Persistent abnormal bleeding after normal pregnancy, abortion, or ectopic pregnancy should lead to a consideration of the diagnosis of gestational trophoblastic disease (GTD). Pulmonary nodules present on chest radiographs after a normal pregnancy suggest GTD. Beta human chorionic gonadotropin ( β -HCG) levels is elevated in these situations.
Investigation of a young woman with metastatic disease of unknown primary should include a β -HCG level measurement.
The risk of GTN after complete hydatidiform mole (HM) is 15% to 20% and is 1% to 5% after partial HM.
Approximately 50% of cases of GTN occur after molar pregnancy, 25% occur after normal pregnancy, and 25% occur after abortion or ectopic pregnancy.
The major risk factors for molar pregnancy include maternal age (older than 45 and younger than 15 years) and a history of prior HM.
The risk of HM is approximately 1 in 1000 pregnancies in North America.
The risk of a subsequent HM after a primary mole is 1 in 100.
Complete moles are of paternal origin, are diploid, and carry a 15% to 20% risk of GTD sequelae.
Partial moles are of maternal and paternal origin, are triploid, and are rarely (2% to 4%) followed by GTD. They nonetheless require follow-up for potential malignant sequelae, as done for a complete mole.
The monitoring of trophoblastic disease and its follow-up is accomplished by measurement of the β -HCG level.
The diagnosis of a molar pregnancy can be established with ultrasonography and may coexist with a normal pregnancy.
Hydatidiform moles are effectively and safely evacuated from the uterus using suction dilation and curettage.
Medical complications of HM are rare but may include anemia, gestational hypertension before 20 weeks, hyperthyroidism, hyperemesis gravidarum, cardiac failure, and, rarely, pulmonary insufficiency.
Patients are classified into low- or high-risk categories. Low-risk patients are treated with single-agent methotrexate or actinomycin D; high-risk patients receive combination chemotherapy, usually with EMA/CO (etoposide, methotrexate, actinomycin D, cyclophosphamide, vincristine [Oncovin]).
The cure rate for low-risk patients approaches 100%.
Patients with high-risk metastatic GTN are successfully treated with chemotherapy in up to 80% of cases.
Surgery plays an important role in the treatment of placental site trophoblastic tumor (PSTT) and epithelioid trophoblastic tumor (ETT).
PSTT and ETT are both relatively chemoresistant.
Patients treated for GTD should not become pregnant for approximately 6 months after treatment to allow accurate follow-up of β -HCG levels.
Fertility rates and pregnancy outcomes are similar in patients treated for GTD compared with those in the general population.
Patients treated with the EMA/CO regimen have an increased rate of secondary malignancies, particularly hematologic malignancies.
Gestational trophoblastic disease (GTD) is considered one of the most curable gynecologic malignancies. Early disease recognition, effective chemotherapy regimens, and sensitive beta human chorionic gonadotropin ( β -HCG) assays have contributed to the excellent oncologic outcome of these patients. Understanding the disease process cannot be overstated to the general gynecologist, who is usually responsible for the initial diagnosis and management of GTD, as well as the timely referral to gynecologic oncology for further management of gestational trophoblastic neoplasia (GTN). A structured approach to diagnosis and management will result in cure for most patients, even in the setting of advanced disease, without adversely affecting future fertility.
GTD describes a heterogeneous spectrum of diseases of abnormal trophoblastic proliferation ranging from benign to premalignant and malignant, with varying predilections toward local invasion and distant metastasis. Benign trophoblastic lesions include placental site nodule and exaggerated placental reaction. Hydatidiform moles, which include complete hydatidiform mole (CHM) and partial hydatidiform mole (PHM) , are considered premalignant conditions given their potential for invasion. An intermediate lesion arising from chorionic-type intermediate trophoblast, termed atypical placental site nodule (APN), has been described ( ; ), but it is not yet part of the World Health Organization (WHO) classification. APN have been associated with a diagnosis of placental site trophoblastic tumor (PSTT) or epithelioid trophoblastic tumor (ETT) 14% of the time, either concurrently or within months of diagnosis ( ). Gestational trophoblastic neoplasia (GTN) includes four subtypes: invasive mole, choriocarcinoma, PSTT, and ETT ( Box 34.1 ).
