Hydatidiform mole: Epidemiology, clinical features, and diagnosis

INTRODUCTION — Hydatidiform mole (HM) was first described by Hippocrates around 400 BCE as "dropsy of the uterus." Since that time, HM (also referred to as molar pregnancy or mole) has been of clinical and research interest. Molar pregnancy is part of a group of diseases classified as gestational trophoblastic disease (GTD), which originate in the placenta and have the potential to locally invade the uterus and metastasize. The pathogenesis of GTD is unique because the maternal tumor arises from gestational rather than maternal tissue [1].

HM is made up of two distinct entities, complete HM and partial HM. These differ on the basis of chromosomal pattern, gross and microscopic histopathology, clinical presentation, and outcome [2-4]. Molar pregnancies, although benign, are considered to be premalignant because they have the potential to develop into a malignancy. Malignant disease is referred to as gestational trophoblastic neoplasia (GTN); the histologic entities included in this group are:

●Invasive mole

●Choriocarcinoma

●Placental site trophoblastic tumor (PSTT)

●Epithelioid trophoblastic tumor (ETT)

The epidemiology, clinical features, and diagnosis of HM will be reviewed here. The pathology of GTD, management of HM, and topics relating to GTN are discussed separately:

●(See "Gestational trophoblastic disease: Pathology".)

●(See "Hydatidiform mole: Treatment and follow-up".)

●(See "Gestational trophoblastic neoplasia: Epidemiology, clinical features, diagnosis, staging, and risk stratification", section on 'Differential diagnosis'.)

●(See "Initial management of low-risk gestational trophoblastic neoplasia".)

●(See "Initial management of high-risk gestational trophoblastic neoplasia".)

●(See "Management of resistant or recurrent gestational trophoblastic neoplasia".)

CLASSIFICATION AND PATHOGENESIS — HM is a premalignant disease. It can be categorized as complete mole or partial mole, which differ by gross morphology, histopathology, karyotype, and risk of malignancy (table 1) [2].

HM occurs after an aberrant fertilization, not after a nonmolar pregnancy (spontaneous or induced abortion, ectopic pregnancy, or preterm or term birth). Molar pregnancies originate in villous trophoblast and are characterized by abnormal chorionic villi with trophoblast hyperplasia as a consequence of overexpression of paternal genes [5,6].

Complete moles are diploid and partial moles are triploid. Complete moles typically produce higher levels of human chorionic gonadotropin (hCG).

The risk of gestational trophoblastic neoplasia (GTN), malignant disease, is higher for complete than partial mole. GTN, usually invasive mole (IM), develops in 15 to 20 percent of patients with complete mole and 1 to 5 percent of patients following partial mole [1,7].

GTN includes IM, choriocarcinoma, placental site trophoblastic tumor, and epithelioid trophoblastic tumor, with the majority being postmolar GTN. These entities represent malignant disease because of the potential for local invasion and metastases. This topic is discussed separately. (See "Gestational trophoblastic neoplasia: Epidemiology, clinical features, diagnosis, staging, and risk stratification", section on 'Classification and clinical behavior'.)

The histopathology of gestational trophoblastic disease (GTD) is discussed in detail separately. (See "Gestational trophoblastic disease: Pathology".)

Complete hydatidiform mole — Approximately 80 percent of complete moles are homozygous 46XX, resulting from duplication of the haploid genome of a single sperm following fertilization of an ovum in which the maternal chromosomes are lost during meiosis, or as a consequence of postzygotic diploidization of a triploid conception [2,5,8]. Approximately 20 percent of complete moles arise by dispermic fertilization of an ovum and may be 46XX or 46XY. However, while complete moles are generally androgenetic because nuclear DNA is paternally derived, mitochondrial DNA is still maternal in origin [9].

Rarely, complete moles are biparental and are associated with an autosomal recessive condition predisposing to molar pregnancy. These patients often have recurrent HMs and may require ovum donation to achieve a livebirth. This is discussed in detail separately. (See "Gestational trophoblastic disease: Pathology", section on 'Complete mole genetics'.)

The marked elevation in hCG associated with a complete mole results in some of the clinical features of this condition. These include ovarian enlargement due to theca lutein cysts, hyperemesis gravidarum, early development of preeclampsia (before 20 weeks of gestation), and hyperthyroidism [10-13]. (See 'Less common or late features' below and 'hCG' below.)

Partial hydatidiform mole — Partial moles are triploid, usually resulting from fertilization of an apparently normal ovum by two sperm or occasionally by a diploid sperm, and may therefore be 69XXX, 69XXY, or 69XYY [2,4]. The existence of diploid partial mole is unlikely, with most reported cases representing misdiagnosed complete moles, hydropic abortions, or twin pregnancies [14].

