The principal endocrine marker of pregnancy is hCG, which is synthesized by the trophoblast. hCG is a glycoprotein consisting of two subunits: α and β. The α-subunit has significant homology with other glycoprotein hormones, such as follicle-stimulating hormone, luteinizing hormone, and thyroid-stimulating hormone. The β-subunit, on the other hand, is specific to hCG, and antibodies against the β-subunit form the basis for current immunoassays. Current commercial automated β-hCG immunoassays use enzyme fluorometry, enzyme spectrometry, or chemiluminescence methods, as opposed to the traditional radioimmunoassay methods commonly used in the past. An immunoassay for β-hCG can detect levels of hCG as low as 5 IU/L of serum with less than a 0.2% incidence of false-negative results. β-hCG can be detected in maternal serum as early as 7 to 8 days after ovulation or approximately the day after blastocyst implantation.

The quantification of serum β-hCG levels is useful in determining the viability of pregnancy. To optimally use β-hCG data in treating a patient with a problematic pregnancy, one should first have a thorough understanding of the particular assay used. The World Health Organization has established reference standards for β-hCG assays. The Third or Fourth International Standards (Third or Fourth IS) are the most commonly used reference standards used by the available commercial kits of today. These standards are roughly equivalent to the First International Reference Preparation (First IRP) but are quite a bit different from the Second International Standard (Second IS). The Second IS contains about 20% intact hCG and was initially developed for use in hCG bioassays. One international unit of β-hCG based on the First IRP is equal to approximately 0.58 IU of β-hCG using the Second IS. Fortunately, the Second IS has been exhausted and is no longer used. The First IRP, Third IS, and Fourth IS are highly purified preparations that were developed to overcome the deficiencies seen in the use of a heterogeneous standard. This notwithstanding, due to the continuing variation of assay methodologies and β-hCG standards, there remains considerable between-method variation in β-hCG assay results.

Serum hCG concentrations increase in an exponential fashion in early pregnancy. During the period of gestation in which the hCG concentration is less than 10,000 IU/L (First IRP), or about 25 to 30 days postovulation, the time required for doubling of hCG levels remains relatively constant, with a mean of 1.9 days. Kadar and colleagues reported that 87% of women with ectopic pregnancies and 15% of women with normal intrauterine pregnancies could expect to have hCG doubling times of more than 2.7 days when the hCG concentration measured less than 6,000 IU/L. The lower limits of the increase in serum hCG for viable intrauterine pregnancies have been established by Barnhart and colleagues in a large cohort study. In this study, the serial β-hCG titers of 287 women who presented with pain and/or bleeding in the first trimester and were ultimately diagnosed with a viable intrauterine pregnancy were evaluated. For viable intrauterine gestations less than 10 weeks from last menstrual period or those with an initial β-hCG titer less than 5,000 mIU/mL, the investigators noted that the curve generated for serial hCG concentrations best fit a log-linear model. Overall, β-hCG concentrations tended to
double every 2 days. The median rise of hCG after 1 day was 50% and after 2 days was 124%. The slowest or minimal rise for a normal viable intrauterine pregnancy, however, was 24% at 1 day and 53% at 2 days. Interval β-hCG determinations interpreted within the context of several values can therefore be of prognostic significance in the differentiation between normal intrauterine and extrauterine pregnancies. A normal rise in hCG production, however, does not always differentiate an ectopic from a viable intrauterine pregnancy. Shepherd and associates reported that, in their experience, a normal rise in hCG production did not reliably differentiate an ectopic from a viable intrauterine pregnancy in the symptomatic patient. Early ectopic pregnancies can initially secrete appropriate amounts of hCG because of a well-vascularized placental bed.

