Epidemiology

3
EPIDEMIOLOGY
Min C Choi, Chan Lee, Harriet O Smith, Seung J Kim

Gestational trophoblastic disease (GTD) encompasses an intriguing

group of interrelated diseases derived from placental trophoblasts.
Variants differ in propensity for spontaneous resolution, local
invasion, and metastasis [1]. The most serious form, choriocarcinoma,
is usually preceded by a hydatidiform mole (HM), but can occur
following any type of gestational event. As the name implies, all GTD
variants arise following an antecedent pregnancy, although recent
studies have shown that related pregnancy is sometimes not the one
most temporally related [2].
Since pregnancy is an obligatory precursor, ideally, incidence ratios
should be described using the total number of pregnancies, including
unrecognized pregnancies. Because the total number of pregnancies is
difficult to estimate, most epidemiologic studies of GTD have used
surrogate denominators - usually, the number of deliveries and live
births within a given hospital or institution. Over the past three or
four decades, in many regions of the world, population-based registries
dedicated to the study of GTD incidence rates have been established,
and newer epidemiologic studies from these registries have begun to
report GTD rates using population based denominators.
In recent years, it has also been shown that the determination of
the absolute number of incident cases, or the true numerator for
GTD, is almost as elusive as the absolute denominator. Newer
approaches to clinical and laboratory diagnosis, including DNA flow
cytometry, cDNA microarray, in-situ hybridization, polymerase chain
reaction (PCR), and recognition of particularly rare variants (placental
site tumors, epithelioid trophoblastic tumor, placental site nodule)
have improved the accuracy of estimations of the absolute number of
GTD cases in some centers; however, these technological advances
are not yet universally applied.
Dramatic differences in incidence rates for GTD have been
reported from hospitals and regions throughout the world. These
findings have led to speculation that there are environmental and
perhaps, genetic differences in risk for GTD among different ethnic
and racial groups. In this chapter, the epidemiology of GTD, including
the difficulties encountered in determining incidence rates, are
addressed. Despite the limitations of earlier investigative efforts, newer
population-based data have enhanced our understanding of trends in
GTD incidence rates. These studies, which suggest that GTD
incidence rates are declining throughout the world, are presented, and
possible etiologic risk factors are discussed.
This edition will conserve, but not review the prior epidemiology
studies. We will focus the new changed incidence of GTD and try to
analyze how the improvement of medicine and society has caused

1
 Epidemiology

effect on it.

3.1 DEFINITIONS OF GESTATIONAL

TROPHOBLASTIC DISEASE

For all GTD variants, the involved trophoblastic tissue is genetically

distinct from the mother. That is, gestational trophoblastic diseases are
allografts. Several different classification systems continue to be used
for staging, including the World Health Organization (WHO) Scoring
Index, the FIGO system, an NIH classification, and combination
classification, although implementation of recent revisions should help
to consolidate reporting [3-5]. WHO divides GTD variants
histologically into distinct groups, including HM, choriocarcinoma,
placental site trophoblastic tumor (PSTT), miscellaneous trophoblastic
tumor (exaggerated placental site, placental site nodule or plaque), and
unclassified trophoblastic lesions. Hydatidiform molar pregnancies
(complete, partial, invasive) are thought to represent abnormally
formed placentas with specific genetic abnormalities in the villous
trophoblasts, whereas choriocarcinomas and PSTTs are considered
true neoplasms, the former of previllous and the latter, of extravillous
origin, and characterized by intermediate trophoblasts [6].
In 1977, Kajii and colleagues demonstrated the androgenetic origin
of complete, or classic hydatidiform mole [7]. In complete moles, the
nuclear DNA is exclusively of paternal origin. The majority are 46XX,
and arise from fertilization of an empty pronucleus of ovum by a
haploid sperm that undergoes duplication [8]. This is usually referred
to as diandric diploidy. The remainder, 46XY, arise from the fertilization
of an empty egg by two sperm, termed diandric dispermy [9-11].
Typically, the karyotype of partial moles is triploidy (69
chromosomes), and usually results from the fertilization of an oocyte
by two spermatozoa, or diandric triploidy. These are cytogenetically
distinguished from triploid conceptuses with a diploid set of maternal
DNA and a haploid set of paternal DNA, or digenic tripoidy [6, 12-13].
In studies that include both, enormous differences in ratios of
complete versus partial mole variants have been reported. Ratios of 1
to 3 and 3 to 1 have been reported, although more recent studies tend
to include more partial mole cases [14-18]. The difficulties in
distinguishing complete from partial mole by microscopic morphology
alone are underscored by a study conducted by pathologists with
special interest in trophoblastic diseases. In this study, the diagnosis
was agreed upon by all in only 35 of 50 cases (70%) [19]. While
complete mole does not usually present a diagnostic dilemma, it is
difficult and sometimes impossible to distinguish a partial mole from a
hydropic abortus using light microscopy alone [6, 14-24].
P57kip2 is the protein product of a paternally imprinted or maternal
gene that inhibits cyclin-dependent kinases (CDK), thus serving to
inhibit cell proliferation and to suppress tumor growth. Its lack of
expression in trophoblastic disease plays a role in its abnormal
proliferation and differentiation. P57kip2 immunohistochemical
staining was absent in the trophoblastic layers of complete HM and as

2
 Epidemiology

positive in the trophoblastic layer of non-molar villi and mesenchymal

dysplasia [25].
The possible role of genetic and environmental factors in the onset
of GTD has been investigated. A cDNA microarray analysis consisting
of 10,305 human genes revealed significant changes, with 91 genes
being upregulated and 122 being down regulated. Significance analysis
of microarray revealed significant changes in gene expression,
including those associated with signal transduction, cell culture,
transcription and apoptosis. Further RT-PCR analysis of altered gene
expression in molar tissues supported the microanalysis results. This
study confirmed the upregulation of TLE4, CAPZA1, PRSS25,
RNF13, and USP1 in complete HM tissues. Moreover, this study
provides the first evidence that ELK3, LAMA3, LNK, STAT2, and
TNFR SF25 are downregulated in complete HM compared to normal
early placenta tissues. This approach allowed us to identify genes that
are differentially expressed in complete HM tissues as compared to
early normal placenta tissues [26].

3.2 LIMITATIONS OF MOST EPIDEMIOLOGIC STUDIES

Despite extensive epidemiologic data spanning at least 50 years,

considerable difficulty exists in determining racial and geographic
differences in incidence rates for GTD, including HM and
choriocarcinoma, throughout the world. The major limitations of data
published between the 1960s and early 1980s include imprecise
definitions of included cases, extensive variability in the methods used
to detect cases, and differences in the denominators used [27].

3.2.1 PROBLEMS WITH CASE DEFINITION

For rates of GTD to be compared, reproducible definitions of the

conditions encompassed by the term “GTD” had to be developed.
Before the recommendations of the WHO Scientific Group meeting
were accepted most reports lacked clear, precise, and reproducible case
definitions [28-29]. Earlier epidemiologic studies also predate the
discovery by Szulman and Surti of two genetically distinct and specific
types of HM - partial hydatiform mole (PHM) and complete
hydatidiform mole (CHM) [12]. Other terms were routinely applied
before these discrete entities were recognized, and included
“destructive mole”, “invasive mole”, “transitional mole”, and
“hydropic degeneration”. At best, this terminology was confusing, and
at its worst, comparisons of all GTD cases across studies were not
possible [28]. As has been discussed, using light microscopy hydropic
degeneration can be easily confused with PHM, and destructive and
transitional mole with invasive mole [6, 15-24].
Studies that included all GTD variants invariably also included
cases that would be reclassified as first trimester abortions, if flow
cytometry and/or cytogenetic analyses available today were applied [6,
15-24]. By design, studies in which the scope of investigation was

3
 Epidemiology

restricted to CHM and/or choriocarcinoma invariably reported lower

incidence ratios, because partial moles were excluded [15-24, 28].
However, older studies that attempted to include all PHM cases,
including clinically ambiguous ones, probably had falsely elevated
numerators, because triploid conceptuses have many histological
features similar to partial mole. The diagnosis of invasive mole
requires histologic evidence of myometrial invasion. While this
diagnosis can easily be made from microscopic evaluation of the
uterus, hysterectomy is usually unnecessary and not routinely
performed [27, 30]. The rare variants, including PSTT [31-33],
epitheliod trophoblastic tumor [32, 34], and exaggerated placental
site/placental site nodule [35], have only recently become recognized.
PSTTs were initially thought to be clinically benign [31] and may not
have been included in registry data until case reports of metastasis and
even death despite conventional therapy began to appear [32-33].
While some early reports may have included metastatic PSTT as GTD
or choriocarcinoma cases, prior to 1977 benign forms probably went
unrecognised [31]. This is exemplified by an analysis of
choriocarcinoma using data collected by the Surveillance,
Epidemiology, and End Results (SEER) database. In the first 10 years
of data entry (1973-1982), no PSTT tumors were listed [36].
Unlike most malignant tumors, GTD is unique in that surveillance
and therapy is based less upon histopathologic diagnosis, and more
upon the biologic behavior of the disease. Regardless of histology, and
even in the absence of histologic confirmation, therapy has been based
upon persistently elevated or rising human chorionic gonadotropin
(hCG) levels, the most recent antecedent pregnancy, and presence or
absence of one or more radiographically detectable metastatic lesions
[3-5, 37]. The pitfalls of treatment based upon serum hCG levels alone
have been addressed [37]. However, biopsy, especially when the vagina
or distant sites are involved, is not feasible because of the highly
vascular nature of these lesions [38]. Many excellent epidemiologic
studies are limited by inclusion or exclusion of cases that lack
pathologic confirmation [36, 39]. Excluding cases that lack pathologic
confirmation from population-based studies is also problematic, as
clinically relevant cases are missed [39]. In the study of
choriocarcinoma using the SEER database, previously mentioned,
histologic confirmation was obtained in 89.7% of cases, and the others
the remainder were diagnosed on radiographic, clinical grounds, or
observation at surgery [36].
Fortunately, other radiographic diagnostic tools including
ultrasound, color Doppler ultrasonography, magnetic resonance
imaging (MRI), computerized tomographic imaging (CT) and PET-CT
are aiding in anatomic diagnosis. These studies have begun to be used
in the clinical setting when a tissue diagnosis is not feasible [40-42].
When modern techniques that determine the cytogenetics of HM are
applied with greater consistency in the clinical setting, more accurate
classification, and comparison of other epidemiologic considerations
including racial, regional, and cultural differences in risk, will be more
feasible. Because the biologic behavior of molar pregnancies
(especially PHM) detected using newer methods is less well