Partial moles are the only type of GTD that are associated with the presence of a fetus, and fetal cardiac activity may be detected. However, there is a high rate of intrauterine death related to triploidy. Thus, a partial mole is often misdiagnosed as an incomplete or missed abortion, and the correct diagnosis is made only after pathology review of the surgical specimen [15]. (See "Gestational trophoblastic disease: Pathology", section on 'Partial mole genetics'.)

Partial mole results in hCG levels that are generally lower than in complete mole, and thus is less likely to be associated with sequelae of hCG stimulation. (See 'Less common or late features' below and 'hCG' below.)

EPIDEMIOLOGY — The incidence of HM is difficult to establish with certainty because of the low frequency of the disease and regional variation in reported rates [16,17].

The reported incidence of HM varies widely in different regions of the world [16,18-20]. North American and European countries report low or intermediate rates of HM (66 to 121 per 100,000 pregnancies), whereas Latin American, Asian, and Middle Eastern nations report a wide range, including high rates (23 to 1299 per 100,000 pregnancies) [19]. This geographic difference in incidence, however, is based on older studies, and subsequent data suggest that the incidence in Southeast Asia is much lower, and in fact may be similar to that seen in the Western Hemisphere [21-24]. This apparent demographic change likely reflects more accurate reporting of population-based rather than hospital-based data, improvement in diagnosis, as well as significant increases in living standards and health care in resource-limited countries. However, studies continue to report an increased incidence among American Indians, Inuits, Hispanics, and African Americans, and investigations have not been able to attribute this finding to genetic traits, cultural factors, or simply differences in reporting [21,25]. Data from the New England Trophoblastic Disease Center (NETDC) indicate that after adjustment for age and compared with White Americans, complete mole was more common in Asian Americans and less common in Hispanic Americans, but only marginally less common in Black Americans. Compared with White Americans, partial mole was less common in Asian, Hispanic, and Black Americans [26].

Data regarding the relative incidence of complete HM versus partial HM are also problematic because of the discrepancies between hospital-based and population-based data and disparity in availability of expert pathology evaluation. In the United Kingdom, the incidence of complete mole is approximately 1 per 1000 pregnancies, and the incidence of partial mole is 3 per 1000 [27]. A similar ratio has been reported in data from Ireland [18]. A large genetic analysis on the products of conception from over 26,000 consecutive miscarriages using single-nucleotide polymorphism microarray identified paternal triploidy (indicative of partial molar pregnancy) in 638 cases (2.8 percent) and full paternal uniparental disomy (indicative of complete molar pregnancy) in 72 cases (0.3 percent) [28]. Thus, the true incidence of molar pregnancies, especially partial molar pregnancies, is likely much greater than the number of cases identified during routine clinical care. Use of in vitro fertilization techniques does not appear to eliminate the risk of molar pregnancy, as cases of both complete and partial mole can occur even following intracytoplasmic sperm injection [29].

RISK FACTORS — The main risk factors for HM are extremes of maternal age and a history of previous mole [19,30]:

●Prior molar pregnancy – Prior HM predisposes to another molar pregnancy. The risk for repeat molar pregnancy after the first mole is approximately 1 to 1.5 percent (approximately 10 to 15 times the risk for the general population) [20,31,32]. The recurrence rate after two molar pregnancies has been reported to range from 11 to 25 percent [33]. In recurrent mole, one type of molar pregnancy can be followed by the same or a different type in any combination (eg, complete mole after complete mole, complete mole after partial mole).

●Extremes of maternal age – The risk of complete mole is highest at extremes of maternal age (≤15 and >35 years). As an example, a series from the Charing Cross Hospital in London, including 7916 molar pregnancies, reported that the median maternal age was 27 years, and the highest risk of molar pregnancy was ≤15 and ≥45 years [31]. A case-control study found that the risk for complete mole was increased twofold for patients >35 years and 7.5-fold for patients >40 years [31,34]. These data suggest that ova from older patients may be more susceptible to abnormal fertilizations that result in HM. In addition, patients age 40 or older with HM may have a higher rate of severe complications. Most cases of HM, however, still occur in patients under age 35 because of the greater number of pregnancies in this age group [20,31,32]. The risk for molar pregnancy and its relation to maternal age is much greater in complete mole than partial mole [31,35].

The risk for both complete mole and partial mole is increased in patients with a history of prior spontaneous abortion and infertility. Compared with patients with no previous history of miscarriage, the risk for complete mole and partial mole was 3.1 and 1.9, respectively, among patients with two or more prior miscarriages. Difficulty in conception or infertility problems was associated with an odds risk of 2.4 and 3.2, respectively, of complete mole and partial mole [36-38].

