 4% and a false-positive rate of 5%.30 The combined use of serum hCG testing and transvaginal sonography is the current noninvasive approach. Because of more sensitive laboratory tests and improved resolution with ultrasound, we are now able to detect ectopic pregnancies at an earlier stage compared with that possible in the 1970s.30 It is, therefore, possible to have a more elective approach to management in stable patients. In addition, early diagnosis of ectopic pregnancy allows for screening of high-risk patients. Cacciatore, Stenman, and Ylostalo31 in 1994 screened 225 women with risk factors of prior ectopic pregnancy and PID and found that 24% had ectopic pregnancies, 84% of which were diagnosed at first screening (37 days of gestation). They suggested that such early diagnosis prevents tubal rupture, substantial hemorrhage, and the need for emergent care.
If a firm diagnosis is not established at the time of initial scanning, follow-up sonography is important. As mentioned previously, gestational sacs grow at a rate of 0.8 mm per day; therefore, follow-up sonography in 2 to 3 days will demonstrate growth in normal pregnancy. Ectopic pregnancies also grow and hemorrhage, and therefore may become more apparent on follow-up. Among patients diagnosed to have ectopic pregnancy, 5% to 18% are detected or confidently diagnosed only at a repeat ultrasound.31
In the work-up of ectopic pregnancy, the visualization of a normal early pregnancy dramatically reduces the likelihood of an ectopic pregnancy. Quoted risks for heterotopic pregnancies (concurrent pregnancies located both within the uterus and in an ectopic location) range from 1/30,000 to 1/2100 (Fig. 32-4).32–35 The 1/30,000 figure was obtained by multiplying the incidence of ectopic pregnancy and dizygotic twinning.34 However, the incidence of heterotopic pregnancies is increasing, especially in women undergoing ovulation induction, in whom the risk for heterotopic pregnancy is as high as 2.9%. This high incidence results from the high prevalence of tubal damage among in vitro fertilization patients and the use of superovulation and multiple embryo transfer, which predispose patients to the condition.
FIGURE 32-4 Transvaginal sonograms of patient with simultaneous intrauterine and extrauterine gestations. A. Sagittal view of the uterus demonstrates an intrauterine gestational sac with embryo (calipers). B. Transverse view toward the left shows intrauterine gestational sac (UT SAC) and left adnexal sac (L AD SAC). C. Magnified view of the left adnexa shows the extrauterine gestational sac with embryo (calipers) and yolk sac. Both embryos demonstrated cardiac activity. Heterotopic pregnancies are becoming more common with the use of ovulation induction agents. In patients with a history of ovulation induction, increased attention should be paid to the adnexa, even after an intrauterine pregnancy has been identified. This patient was not taking ovulation induction agents; this is a rare case of spontaneous heterotopic pregnancy.
However, in general, the finding of an IUP excludes the diagnosis of ectopic pregnancy, except for patients at high risk. Therefore, attention to the uterine contents and understanding of normal early pregnancy are important in the management of patients at risk for ectopic pregnancy. Because transvaginal sonography allows for earlier diagnosis of an IUP than does transabdominal scanning, it is the preferred modality for evaluation of patients at risk for ectopic pregnancy. However, transabdominal views are important to screen for hemoperitoneum (Fig. 32-5) and to visualize ectopic pregnancies beyond the range of the vaginal probe.
FIGURE 32-5 Tubal ring and hemoperiteum in patient with abdominal pain, vaginal bleeding, and positive pregnancy test. A. Transabdominal sagittal view showing an empty endometrial cavity (arrow) within the uterus (UT). The uterus is surrounded by heterogenous hypoechoic material consistent with hemoperitoneum (H). bl, bladder. B. Transabdominal transverse view shows a tubal ring (arrows). No definite yolk sac or embryo is seen within the ring. The right ovary was seen separately (not shown). C. Transabdominal sagittal view, showing free fluid (FF) in the hepatorenal fossa. This shows the importance of transabdominal scanning.