In 2000, the International Federation of Gynecology and Obstetrics (FIGO) released a new staging for GTD incorporating the modified WHO Prognostic Scoring System, which has standardized the method for reporting the disease ( ). GTD is classified according to histopathologic, cytogenetic, and clinical features, using the WHO classification of GTD. GTN is diagnosed based on clinical, laboratory, and histologic criteria, and those tumors have a tendency to invade and metastasize.
The incidence of GTD is as common as 1 in 1000 pregnancies, but there are marked regional variations worldwide ( ; ). The precise estimate of the incidence of GTD is difficult to establish as a result of a number of factors, such as low prevalence of the disease, inconsistencies between hospital- and population-based data, and disparity in access to centralized pathology review ( ). Prior estimates based on hospital data overestimated the incidence because deliveries (as opposed to pregnancies) were used in the denominator. Nonetheless, with the introduction of census data, true denominators have added validity to reported incidence rates. Similarly, improvements in central reporting through tumor registries have increased the certainty of case ascertainment. Other factors leading to the improved accuracy of incidence estimations include standardized definitions of GTD variants, improvements in cytogenetics, and recognition of rare variants, such as PSTT and ETT.
The incidence of PHM in the United Kingdom, where all GTD cases are registered in a national database, is 3 in 1000 pregnancies, and that of CHM ranges from 1 to 3 in 1000 pregnancies ( ). A similar rate (2.2 in 1000 pregnancies) was reported in a population-based study from Nova Scotia, Canada. Population-based studies suggest that the incidence of HM is higher in Asia than in North America or Europe ( ; ); however, the incidence of hydatidiform mole (HM) in Asian countries has been decreasing in recent decades, perhaps because of a combination of improved diet, decreased birth rates, and improved measurement of population incidence ( ; ). In a North American cohort, race and ethnicity were strongly associated with the risk of both CHM and PHM. After adjusting for age, Asian women were twice as likely to develop a CHM compared with white women, whereas black and Hispanic women had the lowest risk of CHM ( ).
The increased incidence of GTD in certain ethnic groups has not been explained by looking at genetic traits, cultural factors, or difference in reporting. More recent evidence suggests that women from lower socioeconomic status from various geographic and cultural backgrounds (East Asia, Middle East, North and South America) have an increased risk of HM compared with women from higher socioeconomic statuses in the same regions ( ; ).
Pregnancy occurring at extremes of maternal age (younger than 15 and older than 45 years) is a well-established risk factor for HM, with incidence rates following a J-shaped distribution curve ( ; ). Risk increases after age 35, and a 5- to 10-fold increase is seen in women conceiving after age 40, rising precipitously thereafter. This increase is accounted for by abnormal gametogenesis or abnormal fertilization with advanced maternal age; however, due to the decreased fecundity in this cohort, the overall effect on incidence rates is low.
A reproductive history including HM is another risk factor, increasing the risk in future pregnancies by 5- to 40-fold that of the general population. Subsequent pregnancies have an approximate 1% to 1.7% risk, increasing to 11% to 25% when the number of previous HM is two or more ( ; ). The risk is not affected by changing partners. Patients with recurrent molar pregnancies are also at increased risk for the malignant sequelae of GTN. A history of infertility and of two or more miscarriages is associated with a modest increased risk of both CMH and PHM of 1.9 to 3.2 times the population baseline risk ( ; ).
Dietary risk factor analyses have shown conflicting results. Several case-control studies have shown an increased risk of CHM with decreasing consumption of animal fat and beta-carotene (precursor to vitamin A) ( ). Vitamin A deficiency is more prevalent in countries where the incidence of GTD is higher. Dietary factors may somewhat explain geographic variations in the incidence of CHM; however, other studies detailing food intake have failed to show a decreased incidence with increasing consumption of dietary protein or fat. To date, no association between diet and partial mole has been reported.