Dietary factors may also play a role in and may partly explain the regional variations in the incidence of complete mole [36,39]. Case-control studies have observed that the risk for complete mole progressively increased with decreasing levels of consumption of dietary carotene (vitamin A precursor) and animal fat. Global regions with a high incidence of vitamin A deficiency correspond to areas with a high frequency of complete molar pregnancy. However, partial mole has not been related to dietary factors [38,40].

CLINICAL FEATURES — Patients with HM typically present to their obstetric clinician with missed menstrual periods, a positive pregnancy test, and signs and symptoms consistent with early pregnancy or early pregnancy complications (bleeding, pelvic discomfort, hyperemesis gravidarum) [7,41-43]. Molar pregnancy may be suspected based on unusually high human chorionic gonadotropin (hCG) levels or only after pathology evaluation of a failed pregnancy.

In many cases, particularly in partial mole, a patient is presumed to have a spontaneous abortion, and a molar pregnancy is detected only after pathology evaluation of a uterine curettage sample. As an example, in a series of 167 patients with partial mole, clinical diagnosis of molar pregnancy was made in only 41 percent of cases [15]. If the products of conception are not examined histologically, the diagnosis of HM may be missed. In these patients, hCG monitoring to detect persistent disease is generally omitted and early diagnosis of gestational trophoblastic neoplasia (GTN) can be delayed.

Alternatively, patients with early disease may be diagnosed before they develop symptoms, based on an unusually high serum hCG or ultrasound findings (40 percent of patients with complete mole in one study) [44].

The clinical presentation of complete HM began to change substantially in the 1980s. Prior to this time, complete mole typically presented with late effects (hyperthyroidism, preeclampsia). The availability of both ultrasonography and sensitive quantitative measurement of serum hCG has led to earlier diagnosis of complete mole [45-49]. Globally, complete mole is now diagnosed more frequently in the first rather than the second trimester. Thus, there are fewer late onset symptoms. However, there has been no change in risk for postmolar GTN. As an example, in one study, the average age at diagnosis was 11.8 weeks of gestation from 1988 to 1993, compared with 16.5 weeks from 1965 to 1975 [41]. Despite the earlier diagnosis, the incidence of postmolar GTN did not change in the later cohort; 15 of 64 (23 percent) patients with complete mole who did not receive chemoprophylaxis developed postmolar persistent tumor. Even when complete mole is diagnosed before 10 weeks gestation, there is no decrease in the risk of GTN [50].

Common features — Vaginal bleeding, pelvic pressure or pain, an enlarged uterus, and hyperemesis gravidarum are common presenting symptoms. However, these are also common in early pregnancy, and thus, clinicians most often initially suspect a common pregnancy complication or an abnormal nonmolar pregnancy (spontaneous abortion or ectopic pregnancy) rather than HM.

Vaginal bleeding — Vaginal bleeding, which results from separation of the molar villi from the underlying decidua, is common (occurring in 84 percent of patients in one study [41]).

In complete mole, the classic appearance of the blood is characterized as a "prune juice" discharge. This appearance is due to accumulated blood in the uterine cavity, which has undergone oxidation and liquefaction.

Historically, when complete mole typically presented at later gestational ages, vaginal bleeding was often heavy and prolonged and resulted in anemia. Anemia is now uncommon. At early gestational ages, the timing, volume, or pattern of bleeding in HM does not differ from other pregnancy-associated etiologies (spontaneous abortion, ectopic pregnancy). Based on data from 1994 to 2013, vaginal bleeding occurs in only 46 percent of patients [50].

The differential diagnosis of bleeding in pregnancy is discussed in detail separately. (See "Evaluation and differential diagnosis of vaginal bleeding before 20 weeks of gestation".)

Pelvic pressure or pain — Similar to vaginal bleeding, pelvic pain and pressure are common, nonspecific symptoms in molar pregnancy due to the enlarging uterus and/or enlarged cystic ovaries.

Uterine size greater than gestational age — Pelvic examination or ultrasound may reveal a uterine size larger than the estimated gestational age. In current practice, this finding was present in only 21 of 74 (28 percent) patients in one series [41]. Uterine enlargement is a nonspecific finding and may also be present for other reasons (eg, incorrect estimate of gestational age, multiple gestation, uterine leiomyomas).

Uterine enlargement is more likely with a complete mole than partial mole. Although in complete mole there is no fetus, the enlargement is due to both large volumes of molar tissue and retained blood and is usually associated with hCG levels >100,000 mIU/mL. In fact, patients with a partial mole may have a uterine size that is small for gestational age, due to slow growth of a fetus with triploidy.