4
 Epidemiology

understood, hopefully, investigators having access to these newer tools

will be able to determine those cases needing surveillance from those
of little or no clinical significance. However, until the clinical
significance of disputed cases is better understood, it would helpful to
consider these separately from other GTD cases. Without staging and
classification criteria that can be applied across all continents and
registries, global differences in incidence rates will continue to remain
an enigma.

3.2.2 PROBLEMS WITH CASE DETECTION

When GTD and choriocarcinoma cases are managed by regional

referral centers and incidence rates are reported per live births and
deliveries within a given institution or hospital, the number of cases
relative to the regional population is invariably overestimated. It is
now clear that many of the differences in incidence rates reported for
different populations were the result of over-reporting of GTD cases.
Hospital-based studies invariably tend to have falsely elevated
incidence rates, particularly those from regions of the world where
uncomplicated live births and pregnancies do not receive hospital-
based care. Data derived from communities where medical attention
is suboptimal are also likely to underreport GTD cases. For example,
in regions of the world where spontaneous abortions and D&C
specimens are not routinely subjected to histopathologic review, GTD
cases that resolve spontaneously are probably not counted. Even in
developed countries, under-registration of GTD has been reported. In
Sweden, which maintains a national registry, 25% of HM cases
diagnosed between 1971 and 1986, and 66% of treated cases, were not
included within their cancer registry. The most common reason given
for omission was absence of histopathological confirmation [39].

3.2.3 DIFFERENT DENOMINATORS FOR THE

POPULATION AT RISK

GTD occurs as a result of an abnormal fertilization process, and can

therefore occur following any type of conception. Thus, the total
number of pregnancies is the preferable denominator to use when
calculating GTD incidence ratios. Ideally, the pregnancy denominator
should include all live births, stillbirths, spontaneous abortions,
induced abortions, ectopic pregnancies, as well as clinically
unrecognized pregnancies, if accurate estimates can be made.
However, population-based data regarding unrecognized pregnancy
loss is virtually nonexistent. In one study, the rate of early pregnancy
loss in one year identified by a rise in urine hCG was found to 31%,
and 22% of pregnancy losses occurred before the pregnancy was
detected clinically [43]. Most of the earlier epidemiologic studies
reported GTD ratios using deliveries or live births as denominator,
which overestimated the reported incidence ratios. In a more recent
population-based study, it was shown that calculations of

5
 Epidemiology

choriocarcinoma incidence rates using live births and deliveries falsely

elevated incidence rates by approximately 20% [36]. While this study
constitutes the best available published estimate of choriocarcinoma
incidence rates in the United States, since the number of pregnancy
terminations before 20 weeks of gestation and the number of ectopic
pregnancies was unknown, it is likely that even their best estimates of
the number of clinically recognized pregnancies over-estimated the
ratio, because these data were not adjusted for unrecognized
pregnancy losses.
Recent reports have attempted to record the best available
numerator information including all GTD variants, as well as
reasonably accurate denominator information (live births, clinically
recognized pregnancies, and adult female population at risk). Age-
adjusted standardization, which has proved to be a reproducible means
of comparing malignancy rates over calendar time and across
registries, may be an alternative means of estimating GTD incidence
rates, especially if ratios are calculated including cases and women of
reproductive age, who represent the fertile population [36, 44-45].
There are potential pitfalls with this denominator as well, because of
differences in sterilization and hysterectomy practices in fertile
women. However, because most countries maintain fairly accurate
population-based age-specific census data for their adult populations,
and many apply age-adjusted standardization to cancer statistics, this
may prove to be a valuable means of comparing GTD incidence rates.

3.3 GEOGRAPHICAL VARIATIONS IN THE INCIDENCE

OF GTD

3.3.1 HYDATIDIFORM MOLE

For a long time, epidemiologists, genetics, and gynecologists have

noted significant variability in GTD incidence rates between different
regions of the world (Table 3.1). Differences by geographic region
within the same continent or country have also been reported. The
previously cited differences in study methodology clearly account for
some of the reported differences. In China, a nationwide study
conducted by the National Co-ordination Research Group of
Chorioma (NCRG) [46] found an incidence of 0.78 (0.81) per 1000
live births (pregnancies), and Song and Wu [47] a rate of 0.67 per
1,000 deliveries, similar to that reported from Western countries [48].
However, significantly higher rates of GTD, especially from southern
and coastal regions compared with other parts of China, have also
been reported [46]. The incidence of molar pregnancy in Japan (2.0
per 1,000 pregnancies) is approximately three times higher than the
reported incidence in Europe or North America (0.6 to 1.1 per 1,000
pregnancies) [49]. Another well-controlled study involving 20 different
prefectures indicated that the incidence of HM in Japan (16,829 cases
between 1974 and 1982) ranged from 2.83 to 3.05 per 1000 live births
[50]. In South Korea, the nationwide incidence of HM ranged from
1.9 to 2.1 per 1000 deliveries comparable to that in Japan [51-52].

6
 Epidemiology

Table 3.1 Selected incidences of hydatidiform mole reported by various authors

Rate per 1,000

Country Authors Pregnancies Deliveries Live births

Population-based studies
Latin America
Paraguay Rolon and de Lopez [57] 0.2 -- --
North America
Canada Yuen and Cannon [88] 0.83 -- --
Greenland Nielsen and Hansen [90] -- 1.2 --
USA Hayashi et al. [48] 1.1 -- --
USA Matsuura et al. [62] 0.8 (White Hawaiian) --
1.75 (Filipino Orientals) --
New Mexico Smith HO et al. [45] 1.14 (Non-Hispanic whites) 1.04
1.14 (Hispanic whites) 1.34
2.23 (American Indians) 2.27
North Asia
Japan (20 prefectures) Takeuchi [50,118] 2.96 -- 3.0
Korea Kim et al [51,52] 1.9
China NCRG [46] 0.8 -- --
South Asia
Singapore Teoh et al. [114] -- 1.2 --
Middle East Asia
United Arab Emirates Graham and Fajardo [104] -- 2.0 --
Turkey Mungan [106] 1.84 2.48 --
Europe
Italy Mazzanti et al. [100] 0.7 -- --
Sweden Flam and Rutgvist [39,94] 0.9 -- --
Netherlands Franke et al. [91] -- -- 0.68
England and Wales Bagshawe and Dent [92] -- -- 1.54
Northern Ireland Giwa-Osagie [89] -- 2.2 --
Finland Loukovaara et al. [95] -- 0.98 --
Oceania
Australia Olesnicky and Quinn [98] 0.57 -- 0.7

Hospital-based studies
Africa
Nigeria Ogunbode [75] -- 4.8 --
Nigeria Osamor et al [81] -- 1.51 --
Uganda Leighton [82] -- 1.03 --
Latin America
Mexico Marquez-Monter et al. [54] 4.6 -- --
North America
USA Yen and MacMohon [74] 0.63 -- --
Drake et al. [58] -- (Hispanic) 2.38
(Alaska) Martin [59] -- 3.9 --
(Hawaii) Matsuura et al. [62] -- -- 4.6
North Asia
Japan Nakano et al. [116] 1.9 2.6 --
Japan Kanazawa K [117] -- -- 3.70
Taiwan Wei and Ouyang [70] -- 8.0 --
Korea Kim [51,64] -- 2.1 --
China Song et al. [115] -- 6.7 --
South Asia
Indonesia Poen and Djojopranoto [71] 9.9 11.5 --
Malaysia Ong et al. [125] -- 1.5 --
Malaysia Sivanesaratnam [119] -- 2.8 --
Philippines Acosta-Sison [113] 5.0 -- --
Middle East Asia
Iran Javey and Sajadi [103] 3.2 -- 3.7
Turkey Gul et al. [107] -- 12.9 --
Turkey Ozalp SS [105] -- -- 7.8
Saudi Arabia Felemban AA et al. [108] -- -- 1.48
Saudi Arabia Khashoggi TY [109] -- 0.91 --
Saudi Arabia Anfinan et al. [1012] -- 1.26 --
Parkistan Bugti et al. [110] -- 4.06 --
57
 Epidemiology

Iraq Yasin and Chaied [111] -- 1.7 --

Israel Matalon and Modan [102] -- -- 0.8
Europe
Italy Di Fabio and de Aloysio [99] -- 0.8 --
Oceania
Australia (Sydney) Steigrad [96] 0.9 1.0 --