Recurrent HM, defined as the occurrence of two or more HMs in the same patient, may be sporadic or familial. A rare autosomal recessive disorder known as familial recurrent HM has been identified on chromosome 19q 13.3-13.4 ( ; ); it affects 1.5% to 9% of women with a previous HM ( ). Affected women have a mutation of the NLRP7 gene and, more rarely, the KHDC3L gene, and they are predisposed to abnormal pregnancies characterized by CHM. Approximately 60% of cases of recurrent HM occur in women who have a mutation in both alleles of the same gene ( ). HMs in women with biallelic mutations are all genetically normal, with chromosomes from each parent (diploid biparental CHM); HMs in women without mutations are heterogeneous, with only a small percentage constituted by diploid biparental CHM and the majority being PHMs and the more common, diploid androgenic CHM type ( ). The mechanism of recurrent HM is through disruption of normal genomic imprinting, with silencing of maternal imprinted genes and preferential expression of paternal imprinted genes ( ). These patients are unlikely going to achieve a normal pregnancy, and egg donation with in vitro fertilization is often necessary.
Histopathology and cytogenetic features
During early embryonic differentiation, trophoblasts are derived from the outer blastocyst layer, with three distinct trophoblasts recognized: cytotrophoblasts, syncytiotrophoblasts, and intermediate trophoblasts. Cytotrophoblasts are the trophoblastic stem cells that differentiate along a villous and extravillous pathway. The villous trophoblast forms the interface between maternal and fetal tissues (chorionic villi) and is composed of cytotrophoblasts and syncytiotrophoblasts. This layer is responsible for molecular exchange across compartments and, in the case of syncytiotrophoblasts, production of the pregnancy-associated hormones β -HCG and human placental lactogen (HPL). Along the extravillous pathway, they differentiate into intermediate trophoblasts in the placental bed at the implantation site. This layer is responsible for establishing the maternal-fetal circulation and infiltrating the decidua, myometrium, and spiral arteries.
In HM, chromosomal abnormalities differentiate the disease, with complete and partial moles having distinct chromosomal profiles ( Figs. 34.1 and 34.2 ). CHMs are completely derived from paternal origin, with greater than 90% having a 46,XX genotype, produced by fertilization of an empty (anuclear) ovum by a single haploid (23,X) sperm, which then duplicates in the ovum. A small percentage of CHMs have a 46,XY or 46,XX genotype, produced by dispermy, in which a 23,X sperm and a 23,Y or 23,X sperm fertilize an empty ovum. Rarely, complete moles may be triploid or aneuploid. The mechanism for production of the empty ovum is unknown.
In contrast, PHMs are derived from paternal and maternal chromosomes, resulting in a triploid genotype. A haploid ovum is fertilized by two haploid spermatozoa, resulting in a triploid gestation 69,XXX, 69,XXY or 69,XYY. Although a triploid karyotype is usually seen in PHM, not all triploid pregnancies show histologic changes consistent with a partial mole. In addition, PHM may present in conjunction with a viable fetus, showing signs of triploidy such as multiple congenital anomalies or severe growth retardation. Conditions that may be confused pathologically with PHM include Beckwith-Wiedemann syndrome, placental angiomatous malformation, twin gestation with complete mole and an existing fetus, early complete mole, and hydropic complete mole.
The histopathologic differences between CHM and PHM are well defined. The gross appearance of CHM may be impressive, with a large volume of grapelike vesicles made up of edematous enlarged villi ( Fig. 34.3 ). Histopathologic characteristics include the following: (1) lack of fetal or embryonic tissues, (2) hydropic (edematous) villi, (3) diffuse trophoblastic hyperplasia, (4) marked atypia of trophoblasts at the implantation site, and (5) absence of trophoblastic stromal inclusions. In comparison, the gross appearance of PHM may only show subtle abnormalities, with generally a smaller volume of hydropic villi and the possible presence of a fetus or fetal tissue. The histopathologic features are the following: (1) presence of fetal or embryonic tissues; (2) less diffuse, focal hydropic swelling of villi; (3) focal trophoblastic hyperplasia; (4) less pronounced trophoblastic atypia at the molar implantation site; and (5) presence of trophoblastic scalloping and stromal inclusions.
