Description of the condition

Gestational trophoblastic disease (GTD) is a spectrum of disease characterised by an autonomous overgrowth of foetal chorionic tissue or trophoblast. Hydatidiform mole (HM) is the most common and benign form of the disease. The prevalence of HM is highest in Asia, with rates ranging from 1 to 2 per 1000 pregnancies in Japan and China (Palmer 1994; Song 1987; Takeuchi 1987), to 12 per 1000 pregnancies in Indonesia, India and Turkey (Aziz 1984; Gül 1997; Steigrad 2003). In North America and Europe the incidence is reported to be lower, at 0.5 to 1 per 1000 pregnancies (Lee 2009; Steigrad 2003). The incidence has been reported to vary with race (Tham 2003), maternal age (Parazzini 1986), parity (Bagshawe 1986), and diet (Berkowitz 1985; Coullin 2015; Parazzini 1988). Variations in prevalence may be because of differences in reporting between hospital‐based and population‐based data or in the availability of central pathology review (Seckl 2010; Smith 2003).

HMs are categorised as partial (PM) or complete moles (CM) based on their gross morphology, histopathology and karyotype. CMs usually occur when a duplicated haploid sperm fertilises an anucleate or 'empty' ovum; the ensuing diploid product (usually 46 XX) is therefore paternally derived (Fisher 2009). PMs are usually triploid (69XXX, 69XXY or 69XYY), with two sets of paternal haploid genes and one set of maternal haploid genes, and occur when two sperm fertilise one ovum (Fisher 2009). With CMs there is no evidence of foetal tissue; however, with PM an embryo or foetus frequently dies in early pregnancy and foetal tissue and blood cells may be identified in 20% and 50% of PM specimens, respectively (Sebire 2009).

HMs usually present with vaginal bleeding. Associated features (including excessive uterine enlargement, theca lutein ovarian cysts, hyperemesis, pre‐eclampsia and hyperthyroidism) are more common in CMs; however, they occur less frequently as the routine use of ultrasound has led to earlier diagnosis (Seckl 2010). The management of PMs and CMs is similar (Berkowitz 2009a; Berkowitz 2009b). For women who want to preserve their fertility, an evacuation of the retained products of conception (ERPC) is performed, ideally by suction curettage, to remove all trophoblastic tissue completely (Seckl 2010). After suction curettage, around 85% of CMs and more than 98% of PMs will resolve without the need for further treatment (Ngan 2012). Most women are cured in this way; however, in some women HM persists and becomes malignant (gestational trophoblastic neoplasia (GTN)), requiring treatment with chemotherapy. In Japan, second ERPCs are performed routinely for HMs within one week of the initial ERPC, to ensure that there is no residual molar tissue in the uterus (Sasaki 2009). Second ERPCs may reduce the risk of GTN; however, to our knowledge, there is currently insufficient evidence to support this routine practice. Hysterectomy may reduce the risk of GTN by up to 10% (Bahar 1989; Curry 1975) and is an option for women who do not wish to retain their fertility or who experience life‐threatening bleeding at the time of evacuation; however, it does not avoid the need for subsequent monitoring or chemotherapy (Tidy 2009).

Transformation to GTN is considered to have occurred when trophoblastic activity remains following evacuation, as shown by a plateau or rise in serial β‐human chorionic gonadotrophin (hCG) levels, raised hCG levels six months after evacuation or if the histopathological examination indicates choriocarcinoma (Kohorn 2009). However, the presence of raised but falling hCG levels six months after evacuation of a molar pregnancy is no longer an absolute indication for chemotherapy: it appears to be safe to continue with active monitoring without detrimental effect (Taylor 2016). The risk of developing GTN is reported to be 16% to 20% in women with CM (Berkowitz 1995; Curry 1975; Felemban 1998; Seckl 2009); and 0.5% to 1% in women with PM (Bagshawe 1990; Seckl 2009). In the UK, this translates to a GTN transformation rate of approximately 8% of all molar pregnancies (Seckl 2009). Thresholds for treating persistent GTD differ by region with, for example, more than twice as many women in the USA (20%) receiving chemotherapy for persistent GTD than in the UK (Hancock 2009).