57
 Epidemiology

Table 3.2 Selected incidences of choriocarcinoma reported by various authors

Rate per 10,000

Country Authors Pregnancies Deliveries Live births

Population-based studies
Latin America
Paraguay Rolon and de Lopez [57] 0.2 -- --
Jamacia Segupta et al. [120] -- -- 1.4
Puerto Rico Aranda and Martinez [121] -- 0.3 --
North America
Canada Brisson and Fabia [122] -- 0.4
USA Brinton [38]
North Asia
Japan (20 prefectures) Takeuchi [50, 118] -- -- 0.83
South Asia
Singapore Teoh et al. [114] -- 2.3 --
Europe
Sweden Ringerz [93] 0.2 -- --

Hospital-based studies
Africa
Nigeria Ayangande [80] -- 9.9 --
Latin America
Mexico Marquez-Monter et al. [54] 3.5 -- --
Mexico MacGregor et al. [55] 0.2 0.3 0.3
North America
USA Brewer and Gerbie [127] 0.5 0.6 --
USA Yen and MacMahon [74] 0.3 -- --
(Hawaii) Matsuura et al. [62] -- -- 4.6
East Asia
Hong Kong Chan [123] -- 7.5 --
Japan Nakano et al. [116] 1.2 1.7 --
Japan Kanazawa K [117] -- 1.48 --
Taiwan Wei and Ouyang [70] -- 20.2 --
Korea Kim [64] -- 1.9 --
South Asia
India Pai [124] 19.1 -- --
Indonesia Poen and Djojopranoto [71] -- 17.7 --
Malaysia Ong et al. [125] -- 1.5 --
Philippines Acosta-Sison [113] 8.70 -- --
Thailand Srivannaboom et al. [126] 6.3 6.5 --
Middle East Asia
Israel Matalon and Modan [102] -- -- 0.5
Saudi Arabia Khashoggi TY [109] -- 0.3 --
Iraq Yasin and Chaied [111] -- 0.4 --
Oceania
Australia (Sydney) Steigrad [96] 0.9 1.0 --

58
 Epidemiology

Indonesia has one of the highest reported incidence rates, 1 in 77

pregnancies (1 in 57 deliveries) [53]. It is generally accepted that GTD
incidence rates are higher among non-White Hispanics [54-58],
American Indians and Eskimos [45, 59], and Asian [46-47, 49-52, 60-
72] populations. Alaskan natives appear to have rates 3 to 4 times
greater than Caucasian Americans [59]. In Hawaii, Japanese and
Philippine women have rates higher than Caucasians native to that
state, but Hawaiian Japanese have lower rates than those native to
Japan [60-63].
There is considerable variability in the rates of GTD found in
women of Hispanic descent. Reports from Latin America include rates
from 0.23 per 1000 pregnancies in Paraguay [57] to 4.62 per 1,000
pregnancies in Mexico [54]. Studies from Los Angeles [73], as well as
from New Mexico (USA) [44] found no increased risk for GTD
among Hispanic (1 in 759 live births, 1 in 877 pregnancies) relative to
non-Hispanic whites (1 in 747 live births and 1 in 961 pregnancies)
using conception-based denominators. In this study, there was a
difference observed using the denominator of women-years (age-
adjusted incidence rates per 100,000 women-years for Hispanic and
non-Hispanic whites 5.32 vs. 3.57, respectively, p < 0.001). Higher
fertility rates in Hispanic relative to non-Hispanic whites appeared to
account for the observed differences [44]. Rates reported by these two
USA studies are considerably lower than those reported in hospital-
based studies from Mexico and Central America [54-59]. In more
recent report in USA, Hispanic women had a higher incidence
compared to blacks (2.38 vs. 1.34; P < 0.001), but not Whites (2.00; P
= 0.17) [58].
There is also considerable variability in incidence rates in blacks,
although available data is limited. In Rhode Island (USA) during a 10-
year period between 1956 and 1965, the incidence of HM among
blacks was found to be higher than for whites, but the difference was
not statistically significant. However, only 15 of the 329 cases were in
blacks [74]. Similarly, a case-control study that compared rates of
molar pregnancy among Black, Caucasian, and Latin American women
in Los Angeles, found no racial differences, but again, the sample size
was small. Among all ethnic groups studied there were only 145 molar
pregnancies and 1 choriocarcinoma in 74,548 deliveries [73]. Hayashi
et al. estimated that the ratio of incidence cases per 100,000
pregnancies for blacks (58.2) was approximately half that of whites
(119) and others (117.9) [48]. However, USA rates are substantially
lower than those of older hospital-based studies from Africa, where
ratios per delivery of 1:184 (Lagos, Africa) [75-76], 1:203 (Ibadan,
Nigeria) [77]; 1:329 (University College Hospital, Ibadan, Nigeria) [78];
1:401 (University of Lagos, Nigeria) [79]; and 1:375 (Ogbomoso,
Nigeria) [80] have been reported. More recently, Osamor et al. [81],
Leighton [82], and Egwuatu and colleagues [83] have reported lower
rates (1:655 deliveries, 1:971 deliveries, and 1:1219 pregnancies,
respectively). For choriocarcinomas specifically, Brinton and
colleagues reported that blacks and other races, respectively, had 2.1-
and 1.8-fold higher incidence rates compared with US whites [36]. A

59
 Epidemiology

more recent update of SEER registry data, using all choriocarcinoma

and PSTTs registered between 1973 and 1997 has been conducted,
and results are similar to the earlier SEER report. The relative risk
(common odds ratio, COR) for blacks (COR 2.19, 95% CI 1.71-2.79)
and other non-whites (COR 2.14, 95% CI 1.61-2.84) was significantly
higher than that for whites. However, blacks alone (COR 2.60, 95%
CI 1.28-5.27) were found to have significantly higher mortality rates
[45]. The higher incidence and mortality among blacks for
choriocarcinoma in the USA is similar to data with respect to other
malignancies that are amenable to screening and early detection, and
may be a reflection of reduced access to medical care [84]. A recent
analysis of postmolar surveillance reported that postmolar surveillance
among indigent women, particularly those belonging to minority
groups, is substantially poorer [85].
For the most part, data from the United States [44, 48, 62-63, 73-74,
86-87], other parts of North America [88-89], Western Europe [87-95],
and Australia [96-98] reveal fairly consistent incidence rates, although
regional differences have been reported (Table 3.1). Hayashi et al.
reported a rate of 1.08 per 2,000 pregnancies in American women [48]
and Bagshawe and colleagues, 1.54 per 1,000 live births in England
and Wales [92]. Sweden also maintains a population-based registry of
all GTD cases, and reported incidence rates of 1 in 686 deliveries and
1 in 1103 pregnancies; no trends in incidence rates over the 14-year
period were found [94]. Australian [96-98] and Greenland [90] studies
estimate rates fairly consistent with those from the United States and
Western Europe, although rates are higher from southern Europe [96-
100], the Middle East [101-112], and higher still in Asian women [61-
63, 70, 72, 113-119].
Within the USA, using data derived from the Hospital Discharge
Survey obtained over the 1970s, the incidence ratio of hydatidiform
mole was 1.1 per 1,000 (1 in 923) pregnancies, and rates in New
England, Middle Atlantic, and East South Central states were higher
than rates reported from the Mountain, West, and South Central
regions. The highest rate (1.89) in the Rocky Mountains was 2.7- fold
higher than the lowest rate (0.70) from New England [48]. Taken
within the context of other studies, these observed regional differences
probably reflect the cultural and racial distribution of women living in
these areas, rather than equatorial differences.
The geographical differences in current incidence rates should be
interpreted within the context of obstacles to accurate estimations of
incidence ratios, previously discussed. Nevertheless, the
preponderance of data overwhelmingly supports that the incidence of
GTD is substantially higher in Oriental women, particularly those in
Southeast Asia. Improved epidemiologic studies are needed that
address temporal trends, genetic, and environmental factors within the
context of accurate case detection.

3.3.2 CHORIOCARCINOMA (CC)

As has been shown for molar pregnancy, internationally, the incidence

60
 Epidemiology

rates of gestational CC also widely differ [36, 48, 86-87, 109, 111, 120-
131] although some of these differences are attributable to the same
methodological problems associated with epidemiologic studies of
molar pregnancies. In Europe and North America, one in 30,000 –
40,000 pregnancies, and 1 in 40 molar pregnancies are affected [127-
128] whereas in Southeast Asia, rates as high as one CC in every 500–
3,000 pregnancies have been reported [129]. Estimations of the
incidence of CC have been limited by some of the same problems as
with HM. As indicated in Table 3.2, rates derived from Latin
American (0.2 cases per 10,000 pregnancies) and Europe (0.2—1.5
cases per 10,000 births and pregnancies) are similar, whereas reports
from Japan have indicated higher rates (0.83 cases per 10 000
pregnancies and births). Hospital-based studies have reported rates
ranging from 0.2 to 20.2 per 10000 pregnancies and deliveries.
Problems with definitions and methodological design have an even
greater impact upon epidemiologic studies of CC, because this variant
relative to all GTD cases is relatively rare.

3.4 TIME TRENDS IN INCIDENCE OF GTD

While much of the literature regarding trends in GTD gives conflicting

results, newer data population-based from the United States, and Asia
demonstrate a decline in incidence rates for HM and CC. This section
explores both older and more recent available information regarding
temporal trends for GTD.