Differentiation between CHM and PHM in early first-trimester abortions can be difficult because of less pronounced trophoblastic proliferation and only subtle hydropic swelling of the villi. Absence of the immunohistochemical nuclear stain p57 gene product of CDKN1C (a paternally imprinted, maternally expressed gene) suggests paternal origin and can be used to differentiate between CHM from PHM or nonmolar pregnancies. Further studies such as flow cytometry, ploidy analysis by in situ hybridization, or molecular genotyping have been described to differentiate PHM (triploid) from CHM and nonmolar hydropic abortions. A summary of the genetic and histopathologic differences between CHM and PHM is presented in Table 34.1 .
|Feature||Complete Moles||Partial Moles|
|Fetal or embryonic tissue||Absent||Present|
|Hydatidiform swelling of chronic villi||Diffuse||Focal|
|Trophoblastic stromal inclusions||Absent||Present|
|Karyotype||46,XX; 46,XY||69,XXX, 69,XXY; 69,XYY|
|Persistent human chorionic gonadotropin||15%-20% of cases||1%-5% of cases|
Dramatic presentations of advanced HMs have become less common in the developed world, largely because of the increased use of ultrasonography and improvements in the sensitivity of β -HCG assays, both leading to earlier detection. The average gestational age of diagnosis of CHM today is 9.6 weeks versus 17 weeks in the 1960s. After a delayed menses, CHM typically presents in the first trimester as vaginal bleeding, with or without the passage of molar vesicles. Other classic signs of CHM include a large-for-date uterus, absence of fetal movement, anemia secondary to occult hemorrhage, gestational hypertension before 20 weeks’ gestation, presence of theca lutein cysts, hyperemesis, hyperthyroidism, and respiratory distress from trophoblastic emboli to the lungs.
When uterine enlargement is more than 14 to 16 weeks, 25% of patients will have medical complications related to the high levels of β -HCG commonly seen in CHM and proportional to the volume of trophoblastic hyperplasia. β -HCG is homologous to thyrotropin-releasing hormone, and the β -HCG isoforms seen in CHM may have a greater affinity for the thyrotropin-stimulating hormone receptor than normal β -HCG, causing excessive thyroid stimulation in some patients. Similarly, β -HCG is homologous to luteinizing hormone (LH), the purported mechanism whereby ovarian stimulation leads to the formation of theca lutein cysts in some patients.
Despite the possible medical complications associated with the disease, data from the New England Trophoblastic Disease Center have revealed the changing clinical presentation over time of HM ( Table 34.2 ). This change in clinical presentation of HM was also demonstrated in China and Thailand, suggesting a worldwide phenomenon. Patients are now more likely to present with minimal symptoms; nonetheless, if medical complications are present, the woman should be stabilized, followed by evacuation of the HM as soon as possible.
|Symptom or Sign||1988-1993 |
(N = 74)
(N = 306)
|Size greater than dates||28||51|
PHM usually presents incidentally after histopathologic examination of the products of conception from uterine evacuation of a suspected missed or therapeutic abortion. Medical complications such as gestational hypertension, hyperthyroidism, theca lutein cysts, and respiratory distress are rare with PHM. With a low clinical suspicion for PHM, missed diagnosis is a risk, reflecting the importance of a thorough histopathologic examination of curettage specimens to ensure quality care. It is recommended to measure serum or urine HCG level 4 to 6 weeks after uterine evacuation of a suspected missed or therapeutic abortion if symptoms of pregnancy persist (persistent vaginal bleeding or amenorrhea).