Hyperemesis gravidarum — Hyperemesis gravidarum was present in approximately 8 percent of patients in one study [41]. Nausea and vomiting may develop earlier than in a nonmolar pregnancy and/or be more severe. However, early or severe symptoms may also occur in a normal pregnancy, particularly in multiple gestation. (See "Nausea and vomiting of pregnancy: Clinical findings and evaluation".)

Less common or late features — HM may also be associated with other clinical features. These are less common and some develop only later in the course if the molar pregnancy presents in the second trimester.

In complete mole, sequelae associated with the high hCG levels include theca lutein cysts, hyperthyroidism, and early onset of preeclampsia. For complete mole diagnosed in the first trimester, the risk of complications such as preeclampsia or significant hyperthyroidism is low [41,46,50,51]. These complications occur in approximately 25 percent of patients with uterine size greater than 14 to 16 weeks of gestation [52].

hCG levels are typically far lower in partial mole. Thus, partial mole is less likely to be associated with sequelae of hCG stimulation [3,53]. As an example, a review of 81 patients with a partial HM found that excessive uterine size occurred in only three patients (4 percent), and hyperemesis, hyperthyroidism, theca lutein cysts, and respiratory distress were observed in only one patient [54]. Only 2 of 30 (6.7 percent) patients had serum hCG concentrations >100,000 mIU/mL. Furthermore, in current practice, partial mole is now primarily evacuated in the first trimester at a median gestational age of 12 weeks [55].

Hyperthyroidism — Clinical hyperthyroidism is present in HM primarily at later gestational ages [41,56]. The development of hyperthyroidism requires the elevation of hCG >100,000 mIU/mL for several weeks. These patients may present with tachycardia, warm skin, and tremor. Laboratory evidence of hyperthyroidism is commonly detected in asymptomatic patients with HM [11,57]. Hyperthyroidism may be accompanied by thyroid storm when the patient undergoes anesthesia, due to the release of catecholamines from the adrenal gland. When pre-evacuation hCG levels exceed 400,000 mIU/mL, administration of beta blockers and antithyroid drugs before complete molar evacuation may avert thyroid storm during or after surgery [57].

The source of the thyrotropic factor in patients with complete mole remains elusive. Some investigators have suggested that hCG is the thyroid stimulator because of the homology between hCG and thyroid stimulation hormone and because positive correlations have been observed between serum hCG and total T4 or T3 concentrations [58]. However, one study measured thyroid function in 47 patients with a complete mole and reported no correlation between serum hCG levels and the serum-free T4 index or free T3 index [59]. Although some investigators have speculated about a separate chorionic thyrotropin, this substance has not yet been isolated.

Ovarian theca lutein cysts — Theca lutein cysts are a form of ovarian hyperstimulation resulting from high circulating levels of hCG and prolactin [10,60]. These cysts are multiloculated, often bilateral, and resolve a few weeks or months after treatment of HM.

Bilateral theca lutein ovarian cysts may be noted on pelvic examination or pelvic ultrasound. When the uterus is markedly enlarged, theca lutein cysts are difficult to palpate and are usually detected by ultrasonography.

Theca lutein cysts may develop in infertile patients treated with ovulation induction, patients with a multiple gestation, or in patients with a normal pregnancy but a hypersensitivity to hCG (see "Adnexal mass: Differential diagnosis", section on 'Theca lutein cysts'). In other patients, a finding of theca lutein cysts should raise a high suspicion of gestational trophoblastic disease.

In practice these cysts can be observed and do not require surgical intervention unless they are large, causing symptoms, or concern for torsion. (See "Hydatidiform mole: Treatment and follow-up", section on 'Ovarian theca lutein cysts'.)

Preeclampsia <20 weeks of gestation — Preeclampsia (pregnancy-induced hypertension and proteinuria or end-organ dysfunction) typically develops after 34 weeks of a normal gestation. When it occurs at <20 weeks, complete mole should be suspected.

In current practice, few cases of HM are associated with preeclampsia (1 to 3.5 percent at 8.5 weeks of gestation) [41,42,46,51]. Historically, when complete mole typically presented at later gestational ages, preeclampsia was a more common presentation. As an example, in a series of patients with complete mole from 1988 to 1993, preeclampsia was present in only 1 of 74 patients [41]. Previously at our center, it was present in 27 percent of patients with complete mole. Eclampsia in molar pregnancy is rare and only anecdotal cases have been reported, usually when a fetus coexists [41,61,62]. (See "Preeclampsia: Clinical features and diagnosis", section on 'Onset <20 weeks'.)

Other features — Patients with HM may also develop:

●Anemia – HM may be associated with significant bleeding at later gestational ages, resulting in iron deficiency anemia. Anemia is uncommon when HM is diagnosed in the first trimester [41].

●Passage of hydropic vesicles from the vagina may occur spontaneously, usually in patients with rapidly enlarging complete moles.