3.4.1 HYDATIDIFORM MOLE

In the United States, two hospital-based studies spanning the time

period between 1940 and 1964, reported that the incidence of HM
declined during World War II, and then increased, and eventually
surpassed pre-war incidence rates [74]. Incidence rates also steadily
increased in Jewish women living in Israel from 1950 to 1965 [102].
Similarly, among indigenous Greenlandic women from 1950 to 1974,
incidence rates increased significantly after 1965 [90]. A Japanese study
summarizing registered cases of GTD, and including 16,829 incidence
cases of HM from 1974 to 1982, found that the rates per 100,000
women had decreased yearly from 10.2 in 1974, to 7.9 in 1982. While
significant trends per 100,000 women were found, when live birth
denominators were used, the incidence rates were virtually unchanged
(mean ratio and range per 1,000 births 2.92 and 2.83-3.05, respectively)
[118].
In Asian populations, extensive data suggests that GTD incidence
rates are declining [46, 64-69, 117-118]. In Korea, between 1971
through 1995, and using live births as the denominator, a 17.5-fold
decline in incidence rates [from 40.2 (1971-1975) to 2.3 (1991-1995)]
and to 1.9 (2001-2005) was reported. This study, which appears to be
the largest hospital-based study of GTD to date, included HM (5393
cases), invasive mole (2466 cases), GTN (1600 cases), and PSTT (38

61
 Epidemiology

cases) [51]. In Okinawa, the average GTD incidence ratios per 100,000
populations and per 1,000 live births were 3.41 and 2.29, based upon a
study of 417 cases (1986 - 1995). Although incidence rates in Okinawa
were higher than those of mainland Japan, GTD rates from both have
similarly declined, and the higher rates in Okinawa are proportional to
higher crude birth rates [68]. Although Malaysia lacks a centralized
registry, the ratio per live births for the 1991-1995 time periods was
2.8 per 1000 deliveries, much less than for earlier years [119].
The incidence of HM is well-known that it is the highest in South-
East Asia. It ranges from 1/1000 pregnancies (Japan), 2/1000
pregnancies (China) to 12/1000 pregnancies (Indonesia, India and
Turkey). However, the incidence in North America, Europe, and
Oceania is about 0.5-1/1000 pregnancies. The data from South
America and Africa is limited and sparse. The accurate explanation on
these geographical differences is difficult. Researchers point out
maternal age, diet, ethnicity, gravidity, poor social-economic
conditions, and infection as different factors.
Recent data show the significant reduction of GTD incidence with
the increase of GNP. This is prominent in South Korea
from 4.4/1000 deliveries in the 1960s, 1.9/1000 deliveries in the
1990s, to 1.8/1000 deliveries in the 2000's [64]. Similar results are
obtained from Japan from 4.9/100,000 population in 1974 to
1.9/10000 in 1993. This trend is also noted in Taiwan [70].

3.4.2 CHORIOCARCINOMA (CC)

As HM greatly influences GTN as a precursor pregnancy, GTN

trend may be transposed into HM trend. India and Indonesia are the
highest with 19.1/1000 [124] and 17.7/1000 [71] pregnancies
respectively. It is lower in South-East Asia. Korea shows 4.4/10,000
births in 1960s to 1.9/10,000 births in 1990s [64]. The incidence of CC
has decreased to a third in Israel from 1950-1954 to 1960-65 [102].
While many countries show a constant decrease, some are
continuously increasing.
Until recently, data on temporal trends in CC were exceedingly sparse.
In Israel, the incidence ratios for CC significantly decreased between
1950 and 1965, so that by 1960 to 1965, the rate was less than one-
third of that for 1950 to 1954 [102]. In the study of registration
systems for CC in Japan the incidence of CC per 100,000 women
decreased yearly from 0.31 in 1974 to 0.19 in 1982 or by about two-
thirds over this time period. It is also of note that the incidence per
10,000 births decreased slightly, from 0.86 in 1974 to 0.74 in 1982,
with an average of 0.83 [118]. Although overwhelming evidence
suggests that incidence rates for GTD in Asia are higher than in other
regions, the specific factors that account for higher rates may be one
or more of the following: race, maternal age, environmental, dietary,
and lifestyle factors. To date, no study has specifically addressed all of
these factors within a poly-ethnic community.
A recent analysis of CC and PSTT data (1973-1997) obtained from
the SEER database, which has annually recorded population-based
cancer statistics for approximately 10% of the US population, was

62
 Epidemiology

performed. The 25-year annualized age-adjusted incidence rate per

100,000 women was found to be 0.136; 5-year average annualized rates
declined by 45.6% [from 0.180 (1973-77) to 0.098 (1993-97)]. The
decline in incidence rates was seen for all races, although blacks and
others consistently had approximately twice the rates of non-Hispanic
whites [45]. While this study showed no statistically significant
reduction in mortality rates over 25 years, in China between 1959 and
1985, mortality rates for CC have declined drastically, from more than
90%, to less than 20% [130]. Wang and colleagues noted a 90.6%
survival rate in women treated at the National Yang-Ming University,
Taipei, Taiwan [132].

3.5 RISK FACTORS FOR GTD

Many factors, including viral infection [133-134], poor nutrition [46],

defective germ cells [8], as well as the effects of prior pregnancies,
maternal age, consanguinity, genetics, and environmental factors have
all been considered as risk factors for the development of GTD.
Although a variety these, and potentially other factors have been
considered, the etiology of GTD, and the role of each of these factors,
is poorly understood.

3.5.1 AGE FACTORS

Maternal age has consistently been identified as an important risk

factor. Age-specific incidence reports usually reveal a ‘J curve’. HM is
increased in the extremes of reproductive age. That is, teenagers have
higher incidence rates, and reproductive-aged women 40 years of age
or older have incidence rates that are substantially higher [48, 57, 74,
87, 90, 102, 114, 135-136]. With respect to age-specific rates, the risk is
slightly higher for women 15-20 years old, and about 20-fold higher in
teenagers under 15 years of age. There is a progressive (above 10-fold)
increase in risk in women over 40 years of age [92, 102], such that for
women over 50 years old, the risk that a pregnancy will result in HM is
about 200 times greater than for women 20 to 35 years of age [137].
These age-specific trends affecting younger and older women suggests
that defects in ovoid function (ova that are premature or post-mature)
is one etiologic factor contributing to the risk for GTD. Reinforcing
this hypothesis are findings that this pattern is consistent throughout
all regions and races [48, 62, 64, 90, 115-116, 138].
Paternal age has also reported as a significant risk factor in some
studies, although data are inconsistent [62, 139]. In at least one
Japanese study, paternal age did not appear to play a significant role.
[140].

3.5.2 REPRODUCTIVE AND OBSTETRIC HISTORY

The importance of including all conceptions in the denominator, and

not just live births, has previously been discussed. Unfortunately, very

63
 Epidemiology

few existing reports make the distinction between gravidity and parity.
Epidemiologic studies that adjust for potentially important
overlapping effects, including maternal age, age at first pregnancy,
induced abortions, miscarriages, spacing between pregnancies, and
infertility, spacing between pregnancies and history of infertility are
few and far between. No studies have included all of them.
A history of a previous HM appears to be a strong and well-
established risk factor predisposing to another molar pregnancy [62,
102, 138, 141-142]. The risk of HM increases with prior history of
HM. The relative risk for recurrent mole is anywhere from 20 to 40
times that for the general population [74, 102].
After the first HM, a second molar pregnancy occurs in 0.6-2.60%
of pregnancies [142-146]. Women who have given birth to twins also
appear to be at increased risk. In one study, the incidence of HM
following twin pregnancies was substantially higher than that for the
general population [147].
In both Rhode Island (USA) [74] and in Italy [148], parity was not
found to be associated with an increased risk for HM, after adjustment
for maternal age. However, another Italian study found an elevated
risk of HM in nulliparous women with a history of miscarriage [142].
In the same study, difficulties in conception were significantly more
common in women who subsequently developed HM pregnancies
[142], but in a similar study conducted in Baltimore, no relationship
between a history of infertility and menstrual problems and
subsequent HM was found [139].
A higher rate of HM after artificial insemination by donor
compared with normally conceived pregnancies was found in one
report from Australia [98]. Also, repetitive molar pregnancies with
different male partners have been reported. Six patients experienced a
total of 34 pregnancies with 20 different partners. These pregnancies
resulted in 15 HMs, 8 term live births, 7 therapeutic abortions, 3
spontaneous abortions, and 1 preterm delivery. The rate of persistent
GTD was 3 (20%). Interestingly, 3 of the males reported normal
pregnancies with other partners [149]. A significantly high incidence of
triploidy in a woman with recurrent GTD following normal
conception has been reported after two cycles of in vitro fertilization
[150]. Thus, it appears that previous history of conceptions that occur
and proceed normally reduces a woman’s likelihood of developing
GTD. Whether this reduction in risk merely reflects the absence of
adverse genetic or environmental factors that are known to induce
abnormal fertilization, and thereby increase the risk for GTD, is
unknown.