Serum progesterone levels reflect the production of progesterone by the corpus luteum in early pregnancy. During the first 8 to 10 weeks of gestation, serum progesterone concentrations change little; as pregnancy fails, the levels decrease. Matthews and colleagues reported progesterone levels in 29 patients with ectopic pregnancy using a direct radioimmunoassay that offers results within 4 hours. Patients with normal intrauterine pregnancies had serum progesterone levels greater than 20 ng/mL, and all patients with ectopic pregnancies had progesterone levels less than 15 ng/mL. Yeko and associates proposed that all ectopic pregnancies could be potentially diagnosed at the first emergency visit with a single serum progesterone determination using a discriminatory value of 15 ng/mL. Other authors, however, have demonstrated significant overlap in the serum progesterone concentrations in ectopic and normal intrauterine pregnancies. One large study by Gelder and colleagues reported that 98% of patients with a normal intrauterine pregnancy had progesterone levels greater than 10 ng/mL and that 98% of patients with ectopic pregnancies not associated with ovulation induction had progesterone levels less than 20 ng/mL. Unfortunately, 31% of viable intrauterine pregnancies, 23% of abnormal intrauterine pregnancies, and 51% of ectopic pregnancies in this series had progesterone levels that fell between 10 and 20 ng/mL, which greatly limited the clinical usefulness of the test. Hahlin and colleagues reported that a serum progesterone value of less than 9.4 ng/mL combined with an abnormal hCG increase had a positive predictive value of 1.0 for pathologic pregnancy. In 1992, Stovall and colleagues reported that in a group of more than 1,000 first-trimester pregnant patients, the lowest serum progesterone level associated with a viable pregnancy was 5.1 ng/mL. Therefore, these investigators established the lower cutoff limit of serum progesterone levels of 5 ng/mL; patients below this threshold had a nonviable pregnancy with 100% certainty and therefore underwent curettage. Patients with serum progesterone levels greater than 25 ng/mL had a 97% likelihood of having a viable intrauterine pregnancy in this study.

Pelvic ultrasound has revolutionized the diagnostic process of ectopic pregnancy. Transvaginal ultrasonography, in particular, may identify masses in the adnexa as small as 10 mm in diameter and can provide more detail about the character of the mass than clinical exam (Fig. 34.2). At the same time, transvaginal ultrasonography can evaluate the contents of the endometrial cavity and can document the presence of a viable intrauterine pregnancy with great accuracy. In addition, transvaginal ultrasonography allows for the simultaneous assessment for the presence of free peritoneal fluid.

Transvaginal ultrasonography is usually considered superior to transabdominal ultrasonography in the diagnosis of ectopic pregnancy. Although the latter provides a broader perspective of the abdominal cavity and pelvis, transvaginal ultrasonography generally provides better resolution of the internal female genitalia. A 5-MHz transvaginal transducer allows for a deeper penetration of the pelvis than transducers of higher frequency, whereas a 7.5-MHz transvaginal transducer provides for better near-resolution at the cost of shallower penetration. On rare occasions, an ectopic pregnancy may be located beyond the reach of the transvaginal transducer’s scanning field. On these particular occasions, incorporation of transabdominal ultrasonography may be an important adjunctive step.

Jain and colleagues compared endovaginal and transabdominal ultrasound results in 90 patients with a positive serum pregnancy test (Table 34.3). The specific diagnosis of ectopic pregnancy was impossible using only transabdominal ultrasound before 7 gestational weeks. Normal intrauterine pregnancies could be detected earlier with endovaginal ultrasound because the yolk sac, fetal pole, and fetal heart motion could be seen sooner. Fetal heart motion was detected as early as 34 days after the last menstrual period in patients with identifiable fetal poles at the time the crown-to-rump length was 0.3 cm.

Although diagnosis by transvaginal ultrasound can be quite useful, it may at times be confusing. One problem is that a pseudogestational sac that is due to a decidual cast can be mistaken for an amniotic sac. A useful differentiating feature is the “double-line” image, caused by the faint hypoechoic decidual lining of the uterus and the hyperechogenic rim of the trophoblast surrounding the gestational sac. The double-line image can be seen as early as 5 weeks after the last menstrual period. Even in the presence of the double-line image, however, it is important to further follow the course of pregnancy and subsequently confirm a viable intrauterine pregnancy with the ascertainment of ultrasonographically imaged intrauterine cardiac motion.

Although not always seen, Frates and Laing reported that the presence of a noncystic extraovarian adnexal mass, extrauterine cardiac motion, or a “tubal ring” by transvaginal ultrasonography is highly specific for ectopic pregnancy (98.9%), with a high positive predictive value (96.3%). These
authors described that the direct imaging of the ectopic pregnancy using any of these differentiating features is possible in 84% of cases.