DIAGNOSTIC EVALUATION — The possibility of HM should be considered in any reproductive-age female with abnormal vaginal bleeding. A quantitative serum human chorionic gonadotropin (hCG) level should be obtained and, if elevated, ultrasound examination should be performed. A nonmolar pregnancy (normal intrauterine pregnancy, ectopic pregnancy, spontaneous abortion) is the most likely diagnosis. If these types of gestations are excluded, particularly in patients with a history of prior HM, HM is a likely diagnosis.

History — An obstetric history is taken, including a history of prior gestational trophoblastic disease. The obstetric history should include the dates, duration, and outcome of all pregnancies.

A medical, gynecologic, and surgical history should also be taken.

The patient should be asked about symptoms associated with HM. (See 'Clinical features' above.)

If the clinical presentation suggests HM, malignant disease, gestational trophoblastic neoplasia (GTN), must be excluded. Patients should be asked about symptoms of metastatic disease. The lungs (symptoms include dyspnea, hemoptysis, or chest pain) and vagina (vaginal bleeding) are the most common metastatic sites, but other potential sites include the central nervous system or liver. (See "Gestational trophoblastic neoplasia: Epidemiology, clinical features, diagnosis, staging, and risk stratification", section on 'Symptoms of metastases'.)

Physical examination — A complete pelvic examination should be performed. On speculum examination, the vagina should be examined for metastases. (See "Gestational trophoblastic neoplasia: Epidemiology, clinical features, diagnosis, staging, and risk stratification", section on 'Physical examination'.)

On bimanual examination, the uterus in patients with complete mole may be larger than the expected gestational age. (See "Clinical manifestations and diagnosis of early pregnancy", section on 'Findings on physical examination' and 'Uterine size greater than gestational age' above.)

Bilateral adnexal masses may be present if ovarian theca lutein cysts have developed due to hCG stimulation. A unilateral adnexal mass suggests another type of adnexal abnormality.

A focused general physical examination should be performed to exclude metastases. In GTN, common sites of metastases include the lungs, vagina, liver, and central nervous system. (See "Gestational trophoblastic neoplasia: Epidemiology, clinical features, diagnosis, staging, and risk stratification", section on 'Metastatic sites'.)

Laboratory evaluation

hCG — A quantitative serum hCG should be measured. The serum hCG concentration in patients with HM is usually higher than that observed with singleton intrauterine or ectopic pregnancies of the same gestational age.

If the initial hCG level is high (>100,000 mIU/mL), a transvaginal ultrasound should be performed and will likely demonstrate molar disease if present. If the hCG level is high and the ultrasound shows an apparently normal singleton gestation, the ultrasound and hCG should be repeated in one week to exclude the possible presence of a twin conception with normal fetus and coexistent molar pregnancy.

Markedly elevated hCG levels are more commonly seen in patients with complete mole in contrast to partial mole [6]. As an example, in a series of 153 cases of complete mole managed at the New England Trophoblastic Disease Center (NETDC) between 1980 and 1990, pre-evacuation hCG levels were >100,000 mIU/mL in 46 percent of the patients [63]. Similarly, another study reported that 30 of 74 patients (41 percent) with complete mole had pre-evacuation hCG values >100,000 mIU/mL [64]. In contrast, only 2 of 30 patients (6 percent) with partial mole managed at the NETDC had pre-evacuation hCG levels >100,000 mIU/mL [54]. Another study of 25 patients reported that only 1 of 17 patients (6 percent) with partial mole had a pre-evacuation urinary hCG level >300,000 mIU/mL [53].

The characteristics of hCG assays and etiologies of persistently low or false-positive hCG tests are discussed in detail separately. (See "Human chorionic gonadotropin: Biochemistry and measurement in pregnancy and disease".)

Blood type and antibody screen — Guidelines and recommendations for RhD testing and anti-D immune globulin administration are variable and changing (see "Overview of pregnancy termination", section on 'Alloimmunization prevention'). Given the lack of any specific data of alloimmunization prevention in molar pregnancies and the difficulties of differentiating between early complete moles from partial moles or spontaneous miscarriages, our practice is to check RhD status in all patients with molar pregnancy and administer anti-D immune globulin to Rh-D negative patients. However, future guidance from organizations such as the American College of Obstetricians and Gynecologists or other professional societies may change this practice. (See "RhD alloimmunization: Prevention in pregnant and postpartum patients", section on 'Selective prophylaxis for pregnancy complications associated with fetomaternal bleeding'.)