3.5.3 ETHNICITY

Several studies have been undertaken in polyracial societies, and give

insight into possible racial and ethnic factors in the risk for GTD. As
previously discussed, in the United States from 1970 to 1977, black
women had about half the rate of HM compared with non-black
women [48]. Matsuura et al. reported that the rate of HM was 8.0 per

64
 Epidemiology

1,000 pregnancies in Caucasians, 17.5 in Filipinos, 16.5 in Japanese and

7.7 in Hawaiians [62]. There were also significant differences in
incidence rates among Oriental women native to Hawaii compared
with those born in Japan [62]. A study of Alaskan natives shows that
the incidence is 3-5 times higher than the whites [59].
A Singapore study including cases and women between 1953 and
1965 found that the incidence of HM was 1 in 811 viable Chinese
pregnancies, and occurred more commonly in Indian and Eurasian
populations; in fact, no GTD cases occurred in Caucasians within the
region during the entire 11-year study period [114]. HM was almost
twice as likely to occur among Eurasians in Singapore compared with
Chinese and Malaysian women [114]. Considerable differences by
ethnicity within different provinces of China have also been reported
[46, 47].
The differences in risk between white and Chinese populations may
be genetic, environmental, or both. A comparison of differences in
human leukocyte antigen (HLA) sharing in Taiwanese was compared
with Pittsburgh (USA) couples. Taiwanese couples alone were found
to share HLA B (p < 0.04), HLA DQ (p < 0.007) and also shared 3 or
more HLA A, B, DR, and DQ (P < 0.02) compared with USA
couples. These results support the hypothesis that GTD development
occurs on a sporadic basis in whites and on a genetic basis in Chinese.
[151]. Indians and Pakistani women have also found to be at increased
risk for recurrent molar pregnancy, compared with Caucasians [152].
In this study, those who presented with a partial mole tended to have a
partial mole as the second event, but those with a history of CHM
were at increased risk for PHM, CHM, or CC [152].
There is much evidence to show the correlation of ethnic difference
with molar pregnancy rates. However, it still remains a problem to
study the exact mechanism behind ethnicity.

3.5.4 GENETIC FACTORS

The importance of genetic factors in the etiology of HM is exemplified

by several observations [153-154]. It is speculated that some of the
wide variability in incidence rates observed among different
populations may be due to genetic differences. The high incidence of
HM and other trophoblastic tumors in certain populations, including
women in Indonesia, Taiwan, the Philippines, and the Far-East has
been attributed to inbreeding [150-151]. Among these groups, the risk
of recurrent HM in a subsequent pregnancy is high: 0.6-3% [102, 146,
155], and similar to certain polygenic conditions. A case report of
three families has described recurrent HM in female siblings, and
suggests that the proposed defect in the ovum could be due to a
mutant gene [156].
In some studies, the ABO blood groups of patients and their
husbands have not been found to be different from normal
populations [157], but a significantly higher incidence of blood group
B has been identified in women with recurrent molar pregnancies
[149]. In one investigation, women with molar pregnancy in Japan had

65
 Epidemiology

a lower frequency of Rhesus (Rh) negative blood type compared with

the general population [158].
Cytogenetic studies in two series [8, 159] have reported a frequency
of balanced translocations of 4.6% in women with CHM, compared
with 0.6% for unaffected populations. It is possible that women with
a balanced translocation have a greater chance of abnormal events
during meiosis, increasing the probability of inducing empty or
inactivated eggs.
Karyotypic studies using molar tissues have found that in most
cases, PHM cases demonstrate triploidy or trisomy [13, 160].
However, CHM cases are exclusively or predominantly 46 XX female
karyotype by Q polymorphism [7-8]. Some CHM cases do result from
doubling of a haploid sperm, and in such cases, they are genetically
homozygous. Less commonly, CHM may be genetically heterozygous,
resulting from the fertilization of an anucleate ovum (empty egg) by
two sperms [9-10, 161-163]. This basic classification appears to explain
the vast majority of all CHM cases, although other cytogenetic types
including tetraploidy variants exist [164]. Current technology permits
the accurate classification of HM into PHM and CHM, also allows
CHM cases to be subclassified genetically into monospermic or
dispermic variants, using the PCR with primers for a variable number
of tandem repeat sequences and the Y chromosome-specific
sequences [165].
More recent technological advances that may further elucidate
GTD variants include gene amplification [166], identification of
specific tumor suppressor genes (or their receptors) such as DOC-
2/hDAb2 [167], TNF-α [168], and paternally derived H19 [169]. The
imprinting of IGF2 and H19 was maintained in all normal placenta
tissues but relaxed in GTD. Promoter usage pattern of IGF2 changed
with gestational stage of normal placenta and GTD. These results
suggest that LOI, deregulation of IGF2 promoter, and altered
expression levels of IGF2 and H19 genes might be associated with
progression of GTD [170].
The oncoproteins p53, p21, Rb, and mdm2 have been investigated,
but the overlap in differences in expression among GTD variants
appears to be too great for clinical utility [171]. There is an increase in
circulating soluble interleukin-2 receptor (SIL-2R) coincident with
reduced serum IL-2 levels in women with GTD compared with
normal controls; SIL-2R levels were 3.9-fold higher in complete and
partial moles, and 6.1-fold higher in high risk CC cases [172]. PCR
polymorphisms have also been found to identify very specific genetic
origins of trophoblast tumors [173]. In addition to flow cytometry,
previously discussed [15-24], in-situ hybridization [174] and
deoxyribonucleic acid image and interphase cytogenetic analysis [175]
are being investigated. GTD in a 42-year old with congenital trisomy
8 mosaicism has been described [176]. It is thought that there is an
increased risk for cancer in tissues with constitutional genetic
imbalance, including trisomies or translocations. Nonetheless, using
data from the Danish National Register, no increased risk for other
cancers, except subsequent CC was seen in a study of 1520 women
and their male partners [177].

66
 Epidemiology

The relationship between genetics and GTD will be discussed in

greater detail in a separate chapter devoted to this topic in this text.

3.5.5 FAMILIAL OCCURRENCE AND CLUSTERING

There are several reports of the influence of consanguinity, or family

history, on the incidence of mole [85,154]. One study reported three
families in India each having two or more sisters with HM in one or
more pregnancies [156]. La Vecchia et al. reported that among patients
with molar pregnancy in Italy, there was a history of HM in four of
100 relatives of the woman (three CHM, one CC), and in one of 100
relatives of the father (CHM), compared with no family history among
200 age-matched controls [142]. There have been numerous reports of
recurrent moles in single individuals, but very few describe familial
occurrence of molar pregnancies.

3.5.6 DIET AND NUTRITION

Information regarding the role of diet and nutrition is inconclusive,

and inextricably linked to other potential factors including
socioeconomic status and geographic location. Although there has
been much speculation that one or more dietary factors predispose to
the development of HM, there is very little evidence supporting this
hypothesis. Early reports from the Philippines [113], Mexico [178] and
Taiwan [179] all suggested that nutritional deficiencies play a role,
especially in regions of the world where malnutrition is common.
In one study, which included 25 women with HM, the serum
albumin and total protein were significantly lower compared with
healthy controls [180]. However, two other reports, including a study
from Hawaii [59] and another on native Alaskans, who have a diet
high in animal protein [59] were unable to detect a similar association
between HM and diet. Berkowitz et al. reported that dietary carotene
above the median level of consumption in the control group appeared
to reduce the risk of CHM (relative risk 0.6; p = 0.02) [143]. In
another study, Parazzini et al. identified an inverse relationship
between HM and certain foods, such as carotene, and a significant
trend toward lower GTD rates in women with higher rates of
consumption of carotene and animal fat [181]. However, these results
differ from others, including a case-control study from China that
found no association with dietary intake of vitamin A, or any other
specific food group [182].

3.5.7 ENVIRONMENTAL FACTORS

To date, no specific environmental factors have been identified that

can be linked with certainty to an increased risk for GTD. However,
because these are potentially correctable risk factors, preliminary data
addressing the role of environmental agents are discussed.

67
 Epidemiology

Cigarette smoking

Despite extensive data regarding the effects of cigarette smoking,

including second-hand smoke, on neonatal birth weight and preterm
labor, there have been relatively few studies that have examined the
effects of cigarette smoking on risk for GTD. Parazzini and colleages
found that cigarette smoking increased risk of HM [148]. In women
who smoked more than 15 cigarettes per day, the relative risk for
CHM was 2.6, compared with 2.2 for women who smoked less than
15 cigarettes a day. La Vecchia et al. noted that the duration of
smoking was associated with an increased risk [142]. Assuming the
relative risk in women who never smoked is 1, for women who
smoked less than 5 years, the risk for HM was 1.3, compared with 2.3
in women who smoked 5 to 9 years, and 4.2 for a smoking history of
10 or more years. However, in several other studies, no association
with smoking before or during pregnancy and subsequent HM
development was found [74, 139, 143, 182].

Oral contraception and intrauterine contraceptive devices

There is conflicting evidence on the relationship between oral

contraceptives (OCs) and intrauterine contraceptive devices (IUCDs)
before pregnancy and risk of HM. In China, the past use of OCs was
associated with an increased risk of HM; the relative risk for four or
more years of use was 2.6, and there was a significant trend with
increasing years of use [182]. These results are supported by those of
Palmer, who found a six-fold increased risk of CC in women who had
taken OCs for more than five years [183]. However, in the same study,
no association with invasive mole and CHM was found. Because it was
unclear if OCs are truly a risk factor, or a chance finding in women
with underlying difficulties with conception, a large multicenter trial
was performed to evaluate the effects of OCs and subsequent risk for
GTD [184]. In this study, the relative risk for GTD for any prior use
of OCs was 1.9 (95% confidence interval {CI} 1.2-3.0). The highest
risk index was in women using OCs during the cycle in which they
became pregnant (relative risk 4.0, 95% CI 1.0-10). Nevertheless,
because the relative risk was small, the authors cautioned that a change
in OCs use based upon this study was not recommended [184].
During postmolar surveillance, Stone et al. reported that women
taking OCs resulted in a delay in decline of hCG levels in women who
were not receiving chemotherapy for GTD [185]. However, Berkowitz
et al. [186] and Curry et al. [187] found no adverse effect. The
prospective randomized trial by the Gynecology Oncology Group
found that the median range to spontaneous regression for women
using OCs and barrier methods was 9 and 10 weeks, respectively, and
the rate of conception during surveillance was reduced by 50% [187].
Use of OCs in women with a history of HM compared with other
forms of contraception (barrier methods, IUCDs, no method) has
been found to actually reduce the risk for post-molar GTD. Using a
stepwise regression analysis, the type of contraception used compared
with other risk factors including theca-lutein cysts, Asian descent,

68
 Epidemiology

interval from last menstrual period, and advanced maternal age, was
the most important prognostic factor in postmolar GTD development
[188].
Previous use of an IUCD has also been found to be significantly
associated with HM in an Italian study [142], but Berkowitz et al.
found a much weaker relationship [143], and the Baltimore study
showed no correlation [139].
Diethylstilbestrol (DES) may also be a risk factor. In one study, 2
of 129 daughters of mothers who had used DES in pregnancy
developed HM [189]. Although the incidence ratio (15.5 per 1000
pregnancies) is disconcerting, it is based upon only two exposed cases;
however, the effect of DES exposure merits further study. Since
daughters of mothers who took DES are aging, and risk increases with
maternal age, the exposure may be only coincidentally related.