The earliest sign of IUP is a small fluid collection in the endometrium. Sacs as small as 2 mm can be visualized with transvaginal sonography. However, because of the overlap with the decidual cysts and pseudogestational sacs of ectopic pregnancy, specificity for early IUP can be increased if a thin rim of echogenic material around the sac is visualized (Fig. 32-6). The sac should be eccentric to the endometrial cavity. This intradecidual sign was described by Yeh et al36 in 1986 and appears at 4.5 menstrual weeks (2.5 weeks after conception). However, even among experienced radiologists, the accuracy of this finding is as low as 38%.37,38 Laing et al38 found that the intradecidual sac sign has a sensitivity and specificity of 48% and 66%, respectively. In a study by Bateman et al,14 when sacs were less than 5 mm, only one third had a well-defined echogenic rim. This low sensitivity was due to the large number of indeterminate scans. However, in a study by Chiang et al39 sensitivity and specificity of the intradecidual sign was found to be 60% to 68%, and 97% to 100%, respectively. In this study, Chiang and colleagues showed the importance of seeing the echogenic rim around the sac in more than one plane and assessing for lack of change in the appearance (Fig. 32-7). Owing to the possibility that a decidual cyst could be mistaken for a very early IUP, the prudent course is to follow up symptomatic patients with either hCG or sonography, or both.
FIGURE 32-6 Transvaginal sonogram of very early intrauterine pregnancies. A. Sagittal sonogram demonstrates intradecidual sign, with an echogenic ring formed by chorionic villi (C) located eccentric to the endometrial cavity (EC). B. Transvaginal transverse view of a very early intrauterine pregnancy. A 2-mm sac is present. Although there is the suggestion of a decidual reaction (arrows) because of the echogenic surrounding endometrium, it is not sufficient to represent definitely an early intrauterine pregnancy. For this patient, a follow-up sonogram or serial human chorionic gonadotropin levels (hCGs) should be obtained to guide clinical management if an ectopic pregnancy is suspected.
FIGURE 32-7 Pitfall in use of intradecidual sign. Transverse transvaginal scan in a patient suspected of harboring an ectopic pregnancy. A. Initial image demonstrates a 2-mm sac with mildly echogenic rim (arrows), eccentric to the endometrial stripe (curved arrows) suggestive of an intradecidual sign of normal early pregnancy. B. Image obtained 15 seconds after that in Figure 32-7A shows that the configuration of the fluid collection (arrow) has changed and the echogenic rim is no longer present. Small fluid collections in the endometrial cavity may show such changes. This change in appearance is consistent with blood in the endometrial cavity rather than an intrauterine gestational sac.
The double decidual sac (DSS) sign (Fig. 32-8) is a more reliable indicator of an IUP,40 but it is present slightly later than the intradecidual sign. The DSS sign is caused by the inner rim of chronic villi surrounded by a thin crescent of fluid in the endometrial cavity, which, in turn, is surrounded by the outer echogenic rim of the decidua vera. This sign is highly reliable for the diagnosis of an intrauterine gestational sac. However, the DSS sign does not absolutely exclude a pseudogestational sac.41 The presence of the yolk sac (Fig. 32-9) and the embryo (Fig. 32-10) further increase confidence in the diagnosis of an early IUP.
FIGURE 32-8 Transvaginal sonogram of an early pregnancy. This sonogram demonstrates the double decidual sac sign. The gestational sac is measured with calipers. The inner ring is formed by the echogenic chorionic villi. The outer ring is the deeper layer of decidua vera (D). The separating lucent zone is the more superficial decidua vera.
FIGURE 32-10 A. Sagittal transvaginal sonogram demonstrates an intrauterine gestational sac with an early embryo (calipers) with a crown rump length of 6 mm. B. Transverse view of the left adnexa shows the intrauterine gestational sac (gs) and a thick-walled left adnexal cyst (c). Although adnexal cysts in patients with intrauterine pregnancy should be examined to exclude ectopic pregnancy, once an intrauterine pregnancy is visualized, a cyst such as this most likely represents a corpus luteum.