The various symptoms associated with HM, such as vaginal bleeding or a large-for-date uterus, often prompt an ultrasound (US) examination to determine whether a pregnancy is viable. US is the standard imaging modality for the diagnosis of a mole, although CHMs are easier to diagnose by US than PHMs, which are difficult to differentiate from incomplete or missed abortion. A CHM has the appearance on US of multiple hypoechoic foci accompanying an enlarged uterus, the so-called snowstorm appearance ( Fig. 34.4 ). As the molar pregnancy progresses into the second trimester, the anechoic spaces of the molar vesicles become more evident. A transvaginal US may show the interface between molar tissue, endometrium, and dilated vesicles in the first trimester better, but it can worsen vaginal bleeding in the setting of metastatic disease to the vagina. Features suggestive of CHM on US are (1) absence of fetal or embryonic tissue, (2) absence of amniotic fluid, (3) enlarged placenta with multiple cysts, and (4) ovarian theca lutein cysts. Features suggestive of PHM on US are (1) presence of fetal or embryonic tissue, (2) presence of amniotic fluid, (3) abnormal placenta with multiple cysts or increased echogenicity of chorionic villi, (4) increased transverse diameter of gestational sac, and (5) absence of theca lutein cysts ( ).
Although US may be the imaging modality of choice, Fowler and associates reviewed 859 cases of histologically proven HM and have shown that only 44% of cases had a preevacuation US suggesting HM, reinforcing the importance of histologic examination for diagnosis ( ). The accuracy was higher for CHM (79%) than for PHM (29%).
Human chorionic gonadotropin
The anterior pituitary produces a series of glycoproteins that differ only in their beta subunits, including HCG, follicle-stimulating hormone (FSH), LH, and thyroid-stimulating hormone (TSH). Outside of pregnancy, an elevated β -HCG level signifies the following: (1) GTN, (2) nongestational tumors secreting HCG, (3) false positives, and (4) pituitary HCG (secondary to LH elevation and cross-reactivity of assays) ( ).
An unexpectedly elevated β -HCG level during pregnancy may suggest the diagnosis of CHM. β -HCG typically plateaus in pregnancy at approximately 10 weeks’ gestation, with levels peaking at 100,000 IU/L and then falling thereafter. Genest and coworkers found that 46% of patients with CHM managed over a 10-year period had pretreatment β -HCG levels higher than 100,000 IU/L ( ). Conversely, Berkowitz and colleagues reported in one series that PHM presented with an elevated β -HCG greater than 100,000 IU/L in 2 of 30 cases (6%) ( ).
Preevacuation diagnosis of HM allows for optimal treatment planning, but the diagnosis of most patients will be made histologically after surgery. If HM is suspected preoperatively, a chest radiograph should be performed because evacuation may transiently shower the lungs with trophoblastic emboli, complicating the interpretation of a postevacuation chest radiograph. A complete blood cell count (CBC), blood typing with antibody screen, β -HCG level, TSH and creatinine levels, coagulation profile, and liver function testing should also be performed. As the RhD factor is expressed on the trophoblasts, patients who are Rh negative with an Rh positive or Rh unknown partner should be treated with Rho(D) immune globulin after evacuation. The anesthesia team should be made aware of the suspected diagnosis and be ready to manage potential complications such as severe hemorrhage, thyroid storm, and acute respiratory distress from trophoblastic emboli.
Suction dilation and curettage
The preferred method of uterine evacuation of HM is suction dilation and curettage (D&C) under general anesthetic. The cervix is serially dilated and then a 12- to 14-mm suction curette is advanced just past the endocervix into the endometrial canal. After activating the suction device, a solution of crystalloid and oxytocin (20 U/L) is infused to increase uterine tone; this is continued for several hours postoperatively to promote uterine contractility and reduce bleeding. Because of the propensity for heavy bleeding, blood products should be available when the uterine size is greater than 16 weeks’ gestation ( ). A gentle sharp curettage may be performed to complete the procedure. Care must be taken during D&C to avoid perforation of the enlarged soft uterus in HM. The D&C may be performed under intraoperative US guidance to decrease the risk of uterine perforation and to establish when evacuation of all products of conception is complete.