Other environmental factors

Constable and Hatch have recently reviewed several studies from

Southeast Asia that have addressed the role of herbicides (including
agent orange) and an increased risk for HM [190]. Two studies from
South Vietnam of women with HM and CC treated at the Gynaeco-
Obstetrical Hospital of Ho Chi Minh City were reviewed [191-192].
Rates of HM and CC minimally increased from 8.61 per 1,000
pregnancies in 1952 to 8.7 in 1971, after which it rose abruptly to 45.4
in 1979, declining only slightly in the following two years. If herbicides
were to influence directly the development of HM, one would expect
the epidemic and the spraying to be more closely linked in time.
Alternatively, herbicides enter the food chain and take several years to
affect reproductive processes. The limitations of such studies are
failure to control for maternal age, parity and other variables that are
known risk factors.
A follow-up case-control study of agent-orange and risk for GTD in
this same population has been conducted. Cases were found to eat less
meat, own fewer consumer goods, and an increased risk was
associated with the breeding of pigs. When a cumulative agent orange
exposure index was constructed, using complete residence histories,
no statistical differences between cases and controls for other
agricultural pesticides, or agent orange (OR 0.7, 95% CI 0.2-1.8) was
found [190]. In another review, the use of herbal medicine during the
first trimester was associated with an increased relative risk for
hydatidiform mole, major birth defects, and liver cancer [192]. An
ecological investigation of exposure to ionizing radiation from 1974 to
1976 in Japan, and its association with a higher incidence of GTD,
supports the hypothesis that ionizing radiation may have a causal role
[193]. Although these studies have some methodological problems
that limit their value, nevertheless, environmental factors appear to be
implicated. Potentially, temporal exposure to these factors at the time
of, or near fertilization, could result in abnormal gametogenesis,
thereby inducing a complete or partial mole.

69
 Epidemiology

3.6 RELATIVE RISK (ODDS RATIO) FOR CHM AND PHM

According to several reports, complete and partial mole share several

epidemiological features. In studies where complete mole and partial
mole are analyzed separately, many social and demographic risk factors
were common to both [194-196]. For example, nulliparous women,
those with low parity, and those with a personal history of infertility
problems, spontaneous miscarriages, previous GTD or a family history
of GTD are at increased risk for both PHM and CHM [197]. Other
risk factors including smoking and higher levels of education are more
commonly seen with PHM, and maternal AB or A blood groups with
CHM [193]. Neither has been linked to induced abortions, age at first
pregnancy or alcohol consumption (Table 3.3).
In one study within a poly-ethnic community, Filipino, but not
Japanese women, were found to be at increased risk for CHM but not
PHM [194]. The increased rate of PHM seen in better educated and
presumably, more health-care conscious women may be a reflection of
better access to pathologic assessment of abortion specimens and to
follow-up in specialist centres. Several identified risk factors, including
the relationship with maternal age and ethnicity are similar for both
variants, although the data available is insufficient for formal statistical
evaluation or inference. In a case-controlled study of PHM vs. CHM,
using multivariate analysis, irregular cycles, pregnancy histories
including only male infants among prior live births, and oral
contraceptive use for over 4 years independently and significantly
increased the risk for PHM, and dietary factors had no effect [196].
Of note, while maternal age is a well-recognized risk factor for CHM,
maternal age has not been found to increase the risk for PHM [194-
196].
As previously discussed, the use of flow cytometry and other
cytogenetic studies increases the detection of PHM [15-24]. However,
the clinical ramifications of identification of these cases are uncertain.
While the risk of GTN following a CHM is approximately 20%, in
one review, the risk following PHM was found to be only 2.9% [199].
Berkowitz reported that the risk for gestational trophoblastic
neoplasia (GTN) following PHM was 9.9%, but in this series, all had
non-metastatic disease [200]. CC following partial mole is a rare event;
as of July 1, 2000, only 21 cases have been reported [201-203]. Using
data derived from a large trophoblastic disease referral system, Seckl
and colleagues reported that 15 of 3000 partial mole pregnancies
eventually required chemotherapy for GTN and 3 (0.1%) developed
CC [204]. Although the risk is rare, the importance of appropriate
surveillance, and inclusion of PHM in registry data, is emphasized by
these studies.

70
 Epidemiology

Table 3.3 Corresponding relative risks of various factors reported to correlate with molar pregnancy occurrence.
Odds ratio

Factors Complete mole Partial Mole

Maternal age (years)
< 20 1.5
>40 5.2

Reproductive history
Parity at conception
0 0.9 0.7
3 0.8 0.5
Spontaneous miscarriages
>2 1.5-3.1 1.9
Problems with Infertility 2.4-3.7 3.2
Contraception
Use of oral contraceptives 1.1-2.6 1.3
IUCD user 1.7-3.7
Age of 1st pregnancy < 25 0.6 1.3
Previous molar pregnancy 16.0

Family history
Spontaneous abortion (yes) 1.5

Socioeconomic and lifestyle

Education (years)
> 12 0.9-2.1 2.1
Marital status
Never married 2.1 2.1
Smoking
Ex-smokers 1.1 0.7
Current smokers
> 15 cigarettes per day 2.2 1.8
Alcohol consumption
< 2 drinks 2.1 1.4

ABO blood types

Maternal blood
AB 2.1 1.2
A 1.7 0.9
Maternal A, husband O 1.5 1.5

Nutrition
Vitamin A in diet above control median 0.6
Source: Grimes [28], Matsuura et al. [62], Bracken [87], Olesnicky and Quinn [98], Messerli et al. [139], La Vecchia et al
[142], Berkowitz et al. [143], Brinton et al. [182], Steigrad [197], Altieri et al [198].

71
 Epidemiology

3.7 RISK FACTORS FOR CHORIOCARCINOMA

In the vast majority of cases, CC arises following an antecedent molar

pregnancy. However, CC may follow any antecedent pregnancy,
including a delivery, stillbirth or spontaneous abortion. Cytogenetic
studies have shown that there can also be intervening pregnancies that
are unrelated. Studies have shown that CC is approximately 1,000 times
more likely to develop following a CHM than after a normal pregnancy,
and the risk for CC following CHM is much greater than for PHM.
Clinically, it would be invaluable to determine factors responsible for
malignant transformation following a gestational event. Potentially, if
these factors could be reliably ascertained, selective surveillance for
high-risk pregnancies (including term births) could be offered, instead
of the protocols for currently following all molar pregnancies.

3.7.1 AGE

The median age for women with CC is generally somewhat higher than
that for normal pregnancy. Some reports [102] described slightly higher
rates of CC for teenagers than for women aged 20-40 years, while others
noted lower rates. As with HM, older women account for a minority of
cases of CC due to their low fertility. The potential effect of the age of
the father is unknown.

3.7.2 ETHNICITY

Little information is available on ethnic variability. In the Singapore

study [114], the incidence was highest in Malaysians compared with
other ethnic groups, but the data were rather inadequate, so that no
definite conclusions could be drawn.

3.7.3 REPRODUCTIVE AND OBSTETRIC HISTORY

The data concerning the effect of gravidity, independent of age, are

inadequate and no firm conclusions can be drawn. Poor obstetric
histories associated with increased fetal wastage may increase the risk of
GTN [205-206]. It was found that the risk of GTN was 21, 31 and 34
times higher for women with one, two and three or more fetal losses,
respectively, than those whose known pregnancies all ended in live
births [206].
The higher proportion of CC following HM could be attributable
either to the high incidence of HM in some countries or to a higher
percentage of HM patients in those countries developing CC. The
percentage of cases of CC preceded by HM ranges from 39% [193] to
83% [194] in reports published since 1960. Since HM is an uncommon
pregnancy outcome, the high percentage of cases of CC which follow
HM indicates that HM is a powerful risk factor for CC. Brinton and
colleagues reported that the overall incidence of CC in the United States

72
 Epidemiology

was 1 in 19,920 live births and 1 in 24,096 pregnancies [36]. The risk for
HM developing into CC is probably less than 5% [131].

3.7.4 OTHER FACTORS

Palmer found a six-fold increased risk of CC in women who had taken

OCs for more than five years, but no such association with HM in the
same study [184]. Based on the result of their case control study,
Buckley et al. [207] suggested that below-normal estrogen levels may
predispose to CC. Two different mechanisms were proposed that may
explain these results: hormonal changes may effect oocyte development,
and low estrogen levels may be the result of an ovarian abnormality.
Pregnancies occurring within one year of diagnosis of HM do not
appear to be at higher risk for adverse effects [208-210].