Other tests — Further testing depends on the likelihood of complications and the clinical assessment. For molar pregnancies that present in the first trimester, the risk of complications such as preeclampsia or significant hyperthyroidism is low [46,51,65]. For patients who present in the second trimester, there is a higher risk of these complications. In such cases, testing should include: complete blood count, renal panel, liver function tests, urine protein, and thyroid function tests. (See "Preeclampsia: Clinical features and diagnosis" and "Diagnosis of hyperthyroidism", section on 'Laboratory tests'.)

If significant bleeding has occurred, a complete blood count should be ordered.

Pelvic ultrasound — If molar pregnancy is suspected, a transvaginal ultrasound should be performed. However, the ultrasound may be indeterminate. In particular, partial molar pregnancies are frequently misdiagnosed as incomplete or missed abortion.

The diagnostic performance of ultrasound appears to be better for complete than partial mole and at later rather than earlier gestations. The mean gestational age of complete mole evaluated by ultrasound at our center from 1988 to 1996 was 8.7 weeks [66].

Ultrasound is preferred to other imaging studies (eg, computed tomography, magnetic resonance imaging) because it provides high resolution images of the female reproductive tract and is cost effective.

Complete mole — Sonographic features suggestive of a complete mole include [66-68]:

●Absence of an embryo or fetus.

●Absence of amniotic fluid.

●Central heterogeneous mass with numerous discrete anechoic spaces – This has classically been described as a "snowstorm or Swiss cheese pattern" on older ultrasounds (image 1A-B). The small cystic areas correspond to the hydropic villi, which are one of the characteristic gross and microscopic pathological findings. The sonographic appearance of complete mole is variable. There may be a large, central fluid collection that mimics an anembryonic gestation (missed or incomplete abortion) or a central mass of variable echogenicity without anechoic spaces, presumably because the hydropic villi are too small to visualize sonographically.

●Ovarian theca lutein cysts (image 2).

These features may be absent in early first trimester mole, and a diagnosis of complete mole may be missed on ultrasound. Estimates of the sensitivity of pelvic ultrasound for the diagnosis of complete mole range from 70 to 90 percent; detection rate appears to increase with increasing gestational age [66,67,69,70]. A series of 1053 patients reported that an ultrasound finding of a molar pregnancy had a specificity of 74 percent for the diagnosis of either complete or partial HM [67]. In some cases, ultrasound findings suggestive of molar pregnancy are ultimately found to be a nonmolar hydropic abortion. The specificity of the sonographic findings in complete mole can be increased by correlating the hCG level. For example, a complete mole can be differentiated from a missed abortion by considering the hCG value. An hCG level >100,000 mIU/mL in the first trimester should raise suspicion of molar pregnancy [71].

Partial mole — Based on ultrasound findings, a partial mole is diagnosed as a missed or incomplete abortion in 15 to 60 percent of cases [67,68,72]. These misdiagnoses are more common for partial mole than for complete mole because only partial moles are accompanied by a fetus and amniotic fluid.

Sonographic features suggestive of a partial molar pregnancy include (image 3A-B) [72-74]:

●A fetus may be identified, may be viable, and is often growth restricted.

●Amniotic fluid is present, but the volume may be reduced.

●Placenta with one or more abnormal findings – Enlarged, cystic spaces ("Swiss cheese pattern") and/or increased echogenicity of chorionic villi.

●Increased transverse diameter of the gestational sac – These changes in the shape of the gestational sac may be part of the embryopathy of triploidy.

●Theca lutein cysts are usually absent.

In one study of sonograms of 22 partial moles and 33 first trimester missed abortions, three radiologists independently evaluated the appearance of the placenta, gestational sac, and the presence or absence of ovarian cysts and found that two sonographic findings were associated with the diagnosis of partial mole: (1) cystic spaces in the placenta and (2) ratio of transverse to anteroposterior dimension of the gestational sac >1.5 [72]. There was high inter-observer correlation for both criteria. When both criteria were present, the positive predictive value for partial mole was 87 percent. When both criteria were absent, the positive predictive value for missed abortion was 90 percent.

When a fetus is present, it can be particularly challenging to distinguish partial moles from a complete mole with a normal twin. (See 'Multiple gestation' below.)

Chest radiograph — When patients present with HM, a chest radiograph is performed only if the patient has pulmonary symptoms such as dyspnea or chest pain.

Uterine evacuation — Suction evacuation of the uterus is both diagnostic and therapeutic for molar pregnancy. There is no role for endometrial biopsy in the evaluation of molar pregnancy, since complete uterine evacuation will eventually be required.

Uterine evacuation for HM is discussed in detail separately. (See "Hydatidiform mole: Treatment and follow-up", section on 'Uterine evacuation'.)

DIAGNOSIS — HM is a histologic diagnosis, based on a uterine evacuation specimen. Complete and partial mole are differentiated based on histopathology, karyotype (complete moles are diploid; partial are triploid), and whether a fetus is present (fetus is only present in partial mole) (table 1) [75-77].