Many authors have reported on correlations between threshold levels of hCG above which an intrauterine gestational sac is expected by ultrasonography in a normal pregnancy (discriminatory zone). Early on, Kadar and colleagues described a threshold level of hCG above which an intrauterine gestational sac was expected by abdominal sonography in a normal pregnancy (discriminatory zone). This threshold hCG level was initially characterized as a titer of 6,500 IU/L or higher using the First IRP. Presently, transvaginal ultrasonography reliably detects intrauterine gestations as early as 1 week after missed menses (β-hCG > 1,500 IU/L; 5 to 6 weeks gestation). Barnhart and associates reported that with a β-hCG concentration of 1,500 IU/L or higher, an empty uterus on transvaginal ultrasonography identified an ectopic pregnancy with 100% accuracy. Even using a discriminatory serum β-hCG concentration of 1,000 IU/L, Cacciatore and associates identified an intrauterine gestation in all intrauterine pregnancies and in none of the ectopic pregnancies. Furthermore, these investigators reported that the detection of an adnexal mass in combination with an empty uterus had a sensitivity of 97%, specificity of 99%, positive predictive value of 98%, and negative predictive value of 98%, provided that serum β-hCG concentrations exceeded 1,000 IU/L. The coupling of hCG titers with transvaginal ultrasonographic findings has therefore greatly facilitated the early diagnosis of ectopic gestation. It must be stressed, however, that considering the variations in β-hCG assays, ultrasound equipment, and sonographer experience, each institution must determine their own discriminatory thresholds for the sonographic detection of an intrauterine pregnancy.

The advent of color-flow Doppler technology may potentially further improve the accuracy of noninvasive diagnostic methods. Kurjak and colleagues reported that ectopic pregnancies are characterized by the identification of peritrophoblastic flow associated with an adnexal mass by color Doppler techniques. Kirchler and coworkers showed that color Doppler qualitative blood flow analyses of the tubal arteries can help localize the side of a tubal ectopic pregnancy. These investigators reported a between-side difference in tubal blood flow of 20% to 45%, with increased blood flow seen on the side of the ectopic pregnancy. Emerson and colleagues demonstrated that color-flow Doppler capability can help differentiate between viable intrauterine pregnancy, completed abortion, incomplete abortion, and ectopic pregnancy by analyzing uterine color Doppler appearance, intrauterine venous flow, the presence or absence of intrauterine peritrophoblastic flow, corpus luteal flow, and the presence or absence of peritrophoblastic flow in the adnexa. Pellerito and associates reported that colorflow imaging increased the sensitivity of detecting an ectopic pregnancy. Of 65 patients with surgically confirmed ectopic pregnancies, 36 (sensitivity 54%) cases were detected by endovaginal sonography alone, whereas 62 (sensitivity 95%) cases were detected by a combination of endovaginal sonography and color-flow imaging.

At one time, the histologic changes in the endometrium that accompany an ectopic gestation were routinely confirmed by dilatation and curettage (D&C). Today, more accurate diagnostic methods—such as the immunoassay for β-hCG, transvaginal ultrasonography, and laparoscopy—exist. In this setting, a D&C may therefore not always be necessary. If, however, the plateau level of β-hCG is low, a D&C can be helpful to establish the presence of degenerating villi, and if a patient is bleeding excessively, a D&C may also be required. In either case, assessment of the removed material and findings of decidua without chorionic villi suggests the diagnosis of ectopic pregnancy. Such findings do not provide absolute proof, however, because they also occur with spontaneous abortion. To distinguish between an ectopic pregnancy and spontaneous abortion, Stovall and colleagues reported that if β-hCG concentrations did not decline ≥15% within 8 to 12 hours following curettage, an ectopic pregnancy was likely.

If available, a frozen section may be obtained immediately after curettage, providing an opportunity to confirm the diagnosis within minutes while the patient is still in the operating room under anesthesia. If no chorionic villi are present, further assessment and treatment by laparoscopy may be considered. In a recent report of 87 consecutive frozen section samples taken from uterine curettings, Spandorfer and colleagues found that 93.1% of these specimens were identified correctly after further analyses of the tissue by permanent section.

The atypical epithelial changes of the gestational endometrium in a case of tubal pregnancy were first described by Polak and Wolfe in 1924, and these changes were further expanded on by Arias-Stella in 1954 (Fig. 34.3). These comprise a highly controversial set of histologic criteria that depend, for accuracy, on the precise definition of the particular cell type involved in the morphologic change, together with ill-defined physiologic events that reportedly produce the changes. Arias-Stella and others were convinced that these histologic changes are a progressive phenomenon resulting from the exaggerated proliferative and secretory endometrial responses to the elevated hormonal levels of pregnancy. Lloyd and Fienberg disagreed, maintaining that these endometrial changes are regressive and involutional and are the result of declining hormonal levels. Whichever hypothesis is ultimately proven, similar endometrial changes may be seen with a normal pregnancy, spontaneous abortion, or ectopic pregnancy. Histologic endometrial criteria, therefore, seem to have limited value in the specific diagnosis of extrauterine pregnancies.