3.7.5 GENETIC FACTORS

As HM has been separated into two clinical and genetic entities (CHM
and PHM), the incidence of developing GTN following the two
separate pathological entities is also different. The incidence of GTN
following a CHM is 2-20% [101, 199, 201], while GTN and CC
following PHM is probably of 0.5% and 0.1%, respectively [201]. The
variable genetic constitution of HM does make it possible to identify
HM and distinguish between a PHM, a monospermic CHM and a
dispermic CHM by using genetic analysis techniques such as cytogenetic
polymorphism, enzyme polymorphism and restriction fragment length
polymorphisms of DNA, and more recently PCR. The relative risk of a
dispermic or a monospermic CHM progressing to a GTN has yet to be
established, although it has been suggested that the dispermic CHM
may have a higher malignancy potential [11, 211]. CC developing in
older women, and certainly women over 50 years of age, is associated
with a poorer prognosis [212-213]. Women who experience repeated
abortion are also at increased risk for CHM [214].

3.7.6. MATERNAL AND PATERNAL BLOOD GROUP

Maternal blood groups of the mother and her consort have been studied
but the data are inconclusive.

ABO group

Blood group A women with incompatible blood group consorts appear

to have a higher risk of CC. Data from the USA [215] and the UK [154]
for patients with CC showed a slight excess in group A above group O.
The UK data suggested that the risk of a woman getting CC is also
influenced by the blood group of her husband, and the effect was most
marked in cases where CC was preceded by term delivery [157].

73
 Epidemiology

In a UK series of 115 patients with CC following term pregnancy or

non-molar abortions at Charing Cross Hospital, for whom mating types
were available, the ratio of incompatible/compatible matings was 2.19,
indicating a predisposition to CC in the incompatible matings [137]. It
is important that more large studies on blood group data should be
done worldwide.

HLA types

Data are available on patients and husbands in a series treated for GTN.
The overall frequencies of the HLA-A and HLA-B locus antigens in 225
Caucasian patients did not differ significantly from those of the normal
control population [216]. However, the degree of incompatibility
between husband and wife, as measured by the number of antigenic
incompatibilities (0, 1 or 2), indicates that there is a trend for patients
who are more compatible with their husbands for the B locus antigens
to fall into the higher-risk treatment categories. Consanguinity as a risk
factor will require evaluation in large-scale studies with modern
cytogenetic techniques.

3.8 TIME TREND OF MORTALITY RATES OF THE GTD

3.8.1 HYDATIDIFORM MOLE

Inadequate management of HM may bring about acute complications,

which results in severe morbidity and mortality. Late complication of
progression to persistent GTN is noted in 57~36% of patients and this
is higher in high risk complete mole.
High risk CHM has base-line hCG over 100,000 milli IU/ml, uterine
size greater than gestation age, ovarian theca lutein cyst over 6cm,
associated medical factors (hyperthyroidism), and maternal age over 40
years or less than 18 years [217-219].
When postmolar evacuation follow-up is inadequate or when GTN
diagnosis is inaccurate, optimal management may not be made in
sufficient time. This results in increase in morbidity and mortality.

3.8.2 GESTATIONAL TROPHOBLASTIC NEOPLASIA (GTN)

Compared with many other human malignant tumors, the decrease in

mortality with chemotherapy treatment is outstanding. This dramatic
improvement in cure rate results from the specific biomarker β-hCG
RIA and sensitive chemo-agents. Currently, low risk GTN has the cure
rate of nearly 100%. However, in spite of tremendous improvement, the
treatment of high risk HM and GTD is still not satisfactory in poorly
developed countries. The mortality of ultra-high risk GTN is around
40~50% [220].
On the global trend, the incidence of GTD is on the decrease, yet the
remission rate of high risk patients is not satisfactory.

74
 Epidemiology

The 5-year survival rate of CC in Japan was very low in the past. It
increased from 23.1% in the 1960s to 52.6% in the 1970s. This has
improved to 86.7% and 85.8% in the 1980s and 1990s [117].
In Korea, the survival rate of GTN patients improved from 73.2% in
1970s, to 92.9% in 1980s, to 96.1% in 1990s, to 96.6% in 2000s [51-52,
64]

3.8.3 OVERALL REMISSION OF GTN

Development of chemotherapy and improvement of general health

services are directly responsible for improved survival. Better
socioeconomic state and improved nutritional conditions are indirect
causes. We hope that studies on genetic factors will shed new light on
our understandings of the disease.
The 26th Annual report from FIGO on Gynecologic Cancer Treatment
shows the total number of GTN has decreased since 1938 (1979-1981).
Patients are mostly 25-29 years old and 82.7% of patients are under the
age of 40, which indicates that this is a tumor of young females [221].
The 5-year survival rate of GTN is 92.7% in the 2000s. However, when
analyzed in depth, the rate is 61.9% for stage IV compared with 97.3%
of stage I, which clearly shows the problem of this disease [221].
Management of recurrent and drug-resistant GTN remains the
biggest challenge in striving towards 100% survival even in the best
centers in the world.

3.9 SUMMARY

Gestational trophoblastic disease is still an important reproductive

health problem worldwide. The problem is that much information of
GTD has come from less developed countries, where proper diagnostic
tools and up-to-date treatment cannot be employed. Maternal age,
previous HM, race and geographical region have been identified as clear
risk factors for GTD. Etiological factors of GTD have long been
studied but no definite causes have yet been found. However, it can be
speculated that during gametogenesis and fertilization, the risk factors
may act synergistically. Genetic sub-classification may be helpful in this
context. Geographic variations of incidence exist but are inextricably
linked with above-mentioned risk factors. Important prerequisites for
accurate evaluation are common denominators, standard classification
and definition of index cases, and same-study methodology. The
development and improvements in suction curettage, termination of
pregnancy, contraceptive techniques, diagnostic imaging and
biochemical testing have been associated not only with a fall in the birth
rate but also with a reduction in the incidence of trophoblastic disease.
Further investigations on possible etiological factors based on case-
control studies in different geographic settings are essential.

75
 Epidemiology

REFERENCES

1. Berkowitz RS, Goldstein DP. (1996) Chorionic tumors [Review]. N

Engl J Med 335:1740-8.
2. Suzuki T, Goto S, Nawa A, Kurauchi O, Saito M, Tomoda Y. (1993)
Identification of the pregnancy responsible for gestational trophoblastic
disease by DNA analysis. Obstet Gynecol 82:629-34.
3. Gestational Trophoblastic Tumor (GTT) (1997). In SGO
Handbook, Staging of Gynecologic Malignancies, 2nd Edition.
4. Benedet JL, Bender H, Jones H 3rd, Ngan HY, Pecorelli S. (2001).
FIGO staging classifications and clinical practice guidelines in the
management of gynecologic cancers. FIGO Committtee on
Gynecologic Oncology. Int J Gynaecol Obstet 70:209-62.
5. Kohorn EI. (2001) The new FIGO 2000 staging and risk factor
scoring system for gestational trophoblastic disease: Description and
critical assessment. Int J Gynecol Cancer 11:73-7.
6. Lage JM, Wolf NG. (1995) Gestational trophoblastic disease. New
approaches to diagnosis. Clin Lab Med 15:631-64.
7. Kajii T, Ohama, K. (1977) Androgenetic origin of hydatidiform mole.
Nature 268:633-4.
8. Wake N, Takagi N, Sasaki M. (1978) Androgenesis as a cause of
hydatidiform mole. J Natl Cancer Inst 60: 51-7.
9. Ohama K, Kajii T, Okamoto E, Fukuda Y, Imaizumi K, Tsukahara
M, Kobayashi K, Hagiwara K. (1981) Dispermic origin of XY
hydatidiform moles. Nature 292: 551-2.
10. Surti U, Szulman AE, O’brian S. (1982) Dispermic origin and
clinical outcome of three complete hydatidiform moles with 46,XY
karyotype. Am J Obstet. Gynecol 144: 84-7.
11. Cheung ANY, Sit ASY, Chung LP, Ngan HYS, O’Hanlan K, Wong
LC, Ma HK. (1994) Detection of heterozygous XY complete
hydatifiorm mole by chromosome in situ hybridization. Gynecol Oncol
55: 386-92.
12. Szulman AE, Surti U. (1978) The syndromes of hydatidiform mole
II. Morphologic evolution of the complete and partial mole. Am J
Obstet Gynecol 132: 20-7.
13. Szulman AE, Philippe E, Boué JG, Boué A. (1981) Human tripoidy:
Association with partial hydatidiform moles and nonmolar conceptuses.
Hum Pathol 12: 1016-21.
14. Womack C, Elston CW. (1985) Hydatidiform mole in Nottingham: a
12-year retrospective epidemiological and morphological study.
Placenta 6: 93-105.
15. Vassilakos P, Riotton G, Kajii T. Hydatidiform mole: Two entities.
(1977) A morphologic and cytogenetic study with some clinical
consideration. Am J Obstet Gynecol 127: 167-70.
16. Jeffers MD, O’Dwyer P, Curran B, Leader M, Gillan JE. (1993)
Partial hydatidiform mole: A common but underdiagnosed condition.
A 3-year retrospective clinicopathological and DNA flow cytometric
analysis. Int J Gynecol Pathol 12: 315-23.
17. Lage JM, Mark SD, Roberts DJ, Goldstein DP, Bernstein MR,
Berkowitz RS. (1992) A flow cytometric study of 137 fresh hydropic
placentas: Correlation between types of hydatidiform moles and nuclear

76
 Epidemiology

DNA ploidy. Obstet Gynecol 79: 403-10.