An initial clinical diagnosis and the exclusion of a viable intrauterine pregnancy allows the clinician to proceed with a uterine evacuation. This can be made based on pelvic ultrasound findings characteristic of complete or partial mole and associated clinical features:

●Complete mole – If the ultrasound is indeterminate, a clinical diagnosis of complete mole can be made based on a human chorionic gonadotropin (hCG) >100,000 mIU/mL [71]. Additional supportive findings are theca lutein cysts or hyperthyroidism without another likely etiology. Patients who present with preeclampsia at <20 weeks should be considered to have gestational trophoblastic disease unless proven otherwise.

●Partial mole – Partial mole is difficult to diagnose on ultrasound. A fetus may be present, but may be growth restricted or have low amniotic fluid. Characteristic sonographic findings are present in a minority of cases. The hCG may be elevated above the expected level, but if it is <100,000 mIU/mL, it is not informative. hCG levels vary in normal pregnancy and there is not a standard level for each gestational age. A uterine curettage may be performed if a fetal heartbeat is not present at an appropriate hCG level and gestational age. For earlier gestations, the patient should be followed with serial ultrasound and hCG levels (eg, weekly) until a normal pregnancy can be ruled out. (See "Pregnancy loss (miscarriage): Ultrasound diagnosis".)

Use of hysteroscopy as an adjunct diagnostic tool has also been described in patients with challenging clinical scenarios [78].

DIFFERENTIAL DIAGNOSIS — The differential diagnosis for HM includes nonmolar pregnancy, spontaneous abortion with hydropic change, and other etiologies of an enlarged uterus, hyperthyroidism, or ovarian theca lutein cysts. During a spontaneous or missed abortion, the chorionic villi may degenerate and develop hydropic change, thereby appearing molar. But these hydropic villi are not accompanied by trophoblastic hyperplasia, the feature that distinguishes molar disease from nonmolar hydropic abortions. Hydropic chorionic villi can develop in any nonviable conception where the villi degenerate and become edematous. Hydropic villi can be identified both on ultrasound and the gross appearance of the products of conception.

Bleeding or pelvic discomfort may occur during a nonmolar pregnancy, including a normal pregnancy, spontaneous abortion, or ectopic pregnancy.

An enlarged uterus may be due to benign or malignant pathology, including leiomyomas, adenomyosis, or uterine malignancy. (See "Uterine fibroids (leiomyomas): Epidemiology, clinical features, diagnosis, and natural history" and "Uterine adenomyosis" and "Endometrial carcinoma: Clinical features, diagnosis, prognosis, and screening" and "Uterine sarcoma: Classification, epidemiology, clinical manifestations, and diagnosis".)

Ovarian theca lutein cysts may also occur in ovarian hyperstimulation syndrome during ovulation induction for assisted reproduction. (See "Pathogenesis, clinical manifestations, and diagnosis of ovarian hyperstimulation syndrome".)

Hyperthyroidism symptoms should be evaluated and the etiology determined. (See "Diagnosis of hyperthyroidism".)

The differential diagnosis of gestational trophoblastic disease, including gestational trophoblastic neoplasia, is discussed separately. (See "Gestational trophoblastic neoplasia: Epidemiology, clinical features, diagnosis, staging, and risk stratification", section on 'Differential diagnosis'.)

SPECIAL CLINICAL SITUATIONS

Multiple gestation — Multiple gestation may be complicated by HM, either a complete or partial mole and a viable fetus. A multiple conception with a normal co-twin with a complete or partial mole is a rare occurrence, developing in only 1 per 20,000 to 100,000 pregnancies [79,80]. It is possible that the incidence will increase with increased use of ovulation induction in older patients and an increased rate of multiple gestation [79].

Multiple gestations that include HM can usually be detected by pelvic ultrasound, but amniocentesis and chromosome analysis is occasionally required. Patients should be advised of the potential risks, including: (1) severe complications such as preeclampsia, hemorrhage, and thyrotoxicosis, which typically develop in the second trimester; (2) preterm birth; and/or (3) malignancy, gestational trophoblastic neoplasia (GTN) [80].

There are limited data about these gestations; representative studies include:

●In a meta-analysis of 14 studies including 244 cases of complete HM with a coexisting fetus, GTN was diagnosed in 34 percent of patients [81]. Of the 182 patients who continued their pregnancy and did not terminate, the overall live birth rate was 50 percent and the rate of intrauterine fetal demise was 40 percent. Maternal complications (eg, vaginal bleeding, preeclampsia, hyperthyroidism, preterm birth and still birth) occurred in more than 80 percent of cases; no maternal deaths were reported.

