Gestational Trophoblastic Disease

Diljeet K. Singh

Gestational trophoblastic disease (GTD) represents a spectrum of cellular proliferations arising from the villous trophoblast of the placenta and encompasses 4 clinicopathologic entities: hydatidiform mole (complete and partial), invasive mole, choriocarcinoma (CCA), and placental site trophoblastic tumor (PSTT). The last 3 conditions are associated with more significant clinical sequelae and together comprise the general term gestational trophoblastic neoplasia (GTN). In the absence of GTD, a normal pregnancy involves functioning trophoblast that invades the endometrium and recruits a robust vasculature to develop the placenta, which supports intrauterine fetal development. In healthy trophoblastic tissue, these “cancer simulating” behaviors are highly regulated; however, in GTD, normal control mechanisms fail, leading to invasive, vascular tumors with a tendency to metastasize.¹

Historically, GTD has been associated with significant morbidity and mortality. Hydatidiform moles were typically accompanied by serious bleeding and other medical complications before the development of early detection and effective uterine evacuation in the 1970s. Over the past 50 years, advances in this field have transformed GTN from a high mortality condition to one of the most treatable of all human cancers, with a cure rate exceeding 90%.^2–4 Collaborative global efforts and specialty care centers have promoted the development of highly predictive staging and prognostic scoring systems, which enhance individualization of therapy. Furthermore, several advances in chemotherapy afford ongoing refinement in treatment protocols.^2–6 For women at highest risk of death, the application of multimodal therapy, including chemotherapy, radiation, and surgery, has led to high cure rates while minimizing disease and treatment-related morbidities. In this setting of potentially high cure rates, the onus to identify and appropriately treat GTD falls on the providers entrusted with the primary care of women.

EPIDEMIOLOGY

Key Points

1. The incidence of GTD is approximately 1 per 1000 pregnancies.

2. The most consistently defined risk factors for GTD include extremes of reproductive age and history of prior molar pregnancy. Of all the environmental factors associated with GTD, only low β-carotene and animal fat intake is consistently associated with GTD.

3. Complete hydatidiform molar pregnancy typically results in 1 sperm fertilizing an empty ovum, with subsequent genetic duplication; in contrast, incomplete hydatidiform molar pregnancy typically develops from dispermic fertilization of a normal ovum.

In general, studies conducted in North America, Australia, New Zealand, and Europe indicate the incidence of hydatidiform mole ranges from 0.57 to 1.1 per 1000 pregnancies. In contrast, studies from Southeast Asia and Japan suggest an incidence as high as 2.0 per 1000 pregnancies. As a result of difficulties in obtaining reliable epidemiologic data, it is unclear whether these findings represent a true difference in prevalence or are related to discrepancies between hospital- and population-based data or disparities in the availability of central pathology review.⁷ Further complicating the identification of true incidence is the uncommon diagnosis of GTD and the unreliable documentation of early pregnancy loss. Epidemiologic studies do support wide regional variations in the incidence of hydatidiform moles.⁷ However, attempts to attribute an increased incidence of hydatidiform mole among American Indians, Eskimos, Hispanics, and African Americans as well as various Asian populations to genetic traits, cultural factors, or differences in reporting have been unsuccessful.⁸ Furthermore, some data suggest a decline in the incidence of molar pregnancies, which may be attributed to improved socioeconomic conditions and improvements in diet, which is consistent with studies that show a decreased risk of molar pregnancy with increased consumption of dietary carotene and animal fat.^9,10

The incidence of CCA and placental-site trophoblastic tumor are even less well known, because these lesions are exceedingly uncommon and because of the difficulty in clinically distinguishing postmolar CCA from invasive mole. In Europe and North America, CCA affects approximately 1 in 40,000 pregnancies, with 1 in 160,000 term pregnancies and 1 in 40 hydatidiform moles. In Southeast Asia and Japan, CCA rates are higher, at 9.2 and 3.3 per 40,000 pregnancies, respectively.^8,11 In the United Kingdom, CCA develops in 1 in 50,000 deliveries, and placental-site trophoblastic tumor accounts for approximately 0.2% of cases of GTD.^12,13 The incidence rates of both hydatidiform mole and CCA have declined over the past 30 years in all populations.¹¹