For patients diagnosed with HM preoperatively for whom continued fertility is not an issue, hysterectomy with preservation of the ovaries is a treatment option. The risk of developing GTN is 53% in women older than 40 and up to 60% in women older than 50 ( ), compared with 15% to 20% after CHM and 1% to 5% after PHM, which has been observed in the general population ( ). A meta-analysis showed that hysterectomy in women 40 years of age and older decreases the risk of postmolar GTN (odds ratio [OR], 0.21; confidence interval [CI], 0.11 0.38; P < .00001) and thus the need for chemotherapy. A hysterectomy eliminates the risk of local (myometrial) persistence ( ), but it does not eliminate the risk of distant metastases. Because of high rates of postmolar GTN (30% to 58%), two out of the six studies included in the meta-analysis were not in favor of hysterectomy, but the pooled analysis favored hysterectomy in this age group. As a result of the ongoing risk of postmolar GTN after hysterectomy, continued β -HCG monitoring is necessary ( ).
After surgical evacuation, postmolar GTN, usually in the form of a locally invasive mole, occurs in 15% to 20% of CHM cases (>50% of cases in woman older than 40 years) and only rarely (<5%) after PHM ( ). Fig. 34.5 outlines a treatment algorithm for GTD. If postevacuation follow-up is anticipated to be compromised, patients with high-risk CHM may be considered for treatment with prophylactic chemotherapy. High-risk HM is defined as the presence of one or more of the following: initial serum β -HCG greater than 100,000, uterine size larger than gestational age, theca lutein cysts larger than 6 cm, maternal age older than 40, previous GTD, or the presence of hyperthyroidism or trophoblastic embolization ( ).
A 2017 Cochrane review evaluated the evidence for the effectiveness and safety of prophylactic chemotherapy to prevent GTN after HM ( ). After excluding nonrandomized studies and studies that enrolled patients who already had a GTN diagnosis, three studies were included in the data synthesis. Two of the three studies were found to have a high risk of bias, and overall the quality of the evidence was low to very low. The meta-analysis showed that prophylactic chemotherapy reduced the risk of GTN by approximately two-thirds in all patients with HM, including in the high-risk cohort; however, patients who received prophylactic chemotherapy had a delayed diagnosis of GTN and required more courses of chemotherapy than the average to treat subsequent GTN. Therefore prophylactic chemotherapy is an option in high-risk patients who have limited access to follow-up; however, its routine use is not recommended because of the morbidity associated with treatment, potential delays in diagnosis of invasive GTN and the development of drug resistance, the requirement for surveillance regardless, and the ultimate high cure rates eventually achieved in GTN.
Surveillance after hydatidiform mole evacuation
After evacuation of a HM, surveillance with serial β -HCG serum measurements are required to ensure a timely diagnosis of postmolar malignant GTN (discussed later in the section Gestational Trophoblastic Neoplasia, Clinical Features). Within 48 hours of evacuation, a baseline β -HCG level should be obtained and repeated weekly until the level returns to normal (<5 mIU/mL). Most cases of postmolar GTN occur within 6 months of evacuation; a retrospective study of 6701 women with HM reported that only 3 were subsequently diagnosed with GTN at 13, 23 and 42 months after evacuation ( ). Therefore weekly β -HCG monitoring is recommended after normalization for 2 additional weeks (three normal weekly values), and then monthly for an additional 6 months (GTN, 2018). Prolonged follow-up is not recommended because of the low yield of diagnosis and the increased financial and emotional burden.