Culdocentesis is a diagnostic tool for identifying the presence of intraperitoneal bleeding. This simple procedure of inserting an 18-gauge spinal needle attached to a 50-mL aspirating syringe into the cul-de-sac between the uterosacral ligaments (Fig. 34.4) provides immediate clinical information when unclotted blood is aspirated from the cul-de-sac. The procedure cannot be used for a definitive diagnosis, of course, because a tubal pregnancy may not have ruptured or leaked into the peritoneal cavity. In addition, a culdocentesis does not provide information concerning whether the blood is from an ectopic pregnancy or from some other cause of intra-abdominal bleeding. The rupture of a corpus luteal hemorrhagic cyst, for instance, may cause a similar bleeding pattern.

The availability of sensitive transvaginal ultrasonographic technology presently limits the usefulness of the culdocentesis procedure, such that it is rarely presently performed. Free intraperitoneal blood has a characteristic ultrasonographic appearance and can be seen in nearly all cases in which a significant intraperitoneal hemorrhage has occurred. In the absence of the immediate availability of transvaginal ultrasonography or in an emergency setting, however, a culdocentesis may still be of potential value.

Laparoscopy remains the gold standard in the detection of ectopic pregnancy, although noninvasive diagnostic methods continue to improve. In addition to permitting the diagnosis of an ectopic pregnancy, it enables surgical treatment. Laparoscopy also provides an opportunity to visualize the entire pelvis and other peritoneal organs. In particular, the condition of the unaffected fallopian tube can be assessed, as well as the presence of pelvic adhesions and endometriosis. This information may be particularly valuable for those patients interested in future fertility. The disadvantage of laparoscopy is that it is an invasive procedure that carries some risk of complications. Using standard methods, it requires general anesthesia and an operating room setting, thereby contributing to increased medical costs. Recent investigators, however, have been exploring the potential of “microlaparoscopy,” in which improved optics and smaller-diameter laparoscopes and trocars allow for a definitive diagnosis and possible treatment in the nonoperating room setting. Several authors have reported
the encouraging use of microlaparoscopy in the office setting, using local rather than general anesthesia. The specific utility of microlaparoscopy, however, for the primary evaluation and treatment of ectopic pregnancy remains to be established.

Laparoscopy may be useful when an ectopic pregnancy is suspected, but no signs of an ultrasonographically visualized extrauterine gestational sac are evident. This includes situations in which there is an inability to visualize an intrauterine gestational sac and serial β-hCG determinations are rising inappropriately. This also includes situations in which a D&C fails to identify products of conception. One must be careful, however, in settings in which the β-hCG determinations are very low or the gestational age is limited. In these settings, the ectopically implanted gestational mass may be so small that it is still not able to be seen at laparoscopy. The clinician and patient might therefore be falsely reassured by negative laparoscopic findings. All patients without a definitive diagnosis established at laparoscopy should continue to be followed closely.

Gleicher et al. described the use of hysterosalpingography and selective salpingography in differentiating early (biochemical) intrauterine from failing intratubal gestations. A characteristic tubal opacification pattern was seen in the cases of early tubal pregnancy. Confino and coworkers reported that selective salpingography was useful in diagnosing early tubal pregnancies in some patients with equivocal clinical, laboratory, and sonographic findings. In addition, these investigators injected a single dose of MTX through the selective salpingography catheter after cannulation of the tubal ostia and identification of a characteristic ampullary radiolucency in seven patients. Each had subsequent complete resolution of the pregnancy without complication. Risquez and colleagues reported the successful visualization of two ectopic pregnancies by transcervical tubal cannulation and falloposcopy. The falloposcope is a microendoscopic instrument 0.5 mm in external diameter that is introduced by a 1-mm coaxial catheter. Although limited by the presence of blood in the tubal lumen, direct visualization of the ectopic pregnancies was accomplished in both cases and confirmed by concurrent laparoscopy. Other investigative teams have explored the potential of other imaging methods, such as magnetic resonance imaging (MRI), in the diagnostic workup for ectopic pregnancy. MRI might be useful if the sonographic image is inconclusive, although it is likely to be rarely needed, particularly if laparoscopy is generally considered in uncertain cases.