The most consistently documented risk factors for the development of GTD include extremes of reproductive age and history of prior molar pregnancy. Advanced or very young maternal age consistently correlates with higher rates of complete hydatidiform mole. Compared with women aged 21 to 35 years, the risk of complete mole is doubled for those older than 35 years and/or younger than 21 years and is 7.5 times higher for women older than 40 years.¹⁴ A diagnosis of a previous hydatidiform mole confers approximately a 1% risk of repeat molar pregnancy.¹⁵ Although this is 10 times the risk of the general population, most women with history of a molar conception will have normal subsequent pregnancies. If a woman has had more than 2 prior molar pregnancies, the risk for recurrence in latter gestations increases to 15% to 28%, and the risk is not influenced by change of partner.^15–21 Although many possible environmental etiologies for complete mole have been studied, the only consistent association is an inverse relationship between β-carotene and animal fat dietary intake and the incidence of molar pregnancy.^9,10

Risk factors for CCA include prior complete hydatidiform mole, ethnicity, and advanced maternal age. GTN (invasive mole or CCA) follows a complete molar pregnancy in 15% to 20% of cases.^22–24 CCA is approximately 1000 times more likely after a complete mole than after another pregnancy event. Fewer than 5% of partial moles will develop postmolar GTN; metastases occur rarely, and a histopathologic diagnosis of CCA has never been confirmed after a partial mole.^22,25 The risk of CCA is also increased in women of Asian and American Indian descent and among African Americans. Similar to molar pregnancies, the median age of women with CCA is higher than that for normal pregnancies.¹¹

Reproductive factors may play a role in the development of GTD. Women with a history of spontaneous abortion have a 2- to 3-fold increased risk of a molar pregnancy as compared with women without a history of miscarriage.²⁶ Other studies have suggested that women with menarche after 12 years of age, light menstrual flow, and previous use of oral contraceptives are at increased risk for GTN.^27,28

Much of the pathogenesis of GTD is well known. In 90% of cases, complete hydatidiform mole occurs when an ovum without maternal chromosomes or with inactive chromosomes is fertilized by 1 sperm that duplicates its DNA, resulting in a 46, XX androgenetic (entirely paternally derived) karyotype.²⁹ The other 10% of complete moles are 46, XY, or 46, XX, as a result of fertilization of an empty ovum by 2 sperms (dispermy). Although nuclear DNA is entirely paternal, mitochondrial DNA remains maternal in origin.³⁰ In contrast, partial molar pregnancies demonstrate a triploid karyo-type (usually 69, XXY), resulting from the fertilization of an apparently normal ovum by 2 sperms.²²

Evidence that recurrent molar pregnancies occur even in the setting of different male partners suggests a poorly understood role of maternal factors in the development of molar pregnancy.³¹ Some researchers have suggested that ova from older women are more susceptible to abnormal fertilizations than are those from younger women.¹ In addition, there appears to be a relationship between excessive paternal chromosomes and trophoblastic hyperplasia.²⁴

DIAGNOSIS

Key Points

1. Classic presenting signs of molar pregnancy include first trimester vaginal bleeding and uterine size greater than expected for gestational dates.

2. Ultrasonography is the imaging modality of choice when GTD is suspected.

3. Markedly elevated human chorionic gonadotropin levels above those of normal pregnancy are a hallmark of hydatidiform moles.

A high index of suspicion on the part of the general obstetrician-gynecologist is essential to the timely diagnosis of GTD and GTN. Clinical features, ultrasound, and serum and urine tests for human chorionic gonadotropin (hCG) aid in the diagnostic process.

In 80% to 90% of cases, complete hydatidiform mole presents with vaginal bleeding, usually at 6 to 16 weeks of gestation (Table 8-1). Other classic clinical signs suggestive of a diagnosis of molar pregnancy include uterine size greater than expected for gestational dates, hyperemesis gravidarum, and pregnancy-induced hypertension. Because the diagnosis of molar disease has shifted earlier in the pregnancy with increasing application of ultrasound technology, these findings are seen much less frequently.³² To a lesser extent, women with molar disease may also present with pelvic pain due to enlarged theca-lutein cysts and/or clinical signs of hyperthyroidism.

Table 8-1 Differential Diagnosis

Partial moles present slightly later in pregnancy because they do not grow as rapidly as complete moles. Most (90%) present with symptoms of incomplete or missed abortion, and vaginal bleeding occurs in approximately 75% of patients.³³ The other signs and symptoms seen with complete mole, such as excessive uterine enlargement, hyperemesis, pregnancy-induced hypertension, hyperthyroidism, and theca lutein cysts, are significantly less common.³³