Although the minimum conventional period for observation is 6 months, such a long duration of follow-up has been questioned, particularly for women with a narrow window of fertility because of advanced age. Older small scale studies ( ; ; ) had suggested that the risk of persistent GTN is zero after a normal β -HCG in patients with PHM and in patients with CHM in whom the β -HCG normalized within 56 days and stated that prolonged surveillance was not necessary. A large population-based study from the United Kingdom ( ) reported the risk of GTN to be extremely low but not zero (1 in 3195, or <1%) after β -HCG normalization and lower still (1 in 9584) after 6 months of normal β -HCG. In contrast, after CHM, the risk of GTN is 1 in 420 after β -HCG normalization, falling to 1 in 839 after 4 months and 1 in 1677 after 12 months. Patients who had β -HCG normalization within 56 days from evacuation were almost 4 times less likely to develop GTN. The resulting recommendations are that for women who are eager to get pregnant after PHM, surveillance may stop after the first normal β -HCG. Further, after CHM, if the patient wishes to attempt another pregnancy sooner, surveillance may stop 6 months after evacuation when the β -HCG normalizes within 56 days but should continue for 6 months after normalization when the β -HCG is still positive at 56 days.
During this period of surveillance, use of reliable contraception is strongly recommended to ensure that a rise in β -HCG level represents postmolar GTN and not a new pregnancy. IUD use should be avoided until normalization of β -HCG because of the potential risk of uterine perforation. One randomized controlled trial (RCT) has shown that use of the oral contraceptive pill (OCP) results in 50% fewer pregnancies during the surveillance period when compared to barrier methods of contraception ( ). In the past there was concern that OCP use increased the risk of GTN, but two RCTs have shown no association between OCP use during postmolar surveillance and the incidence of GTN ( ).
Prognostic factors associated with the development of GTN have been identified in various reports. As noted, advanced maternal age (older than 40 years) increases the risk of invasive mole ( ), as does a history of HM ( ). Ultrasound findings of uterine invasion may also be predictive of the development of GTN. In a retrospective analysis, Garavaglia and coworkers have shown that the presence of hyperechoic lesions (nodules) within the myometrium or increased signal intensity suggesting hypervascularization at baseline US was associated with an OR of 17.57 for the development of GTN ( P < .001) ( ).
Persistent low levels of β -HCG must be evaluated to rule out false-positive assay results or phantom HCG, a rare finding that is secondary to heterophilic antibodies or proteolytic enzymes that mimic HCG. The diagnosis is made when a serum β -HCG test is positive but a corresponding urine β -HCG test taken at the same time is negative. Alternatively, despite serial dilutions of serum, the test result will usually remain positive if heterophilic antibodies are the cause. Finally, physicians can test against multiple β -HCG assays, when available; heterophilic antibodies may cause a positive result in one test and a negative result in another. The reason for the negative urine β -HCG test is that heterophilic antibodies are large glycoproteins unable to cross the glomeruli and thus are not excreted in the urine. These antibodies—typically derived from exposure to mouse, rabbit, goat, or sheet antigens—are acquired through immunizations or time spent in agricultural settings, and they persist over time.
After a hydatidiform mole, choriocarcinoma, or spontaneous abortion, the persistence of low levels (range, 1 to 212 IU/L) of β -HCG for 3 months or longer with no obvious increase or decrease in the β -HCG level trend along with the absence of clinical or radiologic evidence of GTN is termed quiescent GTD. This process is more common after a CHM but may occur after PHM, invasive mole, or choriocarcinoma and has been identified in patients treated with single-agent or multiagent chemotherapy. It is best described as a premalignant condition given that 25% of these cases progress to GTN-choriocarcinoma over a time frame ranging from 6 months to 10 years. Cole and colleagues have shown that the incorporation of hyperglycosylated HCG (HCG-H), a marker of invasive cytotrophoblasts, will detect 100% of quiescent GTD cases that require no further treatment and 96% of self-resolving HM cases that require ongoing surveillance, differentiating these from GTN-choriocarcinoma cases that require further treatment ( ). This methodology, however, requires validation in a prospective fashion.
In perimenopausal or postmenopausal women, persistent low levels (typically <10 mIU/mL) of β -HCG may be due to increased pituitary gland production, which naturally occurs in both older men and women ( ). The form of β -HCG produced by the pituitary is a sulfated variant, and it is detectable in the laboratory as such [( ). The diagnosis can also be made clinically after the administration of a combined OCP for 3 weeks, which inhibits gonadotropin-releasing hormone (GnRH) production and consequently also production of the β -HCG by the pituitary ( ).