●A report from the authors' center reviewed 22 cases of complete mole with a coexisting twin and one case of a partial mole coexisting with a normal placenta and fetus; compared with singleton complete mole, pregnancies consisting of complete mole and coexisting fetus tended to be diagnosed later, have more markedly enlarged uteri, and have higher pre-evacuation human chorionic gonadotropin levels [82]. Persistent tumor developed in 12 (55 percent) of 22 patients. Five (23 percent) fetuses survived, and no fetal anomalies were noted. This report was then compared and combined with Brazilian centers to provide information from a total of 72 patients with complete mole and a coexisting normal fetus [83]. Among 62 patients who chose to continue pregnancy, 36 (58 percent) delivered a live baby. Medical complications developed in 41 (65 percent) of 63 patients with known outcomes, and GTN occurred in 31 (46 percent) of 68 patients with known outcomes.

Patients with complete mole and coexistent normal co-twin should be counseled regarding the increased risk of medical complications, including postmolar GTN [84]. These patients may continue the pregnancy under careful monitoring. They should be evaluated and managed by a gynecologic oncologist with expertise in gestational trophoblastic disease and a maternal-fetal medicine specialist, if possible.

Familial recurrent molar pregnancy — Patients with recurrent molar pregnancies can have biparental molar rather than typical androgenetic disease; this is typically familial [85]. Genetic studies in such families showed that the related genes are at chromosome 19q13.3-13.4, and subsequent analysis noted NLRP7 mutations in this region [86]. Data showing clustering of mutations in the leucine-rich region of NLRP7 suggest that this region is crucial for normal function. Some androgenetic diploid complete moles and possibly even triploid partial mole might also carry NLRP7 mutation [87]. Less common mutations associated with familial recurrent molar pregnancy have been identified in the genes KHDC3L and PADI6 [88,89]. NLRP7, KHDC3L, and PADI6 are all components of the subcortical maternal complex, a multiprotein complex essential for early embryogenesis that appears to regulate DNA methylation in oocytes [90].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Gestational trophoblastic disease".)

SUMMARY AND RECOMMENDATIONS

●Classification and pathogenesis

•Hydatidiform mole (HM) is part of a group of diseases classified as gestational trophoblastic disease (GTD), which result from an aberrant fertilization and have the potential to locally invade the uterus and metastasize. Malignant GTD is referred to as gestational trophoblastic neoplasia (GTN) and includes invasive mole, choriocarcinoma, placental site trophoblastic tumor, and epithelioid trophoblastic tumor. (See 'Introduction' above.)

•HM can be categorized as complete mole or partial mole. Complete moles are diploid and partial moles are triploid. The risk of GTN is higher for complete mole (15 to 20 percent) than partial mole (1 to 5 percent) (table 1). (See 'Classification and pathogenesis' above.)

●Risk factors – The two main risk factors for GTD are extremes of maternal age (over age 35 or under 20) and previous GTD. (See 'Risk factors' above.)

●Clinical features – HM presents with an elevated serum hCG concentration and is often accompanied by vaginal bleeding, an enlarged uterus, and pelvic discomfort. Since nonmolar pregnancy is by far the most common source of human chorionic gonadotropin (hCG), clinicians generally initially suspect a pregnancy complication (threatened or missed abortion, ectopic pregnancy) rather than HM. (See 'Clinical features' above.)

●Evaluation – Initial evaluation consists of a quantitative serum hCG and pelvic ultrasound examination.

•hCG – Serum hCG is often much higher in patients with GTD than in those with intrauterine or ectopic pregnancies. Complete moles typically produce higher levels of hCG than partial moles; hCG is >100,000 mIU/mL in approximately 40 percent of complete moles and 6 percent of partial moles. (See 'hCG' above.)

•Ultrasound – The classic ultrasound appearance of a complete mole is a central heterogeneous mass with numerous discrete anechoic spaces (referred to as a "snowstorm" or "Swiss cheese" pattern (image 1A and image 1B)). The ultrasound is more likely to be indeterminate in a partial mole, but abnormalities of the gestational sac or a placenta with cystic spaces may be seen. In partial mole, a fetus may be present (image 3A). (See 'Pelvic ultrasound' above.)

●Diagnosis – HM is a histologic diagnosis, based on a uterine evacuation specimen. Complete and partial mole are differentiated based on histopathology, karyotype (complete moles are diploid; partial are triploid), and whether a fetus is present (fetus is only present in partial mole). (See 'Diagnosis' above.)

ACKNOWLEDGMENTS — The UpToDate editorial staff acknowledges Jonathan S Berek, MD, MMS; Jing Wang Chiang, MD; and Donald Peter Goldstein, MD, who contributed to earlier versions of this topic review.