Presentation of gestational trophoblastic neoplasia depends on the antecedent pregnancy event, extent of disease, and histopathology. Most often postmolar GTN (invasive mole or CCA) presents as irregular bleeding after evacuation of a hydatidiform mole. Clinical signs of postmolar GTN include an enlarged, irregular uterus and persistent bilateral ovarian enlargement. Occasionally, the diagnosis is made at the time of evacuation when a metastatic vaginal lesion is found. Biopsy of suspected vaginal metastases is discouraged because of the risk of substantial bleeding.³⁴ In patients with postpartum uterine bleeding and subinvolution, the differential should include GTN as well as retained products of conception, endomyometritis, primary or metastatic tumors of other organ systems, or a new pregnancy event (Table 8-1).³⁵

Diagnosis of GTN is often made when metastases induce symptoms. The vascular nature of these lesions can lead to bleeding, including intra-abdominal and/or intracerebral hemorrhage, melena, or hemoptysis. Brain metastases and bleeding from these lesions can cause increased intracranial pressure, leading to headaches, seizures, or hemiplegia. Extensive lung metastases can also cause dyspnea, cough, and chest pain. PSTTs and epithelioid trophoblastic tumors almost always cause irregular uterine bleeding, often distant from a preceding nonmolar gestation, and rarely virilization or nephrotic syndrome. The uterus is usually symmetrically enlarged, and serum hCG levels are only slightly elevated.^36,37

Several diagnostic tests aid in the diagnosis and evaluation of GTD. Ultrasonography has virtually replaced all other means of preoperative diagnosis of both complete and partial mole.^38–40 Characteristic ultrasonographic scans of complete mole show a uterine cavity filled with a heterogeneous mass (snowstorm pattern), without associated fetal development and with theca lutein ovarian cysts, although these features are not always visible in the first trimester⁴¹ (Figure 8-1). Ultra-sonography may also facilitate the early diagnosis of a partial mole by demonstrating focal cystic spaces within the placenta and an increase in the transverse diameter of the gestational sac.³⁹ Although previous work suggested that ultrasound was diagnostic of complete mole in early pregnancy, larger, more recent studies have shown that only 40% to 60% of cases are detected as molar by sonography in routine clinical practice.^39,40,42,43 In addition, 10% of “molar pregnancies” diagnosed by ultrasound were found to be nonmolar hydropic abortions on histologic review.^42,44 In the United Kingdom, the Royal College of Obstetrics and Gynaecologists recommends that all products of conception from non-viable pregnancies should undergo histologic examination irrespective of ultrasonographic findings, and other authors have recommended routine follow-up of hCG after elective termination.^45,46 The American Congress of Obstetrics and Gynecology suggests pathologic evaluation of tissue after spontaneous and therapeutic abortions, although regulations vary by state.

FIGURE 8-1. Pelvic ultrasound of a complete hydatidiform mole with the characteristic vesicular pattern of multiple echoes, holes within the placental mass, and no fetus.

hCG is a disease-specific tumor marker produced by hydatidiform moles and GTNs. hCG assays are readily available, and levels are easily measured quantitatively in both urine and blood. hCG is made up of an α subunit that is shared with pituitary glycoprotein hormones, including thyroid-stimulating hormone and luteinizing hormone (LH) and a β subunit unique to the placenta that confers specificity. Assays that are designed to detect hCG target the β subunit. In healthy pregnancy, hCG is intact and is hyperglycosylated during the first trimester. However, in GTD, many other subtypes of β-hCG can exist, including free β-hCG, β-core, nicked free β, or c-terminal peptide.^47,48 Because the hCG molecules in GTD are more heterogenous and degraded than those in normal pregnancy, an assay that detects all forms of hCG and its multiple fragments should be used to follow patients with GTD. Most institutions currently use rapid, automated, radiolabeled monoclonal antibody “sandwich” assays that measure different mixtures of hCG-related molecules. hCG assays are susceptible to false-positive results, usually caused by cross-reacting heterophile antibodies that are found in 3% to 4% of healthy people.⁴⁹ These cross-reacting heterophile antibodies are present only in serum and do not pass into the urine. In most cases a negative urine hCG test can confirm that the serum value is a false positive, although referral to a specialty laboratory may be required. These so-called phantom hCG results, with levels reported as high as 800 mIU/mL, have led to treatment of healthy patients with unnecessary surgery and chemotherapy.⁵⁰ Additionally, there is some cross-reactivity of hCG with LH, which may lead to falsely elevated low levels of hCG. Measurement of LH to identify this possibility and suppression of LH with oral contraceptive pills prevents this problem.⁵¹

Markedly elevated hCG levels above those of normal pregnancy are a hallmark of hydatidiform moles, and approximately half of patients with complete mole have pre-evacuation hCG levels greater than 100,000 mIU/mL.⁵² However, the differential diagnosis of a significantly elevated hCG level includes the multiple causes of an enlarged placenta, such as misjudged gestational age, multiple gestation, and erythroblastosis fetalis (Table 8-1). Partial molar pregnancies, in contrast, typically do not demonstrate elevated hCG levels; fewer than 10% have hCG levels exceeding 100,000 mIU/mL.³⁸

A clinical diagnosis of postmolar GTN is most often made by the finding of rising or plateauing hCG levels after evacuation of a hydatidiform mole. Chorio-carcinoma is usually diagnosed by the finding of an elevated hCG level, frequently in conjunction with the discovery of metastases, after other pregnancy events. PSTT is commonly associated with only slightly raised hCG levels; human placental lactogen may be elevated in this variant of trophoblastic disease.

