The pathogenesis of hydatidiform mole has remained a subject of considerable speculation for many years. The work of Kajii and Ohama,11 however, demonstrated that hydatidiform mole results from the fertilization of an "empty egg," that is, an ovum without any active chromosomal material. The chromosomes of the sperm, finding no chromosomal complement from the ovum, reduplicate themselves, resulting in a 46,XX molar pregnancy. This has also been called a complete mole or classic mole. In complete mole, there is complete lack of fetal development, so there are no identifiable fetal parts or fetal membranes that can be seen in this situation. Complete moles are associated with different degrees of trophoblastic proliferation and may follow either a benign or malignant clinical course. Only about 20% of cases will eventually pursue a malignant course.6
Some cases of hydatidiform mole are found to contain a small complement of fetal structure, such as a placenta with membranes, or even a developed fetus. This is referred to as a partial mole. These cases usually involve some edema of the villi but relatively little trophoblastic proliferation. Hydropic degeneration is present, but with some elements associated with fetal structures, the designation partial mole has been made.12 Although subsequent malignancy has been reported, partial moles are almost always benign.13 In a partial mole, the fetus usually has significant congenital anomalies and a triploid karyotype.14 Two sets of chromosomes are of paternal origin, and the third set is of maternal origin, usually leading to a karyotype of 69 XXY or 69 XXX.
A fetus with a coexisting molar pregnancy can occur. This is much less common than a partial molar pregnancy; however, it can grossly appear similar to the partial mole. This disorder is thought to result from a dizygotic pregnancy, with one fetus resulting from a normal fertilization and the other being a complete molar pregnancy.15 In these patients, a fetus and normal placenta can usually be identified, in contrast to a partial molar pregnancy where a normal placenta is not present.
Hydropic degeneration of the placenta may give a similar sonographic appearance to complete and partial mole, but, histologically, it is not associated with trophoblastic proliferation. The villi and hydropic degeneration of the placenta are swollen and edematous, and thus may resemble abnormal trophoblastic tissue. Microscopic examination will usually reveal the absence of cisternal formation; only hydropic swelling is seen. Hydropic degeneration may be seen in 20% to 40% of placentas from abortuses.6
The most common presenting clinical sign of molar pregnancy is vaginal bleeding, which may be seen in 89% to 97% of cases; in approximately half of the cases, the bleeding may be severe enough to produce anemia.16 A uterus that is enlarged beyond what is expected based on gestational age is considered "classic" for molar pregnancy; however, this is only seen in 33% to 51% of cases.16 In approximately 50% of cases, patients will present with what is easily recognized as a molar pregnancy based on the passage of vesicular tissue via the vagina.17 If there has been significant expulsion of molar tissue before presentation, the uterus may appear normal for size or even small for dates, both clinically and sonographically.
Molar pregnancies should always be considered in the differential diagnosis of a patient presenting with severe preeclampsia before 24 weeks of gestation without underlying renal disease. Although this presentation is considered classic for hydatidiform mole, it is only seen in 6% to 12% of cases of molar pregnancy.16
Theca-lutein cysts are often encountered in patients with molar pregnancies. These are physiological cysts and not neoplasms due to the high hCG levels. The actual incidence of these cysts in association with molar pregnancy has been reported in various series to range from 18% to 37%.18,19 Compared with clinical examination, sonography can more accurately assess the presence or absence of thecalutein cysts (Figures 29-5 and 29-6).20 In one large series, theca-lutein cysts were detected clinically in 10% of patients with molar pregnancy, compared with 37% of patients when examined by sonography.19 The presence or absence of theca-lutein cysts does not seem to be an accurate predictor of later development of an invasive mole or choriocarcinoma.19 A review of a series at the University of Southern California has suggested, however, that patients with theca-lutein cysts have a 2-fold increase of malignant sequelae.21 In most patients who undergo spontaneous resolution of hCG levels, the cysts also regress spontaneously; however, they may occasionally undergo torsion, and therefore require surgical intervention.
Typical sonographic appearance of hydatidiform mole with small vesicular cystic spaces.
Complete hydatidiform mole demonstrating 2 theca-lutein cysts (curved arrows), one superior to the fundus and one in the cul-de-sac.
The laboratory findings for molar gestation are usually diagnostic. Measurement of hCG, specifically the β subunit, is almost always abnormally elevated in molar gestations. The assay is not foolproof, because it can be spuriously elevated in twin gestations or may occasionally fail to show significant elevation in a molar pregnancy.22 In PSTT, the human placental lactogen level will be elevated.
As soon as the diagnosis of molar pregnancy is confirmed, the uterus should be evacuated. This typically involves suction curettage (dilation and evacuation, or D&E). Before evacuation, a chest x-ray should be obtained to exclude metastatic disease that might already be present. The patient should also be evaluated as to her pulmonary status, cardiac status, thyroid status, and to be sure that adequate coagulation factors are present.16 After D&E, serial β-hCGs are obtained weekly to assess the status of any remaining trophoblastic tissue. The level of this glycoprotein hormone should return to normal approximately 10 to 12 weeks after evacuation; however, it may occasionally take longer.23 Recurrence of GTD is increased 10- to 15-fold; thus, patients should be instructed to seek medical attention, and early hCG testing and sonography should be exercised to assess future pregnancies. As previously mentioned, theca-lutein cysts will usually regress after successful treatment of molar pregnancy. Their presence or absence should not be taken as an absolute indication of the presence or activity of residual trophoblastic disease.24
Sonography has an important role in evaluating those patients in whom the β-hCG subsequently rises. It can detect the presence or absence of intrauterine pregnancies that may occur after the initial evacuation. It can also detect the small number of patients with molar gestation who have a coexisting normal pregnancy.25 It may also detect the presence of residual trophoblastic disease and can detect invasion of the myometrium by this residual trophoblastic tissue.3
The sonographic appearance of a hydatidiform mole is quite distinctive.1–2 In most cases, a sonographic pattern arising from molar tissue consists of echogenic intrauterine tissue that is interspersed with numerous punctuate sonolucencies (see Figure 29-5). Irregular sonolucent areas may occur secondary to internal hemorrhage or an area of unoccupied uterine lumen.