As mentioned earlier, although the finding of an IUP markedly decreases the risk of ectopic pregnancy, all patients should have evaluation of the adnexa. The importance of this evaluation is emphasized in women who have undergone in vitro fertilization. In these women, even visualization of one ectopic pregnancy should not decrease vigilance to identify other gestations, because a patient with twin ectopic pregnancies with simultaneous live triplet intrauterine pregnancies has been reported.42
One of the difficult tasks in evaluating for ectopic pregnancy is distinguishing an abnormal IUP from the pseudogestational sac of ectopic pregnancy. This frequently is not possible. The diagnostic criteria for abnormal IUP are discussed in detail in Chapter 6. If an intrauterine sac is seen with a diameter exceeding that at which a yolk sac should be seen (i.e., mean sac diameter greater than 10 mm without visualization of a yolk sac) (Fig. 32-11) and the patient’s history is suggestive of ectopic pregnancy, a dilatation and curettage can be performed. If this documents the presence of villi, then a nonviable gestation is present, and the patient’s risk for an extrauterine gestation is effectively eliminated. If villi are not detected, the patient is still at risk for ectopic pregnancy. If the patient is stable, interval change in hCG levels can be monitored.
FIGURE 32-11 Transvaginal scan in a pregnant patient with bleeding. Sagittal view of the uterus shows an irregular fluid (F) collection in the uterus. No double decidual sac sign, yolk sac, or embryo is seen; therefore, an intrauterine pregnancy cannot be diagnosed. The mean sac diameter is 29 mm. Because of the large sac size, it can be concluded that if this is an intrauterine pregnancy, it is not viable (see Chapter 6). Therefore, the uterus can be safely evacuated and the products examined for chorionic villi. If chorionic villi are found, an abnormal intrauterine pregnancy is established and the patient’s risk for extrauterine gestation is effectively eliminated.
The sonographic evaluation of ectopic pregnancy should begin with a transabdominal scan to assess for any masses that might be out of the plane of the vaginal transducer. If the bladder is empty at the time of the initial examination, the examination can begin with transvaginal study; however, it must be emphasized that if the adnexal regions are not well evaluated, a transabdominal study through a full bladder should be performed. Cases have been reported of ectopic pregnancies visualized transabdominally that were missed on vaginal scanning alone,43,44 especially in patients with pelvic masses such as fibroids.44 In some cases in which only vaginal scanning is used, either the ectopic pregnancy will not be identified or the ectopic pregnancy will be misinterpreted as intrauterine. The renal regions are evaluated for free fluid in the abdomen (see Fig. 32-5).
Transvaginal sonography should then be performed. Multiple studies have shown that vaginal sonography provides additional clinically useful information above that available with transabdominal sonography for patients with suspected ectopic pregnancy.43,45–49 The uterine contents are closely evaluated for signs of early IUP. As mentioned earlier, the documentation of an IUP greatly diminishes the likelihood of ectopic pregnancy.
As shown in Figure 32-12, common locations of ectopic pregnancies are 75% to 80% in the ampullary portion of the fallopian tube, 10% in the isthmic portion, 5% in the fimbrial end, 2% to 4% in the interstitial end (sometimes called cornual ectopic), and 0.5% in the ovary.4 Abdominal and cervical pregnancies are very rare.
(Modified from Benson RC: Handbook of Obstetrics & Gynecology, 8th ed. Los Altos, CA, Lange Medical Publications, 1983; and Schoenbaum S, Rosendorf L, Kappelman N, et al: Gray-scale ultrasound in tubal pregnancy. Radiology 127:757, 1978.)
Because the tube is the most common location for ectopic pregnancies, it is important to scan above and below the ovaries and between the uterus and ovaries to exclude an adnexal mass (see Figs. 32-3 and 32-13 to 32-15). The location of the corpus luteum is not helpful in directing the examination because contralateral implantation occurs in up to one third of cases (Fig. 32-16).50 When an adnexal mass is visualized, it should be examined for yolk sac, embryo, and heartbeat (Figs. 32-17 to 32-19). If a heartbeat is visualized, this likely will influence the mode of treatment, because methotrexate has a lower success rate in cases of ectopic pregnancy with cardiac activity. However, finding cardiac activity in an ectopic pregnancy occurs only about 10% of the time. A hyperechogenic ring is a fairly specific finding for ectopic pregnancy; but this is easily confused with a normal corpus luteum cyst (see Figs. 32-16, 32-20, and 32-21).
FIGURE 32-13 Transvaginal sonogram of a patient suspected of harboring an ectopic pregnancy. Transverse view of the left adnexa shows a hypoechoic mass (calipers) adjacent to the left ovary (O). No intrauterine pregnancy was visualized. A patient who fails to demonstrate an intrauterine pregnancy and also demonstrates a mass separate from the ovary is worrisome for ectopic pregnancy. A mass of this type generally represents a hematosalpinx.