When a diagnosis of GTN is suspected, a metastatic work-up should be conducted. Most patients who develop GTN after a molar pregnancy are detected early by hCG monitoring, so detailed investigation is rarely needed. Diagnostic testing should be guided by findings on complete history and physical examination and laboratory studies including complete blood count, serum chemistries including renal and liver functions panels, blood type and antibody screen, and quantitative serum hCG level. Pulmonary metastases are most common, so chest radiography (CXR) is essential.² Chest computed tomography (CT) is not needed when CXR is normal, because discovery of micrometastases, which can be seen in approximately 40% of patients, does not affect outcome.⁵³ However, if lesions are noted on CXR, CT of the chest, abdomen, and pelvis and brain magnetic resonance imaging (MRI) are obtained to exclude more widespread disease such as the brain or liver metastases, which would substantially change management. Pelvic ultrasound or MRI may also be useful in detecting extensive uterine disease for which hysterectomy may be of benefit.

Several procedures are critical in the diagnosis and management of GTD. A diagnosis of GTD is usually confirmed by cervical dilatation and suction curettage of uterine contents. Some patients who do not desire future fertility may elect to undergo primary hysterectomy for evacuation of the molar gestation, with concurrent sterilization. Due to risks of hemorrhage at the time of evacuation, hysterotomy or induction of labor is not recommended.

Repeat curettage after hydatidiform mole evacuation is not recommended unless there is excessive uterine bleeding and radiologic evidence of substantial intracavitary molar tissue, because repeat curettage does not often induce remission or influence treatment and may result in uterine perforation and hemorrhage.^54–56

PATHOLOGY

Key Points

1. Complete hydatidiform moles undergo early and uniform hydatid enlargement of villous trophoblast in the absence of a fetus or embryo.

2. Partial, or incomplete, hydatidiform moles demonstrate identifiable fetal tissue and chorionic villi with focal edema that vary in size and shape.

3. Approximately 10% to 17% of hydatidiform moles result in invasive mole, and approximately 15% of these metastasize; the most common site of meta-static spread is to the lungs or vagina.

4. Choriocarcinoma (CCA) is a malignant disease characterized by abnormal trophoblastic hyperplasia and anaplasia, absence of chorionic villi, hemorrhage, and necrosis, with direct invasion into the myometrium and vascular invasion resulting in spread to distant sites.

Molar pregnancies and gestational trophoblastic neoplasia all originate from the placental trophoblast. Hydatidiform moles and CCA arise from villous trophoblast and PSTT from intermediate trophoblast. Normal trophoblast is composed of cytotrophoblast, syncytiotrophoblast, and intermediate trophoblast, all 3 of which may result in GTD when they proliferate.⁵⁷ Normal syncytiotrophoblast invades the endometrial stroma with implantation of the blastocyst and is the cell type that produces hCG. Cytotrophoblast functions to supply the syncytium with cells in addition to forming outpouchings that become the chorionic villi covering the chorionic sac. The villous chorion adjacent to the endometrium and basalis layer of the endometrium together form the functional placenta for maternal-fetal nutrient and waste exchange. Intermediate trophoblast is located in the villi, the implantation site, and the chorionic sac.