Gestational trophoblastic neoplasia
GTN includes invasive mole/postmolar GTN, choriocarcinoma, PSTT, and ETT.
Histopathology and cytogenetic features
Invasive moles are HMs characterized by syncytiotrophoblast or cytotrophoblast hyperplasia, with the presence of villi. The presence of these villi extending into the myometrium constitutes invasion, hence the name. Most of these tumors, as in HM, are diploid; anaplastic tumors are the exception.
The dominant histologic type in metastatic GTN is gestational choriocarcinoma after an HM or nonmolar pregnancy, which occurs in approximately 1 to 50,000 pregnancies ( ). The characteristic appearance of choriocarcinoma is sheets of anaplastic trophoblastic tissue containing cytotrophoblast and syncytiotrophoblast cells without chorionic villi. These cells invade adjacent tissues, with a propensity for vascular infiltration, necrosis, and hemorrhage. Immunohistochemistry in trophoblast cells shows strong staining with markers for β -HCG, inhibin, and cytokeratin. Primary gonadal (nongestational) choriocarcinomas, a type of ovarian germ cell tumors, can develop without pregnancy, and the estimated incidence is 1 in 369,000,000. They are highly aggressive, secrete β -HCG, and share the same histologic appearance as gestational choriocarcinoma. Nongestational choriocarcinomas are derived from differentiation of malignant germ cells into trophoblastic structures. The absence of paternal DNA within the tumor using DNA analysis differentiates nongestational choriocarcinomas from gestational choriocarcinomas. Furthermore, metastatic GTN must be distinguished from extragonadal germ cell tumors. Those rare tumors originating from midline locations such as the anterior mediastinum and retroperitoneum have no primary tumor in the ovaries but do secrete β -HCG. The fluorescence in situ hybridization (FISH) method for identifying single-nucleotide variants in exons and introns on individual RNA transcripts is used to differentiate gestational choriocarcinoma from extragonadal germ cell tumors by quantifying allelic expression and differentiating maternal from paternal chromosomes.
PSTT is a rare tumor composed almost entirely of intermediate trophoblasts, lacking the syncytiotrophoblasts and cytotrophoblasts seen in other forms of GTD. PSTT has an infiltrative pattern, with nests or sheets of cells invading between myometrial cells and fibers. Compared with choriocarcinoma, PSTT is less at risk for vascular invasion, necrosis, and hemorrhage. Immunohistochemical staining is often positive for HPL, CD146, placental alkaline phosphatase, and, in less than 10% of cases, for β -HCG.
ETT was recognized by the WHO tumor classification in 2003. It is considered a rare variant of PSTT and is also derived from intermediate trophoblasts. Similar to PSTT, these cells are arranged in sheets or nests and form tumor nodules in the myometrium. Immunohistochemical staining is positive for multiple markers, such as cytokeratin and inhibin A.
After surgical evacuation of HM, β -HCG values decrease exponentially, with an expected initial steep decline followed by a slower decrease in β -HCG levels. Symptoms associated with invasive mole include irregular vaginal bleeding, uterine subinvolution, and theca lutein cysts. Most GTN is identified in patients undergoing surveillance after evacuation of HM on the basis of β -HCG criteria, as outlined by FIGO ( Box 34.2 ). These include a plateau in β -HCG values (remaining within ±10% of the previous results) over a 3-week period (four values on days 1, 7, 14, and 21); a rising β -HCG value of 10% or more over a 2-week period (three values on days 1, 7, 14); and a histologic diagnosis of choriocarcinoma or evidence of metastases (clinically or radiologically). Experts in the United Kingdom also start chemotherapy to help with heavy vaginal bleeding requiring transfusion, even if the β -HCG is falling, and in any patients in whom the β -HCG is 20,000 or more 4 weeks after evacuation ( ). A persistently elevated β -HCG 6 months after evacuation is no longer an accepted criteria for treatment because evidence suggests that the β -HCG eventually normalizes in almost all patients left on surveillance ( ).