The sonographic appearance of a hydatidiform mole differs according to its gestational duration and the size of the hydropic villi.26 For instance, hydatidiform moles that occur from 8 to 12 weeks typically appear as homogeneously echogenic intraluminal tissue, because the villi at this stage have a maximum diameter of 2 mm (see Figure 29-6). As the hydatidiform mole matures to 18 to 20 weeks, the vesicles have a maximum diameter of 10 mm, which is readily delineated as cystic spaces on sonography (Figure 29-7).
Hydatidiform mole with cystic spaces. A: Transabdominal longitudinal scan demonstrating echogenic material (curved arrow) within the enlarged uterus, and irregular cystic region along the lower uterine lumen (straight arrow). B: Semicoronal transvaginal sonogram demonstrating echogenic material (curved arrow) within the uterine lumen, representing trophoblastic tissue and irregular cystic space inferiorly. A myometrial vein is distended (arrowhead).
Complete moles do not have fetal tissue; in contrast, partial molar pregnancy, hydatidiform mole coexistent with fetus, and hydropic degeneration of the placenta are associated with the presence of a fetus or fetal parts. Although it may be difficult to differentiate between a partial molar pregnancy and a complete mole with a coexistent fetus on the basis of sonography, these 2 entities can be differentiated from a complete mole when an identifiable fetus is present.15,18 In addition, the complete mole with a coexistent fetus typically has a fetus with a separate normal placenta as well as the molar mass. This contrasts with a partial mole in which only a portion of the placenta is normal and most of it has a vesicular pattern.
Sonography can also delineate the theca-lutein cyst that enlarges under the influence of high levels of β-hCG elaborated by the trophoblasts. These cysts appear as multiloculated cystic masses that are usually located superior to the uterine fundus or, less commonly, in the cul-de-sac (see Figure 29-6). There are typically thin septa present.
Rarely, there can be massive enlargement of the ovaries with luteinization of several follicles in apparently normal pregnancies. This condition, which is termed hypereactio luteinalis, may be a result of hypersensitivity of the woman's ovary to high circulating levels of hCG (Figure 29-8).
Hypereactio luteinalis. A: Composite obstetric sonogram demonstrating multiloculated cystic mass (arrow) superior to the uterine fundus. B: Same patient demonstrating a second multiloculated cystic mass (arrow) in the cul-de-sac. At surgery, massively enlarged ovaries that contained multiple luteinized cysts were found. (Courtesy of Bill Wilson, MD.)
Invasive trophoblastic disease can be diagnosed with color Doppler sonography by demonstration of enlarged vessels within the myometrium that usually exhibit low-impedance, high-diastolic flow (see Figure 29-1).5
Sonographic Differential Diagnosis
Hydropic degeneration of the placenta associated with incomplete or missed abortions is the most common condition that can simulate the appearance of a molar pregnancy (Figure 29-9). The hydropic areas may be focal or diffuse and appear as anechoic space within the placenta (see Figure 29-9). This is due to the sonographic similarity of a hydropic placenta with marked swelling of the villi to molar tissue. A fetus may or may not be present with hydropic degeneration of the placenta. Serum β-hCG levels are generally lower in hydropic degeneration than in partial or complete moles, probably due to the reduced number of functioning trophoblasts. Hydropic changes in the placenta can be associated with triploidy.27
Choriocarcinoma appearing as an echogenic area (arrow) within myometrium of the anterior corpus.
Technical factors that may be used to improve the ability to distinguish partial from complete moles or hydropic degeneration have been described.3 Specifically, detailed examination of the entire intrauterine contents with transducers that are optimally focused to a particular region within the uterus has been stressed. Transvaginal scanning may be helpful in some cases in which the molar tissue cannot be adequately delineated transabdominally. Using this technique, the typical vesicular texture arising from molar tissue can be correctly distinguished from tissue texture emanating from retained products of conception or leiomyomata.4
Pulsed and color Doppler sonography have been used to assess the presence of invasive trophoblastic disease.28 This technique is discussed in greater detail in the section devoted to invasive trophoblastic diseases.
Occasionally, the sonographic appearance of a uterine leiomyoma may mimic that of a hydatidiform mole. However, as described in Chapter 33 uterine leiomyomata typically have a whorled internal consistency that is distinctly different from the vesicular pattern encountered in the hydatidiform mole. They may also contain areas of hyaline and myxomatous degeneration that can simulate the sonographic appearance of hemorrhage within a hydatidiform mole. We have also encountered some partially solid ovarian tumors that simulate the appearance of a hydatidiform mole. Patients with this type of mass can usually be distinguished from those with molar pregnancies by clinical and laboratory methods because the β-hCG is not elevated in nonpregnant conditions.
Finally, patients with retained products of conception with hemorrhage can simulate the sonographic appearance of molar pregnancies, due to the appearance of echogenic material within the uterus. One is usually not able to demonstrate a vesicular pattern of tissue associated with retained products, however. The hCG levels likewise would not typically be markedly elevated.
Although absolute distinction between the various trophoblastic disorders may not always be possible on the basis of sonography, the sonographic evaluation of these disorders plays a vital role of clinical importance. Specifically, it is known that the malignant potential of a complete mole is greater than that of a partial mole or hydropic degeneration. Thus, the sonographic findings can have a significant clinical impact on the treatment and management of these disorders.