Hydatidiform mole is pathologically characterized by varying degrees of trophoblastic proliferation (both cytotrophoblast and syncytiotrophoblast) and vesicular swelling of placental villi associated with an absent or an abnormal fetus/embryo. There are 2 syndromes of hydatidiform mole, which are distinguished by their clinical behavior, morphology, and genetic make-up. Complete hydatidiform moles undergo early and uniform hydatid enlargement of villi in the absence of a fetus or embryo, the trophoblast is consistently hyper-plastic with varying degrees of atypia, and villous capillaries are absent (Figure 8-2). Partial, or incomplete, hydatidiform moles demonstrate identifiable fetal or embryonic tissue, chorionic villi with focal edema that vary in size and shape, scalloping and prominent stromal trophoblastic inclusions and a functioning villous circulation, and focal trophoblastic hyperplasia with mild atypia only (Figure 8-3). Invasive mole arises from myometrial invasion of a hydatidiform mole via direct extension through tissue or venous channels (Figure 8-4). Approximately 10% to 17% of hydatidi-form moles result in invasive mole, and approximately 15% of these will metastasize; the most common site of metastatic spread is to the lungs or vagina. Invasive moles are most often diagnosed clinically rather than pathologically based on persistent hCG elevation after molar evacuation and are frequently treated with chemotherapy without a histopathologic diagnosis. CCA is a malignant disease characterized by abnormal trophoblastic hyperplasia and anaplasia, absence of chorionic villi, hemorrhage, and necrosis, with direct invasion into the myometrium and vascular invasion resulting in spread to distant sites, including the lungs, brain, liver, pelvis and vagina, kidney, intestines, and spleen (Figure 8-5). CCA has been reported to occur in association with any pregnancy event. Approximately 25% of cases follow abortion or tubal pregnancy, 25% are associated with term or preterm gestation, and the remaining 50% arise from complete moles. Only 2% to 3% of complete moles progress to CCA. PSTT is an extremely rare disease that arises from the placental implantation site and consists predominantly of mono-nuclear intermediate trophoblast without chorionic villi infiltrating in sheets or cords between myometrial fibers (Figure 8-6). PSTT is associated with less vascular invasion, necrosis, and hemorrhage than CCA, and it has a propensity for lymphatic metastasis. Immunohistochemical staining reveals the diffuse presence of cytokeratin and human placental lactogen, whereas hCG is only focal. Cytogenic studies have revealed that PSTTs are more often diploid than aneuploid. Most PSTTs follow nonmolar gestations.³⁷ Epithelioid trophoblastic tumor (ETT) is a rare variant of PSTT that simulates carcinoma. Based on morphologic and histochemical features, it appears to develop from neo-plastic transformation of chorionic-type intermediate trophoblast. Most ETTs present many years after a full-term delivery.³⁶

FIGURE 8-2. Complete hydatidiform mole with hydropic villi, absence of villous blood vessels, proliferation of hyperplastic cytotrophoblast, and syncytiotrophoblast.

FIGURE 8-3. Partial hydatidiform mole with chorionic villi of varying size and shape with focal edema and scalloping, stromal trophoblastic inclusions, and functioning villous circulation, as well as focal trophoblastic hyperplasia.

FIGURE 8-4. Invasive mole with direct extension of molar tissue, including hydropic villi and covering hyperplastic trophoblast, into the myometrium.

FIGURE 8-5. Choriocarcinoma composed of abnormal cytotrophoblast and syncytiotrophoblast with hyperplasia and anaplasia, absence of chorionic villi, hemorrhage, and necrosis.

FIGURE 8-6. Placental-site trophoblastic tumor with sheets of mononuclear intermediate trophoblast cells without chorionic villi infiltrating between myometrial fibers.

Pathologic diagnosis of complete and partial moles is made by examination of curettage specimens. In the setting of unclear diagnosis, additional testing can be helpful. Immunohistologic staining for p57 (a paternally imprinted, maternally expressed gene) can differentiate complete moles (absent immunostaining) from hydropic abortuses and partial moles (positively staining), and flow cytometry can distinguish diploid complete moles from triploid partial moles.^58,59 Additionally, pathologic diagnosis of invasive mole, CCA, PSTT, and ETT can sometimes be made by curettage, biopsy of metastatic lesions, or examination of hysterectomy specimens or placentas. Biopsy of a vaginal lesion suggestive of a GTN is dangerous because of the massive bleeding that can occur.⁶⁰

In 2002, the International Federation of Gynecology and Obstetrics (FIGO) defined criteria for the diagnosis of postmolar disease and adopted a combined anatomic staging and modified World Health Organization (WHO) risk-factor scoring system for GTN (Tables 8-2 and 8-3).⁶¹ The components needed to diagnose postmolar GTN include at least 1 of the following: (1) hCG plateau for 4 consecutive values over 3 weeks, (2) hCG rise of > 10% for 3 values over 2 weeks, (3) hCG persistence 6 months after molar evacuation, (4) histopathologic diagnosis of CCA, or (5) presence of metastatic disease (Figure 8-1). The FIGO stage is designated by a Roman numeral, followed by the modified WHO score designated by an Arabic numeral, separated by a colon. PSTTs and ETTs are classified separately.⁵

Table 8-2 FIGO Anatomic Staging for Gestational Trophoblastic Neoplasia¹

Stay updated, free articles. Join our Telegram channel

Join