HMs may be categorised as being at a low or high risk of malignant transformation based on criteria first introduced by Bagshawe 1976 (Table 1; Berkowitz 1987). Women with high‐risk HMs have more than one of the following characteristics: an initial serum β‐hCG more than 100,000 mIU/mL; uterine size larger than gestational age; theca lutein cysts more than 6 cm in diameter; maternal age over 40 years; and other associated medical and epidemiological factors, including previous GTD, hyperthyroidism and trophoblastic embolisation (Berkowitz 1995). Approximately 30% to 50% of high‐risk HMs will progress to GTN (Goldstein 1981; Limpongsanurak 2001; Uberti 2009).

GTN, which may also follow a 'normal' pregnancy, an ectopic pregnancy or a miscarriage, is classified as low or high risk using a modified World Health Organization (WHO) scoring system adapted by the International Federation of Gynecology and Obstetrics (FIGO 2009). Low‐risk GTN accounts for 95% of cases in the UK and has a cure rate of almost 100% (Seckl 2010). High‐risk GTN has a cure rate of between 80% and 90%; these lesions require combination chemotherapy regimens and frequently develop drug resistance (Goldstein 2012).

Description of the intervention

Methotrexate was first reported to be active against trophoblastic tissue in the mid‐1950s (Hertz 1956). Since then, GTN has been shown to be a highly chemosensitive disease, with various chemotherapeutic agents achieving good rates of cure. All women with 'low‐risk' GTN and approximately 80% to 90% of women with 'high‐risk' GTN will be cured following treatment with one or more chemotherapy regimens (Seckl 2010; Goldstein 2012). Since chemotherapy drugs are associated with various toxic effects, most commonly myelotoxicity, gastrointestinal toxicity, stomatitis and alopecia, the chemotherapeutic aim when treating GTN is to provide the most effective treatment with the least toxicity. Methotrexate and dactinomycin are considered to be relatively safe agents that are commonly administered as first‐line chemotherapy for GTN, alone or in combination with other agents (Alazzam 2012). They have not been shown to be associated with adverse reproductive outcomes, ovarian failure or second tumours (Goldstein 1995).

The use of prophylactic chemotherapy (P‐Chem) in women with molar pregnancy was first described in 1966 (Lewis 1966). Since then, studies of dactinomycin and methotrexate administered before, during or after evacuation of a molar pregnancy have reported encouraging results (see Table 2). Several studies have found a significant reduction in GTN for high‐risk HMs only (Kim 1986; Fasoli 1982; Park 1996). Various dosing schedules have been described including five‐day dactinomycin (Goldstein 1974; Goldstein 1981; Limpongsanurak 2001; Park 1996), eight‐day methotrexate‐folinic acid (Goldstein 1971; Kim 1986; Park 1996) and single‐dose dactinomycin (Uberti 2006; Uberti 2009).

How the intervention might work

As GTN is a highly chemosensitive disease, prophylaxis with chemotherapy agents that have known activity against trophoblast tumour cells may prevent progression to GTN. The use of P‐Chem has been based on an assumption that the development of GTN is pre‐determined, that metastatic GTN spreads via the bloodstream and that high serum levels of cytotoxic agents around the time of evacuation should reduce the ability of the trophoblast cells to invade or metastasise (Goldstein 1995).

P‐Chem may be particularly useful in women with high‐risk CMs who have poor access to health care, for whom hormonal follow‐up is not available, or where poor compliance may be an issue (Berkowitz 2009a; Limpongsanurak 2001; Uberti 2006). In Latin America, loss to follow‐up may be as high as 44% in some areas; hence numerous referral centres in this region are reported to use P‐Chem (Charry 2009). However, the use of P‐Chem may expose women to toxic side effects (Kaye 2002 and Ratnam 1971 have reported toxicity‐related deaths with methotrexate prophylaxis), may lead to inadequate follow‐up, and incompletely protects women against persistent tumour (Goldstein 1995; Hancock 2009). Furthermore, P‐Chem may favour the development of drug resistance (Kim 1986) and delay the time to effective treatment, thereby having adverse effects on survival.

Why it is important to do this review

P‐Chem for high‐risk HM appears to be routine clinical practice in some regions of the world (Charry 2009). Although several studies have been reported, it remains unclear whether P‐Chem, which may be associated with substantial toxicity, will prevent malignant transformation of HM. Furthermore, if P‐Chem benefits exist, it is not clear which drug regimen might have the best effectiveness‐to‐toxicity ratio. We undertook this review in an attempt to clarify the benefits and risks associated with P‐Chem for HM.