18. Lawler SD, Fisher RA, Dent J. (1991) A prospective genetic study of
complete and partial hydatidiform moles. Am J Obstet Gynecol 164:
1270-7.
19. Howat AJ, Beck S, Fox H, Harris SC, Hill AS, Nicholson CM,
Williams RA. (1993) Can histopathologists reliably diagnose molar
pregnancy? J Clin Pathol 46: 599-602.
20. Bocklage TJ, Smith HO, Bartow SA. (1994) Distinctive flow
histogram pattern in molar pregnancies with elevated maternal serum
human chorionic gonadotrophin levels. Cancer 73:2782-90.
21. Koenig C, Demopoulos RI, Vamvakas EC, Mittal KR, Feiner HD,
Espiritu EC. (1993) Flow cytometric DNA ploidy and quantitative
histopathology in partial moles. Int J Gynecol Pathol 12: 235-40.
22. Redline RW, Abdul-Karim FW. (1995) The pathology of gestational
trophoblastic disease. Semin Oncol 22: 96-09.
23. Paradinas, FJ, Browne P, Fisher RA, Foskett M, Bagshawe KD,
Newlands E. (1996) A clinical, histopathological and flow cytometric
study of 149 complete moles, 146 partial moles and 107 non-molar
hydropic abortions. Histopathol 28: 101-9.
24. Fukunaga M, Ushigome S, Endo Y. (1995) Incidence of
hydatidiform mole in a Tokyo hospital: a 5-year (1989 to 1993)
prospective, morphological, and flow cytometric study. Hum Pathol
26: 758-64.
25. Soma H, Osawa H, Oguro T, Yoshihama I, Fujita K, Mineo S,
Kudo M, Tanaka K, Afeita M, Urabe M, Kudo Y. (2007) P57kip2
immunohistochemical expression and ultrastructural findings of
gestational trophoblastic disease and related disorders. Med Mol
Morphol 40: 95-102.
26. SJ Kim, SY Lee, C Lee, IH Kim, HJ An, TY Kim, KH baek, EJ
Kim, TM Kim, JB Lee, JW Lee, WW Jung, TH Lhun, YK Oh. (2006)
Differential expression profiling of genes in a complete hydatidiform
mole using cDNA microarray analysis. Gynecol Oncol 103: 654-60.
27. Dehner LP. (1980) Gestational and nongestational trophoblastic
neoplasia: A historic and pathobiologic survey. Am J Surg Pathol 4: 43-
58.
28. Grimes DA. (1984) Epidemiology of gestational trophoblastic
disease. Am J Obstet Gynecol 150: 309-18.
29. Gestational trophoblastic diseases. Report of a WHO Scientific
Group. (1983) World Health Organ Tech Rep Ser 692: 7-81.
30. Takeuchi S. (1982) Nature of invasive mole and its rational
management. Semin Oncol 9: 181-6.
31. Kurman RJ, Scully RE, Norris NJ. (1976) Trophoblastic
pseudotumor of the uterus. An exaggerated form of “syncytial
endometritis” simulating a malignant tumor. Cancer 38: 1214-26.
32. Scully RE, Young RH. (1981) Trophoblastic pseudotumor. A
reappraisal. Am J Surg Pathol 5: 75-6.
33. Feltmate CM, Genest DR, Wise L, Burnstein MR, Goldstein DP,
Berkowitz RS. (2001) Placental site trophoblastic tumor: A 17-year
experience at the New England Trophoblastic Disease Center. Gynecol
Oncol 82: 415-9.
34. Shih IM, Kurman RJ. (1998) Epithelioid trophoblastic tumor: a

77
 Epidemiology

neoplasm distinct from choriocarcinoma and placental site trophoblastic

tumor simulating carcinoma. Am J Surg Pathol 22: 1393-403.
35. Shih IM, Seidman JD, Kurman RJ. (1999) Placental site nodule and
characterization of distinctive types of intermediate trophoblast. Hum
Pathol 30: 687-94.
36. Brinton LA, Bracken MB, Connelly RR. (1986) Choriocarcinoma
incidence in the United States. Am J Epidemiol 123: 1094-100.
37. Rotmensch S, Cole LA. (2000) False diagnosis and needless therapy
of presumed malignant disease in women with false-positive human
chorionic gonadotrophin concentrations. Lancet 355: 712-5.
38. Yingna S, Yang X, Xiuyu Y, Hongzhao S. (2002) Clinical
characteristics and treatment of gestational trophoblastic tumor with
vaginal metastasis. Gynecol Oncol 84: 416-9.
39. Flam F, Rutqvist LE. (1992) Under-registration of gestational
trophoblastic disease in the Swedish Cancer Registry. Eur J Epidemiol
8: 683-6.
40. Hoffman JS, Silverman AD, Gelber J, Cartun R. (1993) Placental site
trophoblastic tumor: A report of radiologic, surgical, and pathologic
methods of evaluating the extent of disease. Gynecol Oncol 50: 110-4.
41. Chan FY, Chau MT, Pun TC, Lam C, Ngan HY, Wong RL. (1995).
A comparison of colour Doppler sonography and the pelvic arteriogram
in assessment of patients with gestational trophoblastic disease. Br J
Obstet Gynaecol 102: 720-5.
42. Mutch DG, Soper JT, Baker ME, Bandy LC, Cox EB, Clarke-
Pearson DL, Hammond CB. (1986) Role of computed axial
tomography of the chest in staging patients with nonmetastatic
gestational trophoblastic disease. Obstet Gynecol 68: 348-52.
43. Wilcox AJ, Weinberg CR, O’Connor JF, Baird DD, Schlatterer JP,
Canfield RE, Armstrong EG, Nisula BC. (1988) Incidence of early loss
of pregnancy. N Engl J Med 319: 189-94.
44. Smith HO, Bedrick EJ, Qualls CR, Rayburn WF, Waxman AG,
Stephens ND, Wiggins C, Hilgers RJ, Key CR. (2001). Multiethnic
differences in gestational trophoblastic disease (GTD) rates in New
Mexico: a 25-year population-based study (2001). Proceedings of the
XIth World Congress on Gestational Trophoblastic Diseases, abstract
3A5, Santa Fe, New Mexico, USA.
45. Smith HO, Prairie BA, Padilla LA, Qualls CR, Key CR. (2002)
Incidence and survival rates for gestational trophoblastic disease (GTD)
in the United States: A SEER (1973-1997) population-based study.
Gynecol Oncol 84: 535.
46. National Co-ordination Research Group of Chorioma. (1980)
Incidence of hydatidiform mole (a retrograde study of 2, 023, 621
women in 23 provinces). Zhongua Yizne Zazhi, 60:641-4.
47. Song H, Wu P. (1987) Hydatidiform mole in China: a preliminary
survey of incidence in more than three million women. Bull WHO 65:
507.
48. Hayashi K, Bracken MB, Freeman DH Jr, Hellenbrand K. (1982)
Hydatidiform mole in the United States (1970-1977): A statistical and
theoretical analysis. Am J Epidemiol 115: 67-77.
49. Palmer JR. (1994) Advances in the epidemiology of gestational
trophoblastic disease. J Reprod Med 9:155-62.

78
 Epidemiology

50. Takeuchi, S. (1984) Incidence of gestational trophoblastic disease by

regional registration in Japan. Hum Reprod 2: 729-34.
51. Kim SJ, Lee C, Kwon SY, Na YJ, Oh YK, Kim CJ.(2004) Studying
Changes in the incidence, diagnosis and management to GTD the South
Korean Model. J Reprod Med 49: 643-54
52. Kim SJ, Na YJ, Jung SG, Kim CJ, Bae SN, Lee C (2007)
Management of High-Risk Hydatidiform Mole and persistent
Gestational Gestational Trophoblastic Neoplasia. J Reprod Med 52:
819-830
53. Aziz MF, Kampono N, Moegni EM, Sjamsuddin S, Barnas B, Samil
RS. (1984) Epidemiology of gestational trophoblastic neoplasm at the
Dr. Cipto Mangunkusumo Hospital, Jakarta, Indonesia. Adv Exp Med
Biol 176:165-75.
54. Márquez-Monter H, de la Vega GA, Ridaura C, Robles M. (1968)
Gestational choriocarcinoma in the general hospital of Mexico: Analysis
of forty cases. Cancer 22:91-8.
55. MacGregor C, Ontiveros EC, Vargas EL, Valenzuela SL. (1969)
Hydatidiform mole: Analysis of 145 patients. Obstet Gynecol 33: 343-
51.
56. Aguero O, Kizer S, Pinedo G. (1973). Hydatidiform mole in
Concepcion Palacios Maternity Hospital. Am J Obstet Gynecol 116:
1117-20.
57. Rolón PA, de Lopez BH. (1977) Epidemiological aspects of
hydatidiform mole in the Republic of Paraguay (South America). Br J
Obstet Gynaecol 84: 862-4.
58. Drake RD, Rao GG, McIntire DD, Miller DS, Schorge JO (2006)
Gestational trophoblastic disease among Hispanic women: a 21-year
hospital-based study. Gynecol Oncol 103:81-6.
59. Martin PM. (1978) High frequency of hydatidiform mole in native
Alaskans. Int J Gynaecol Obstet 15: 395-6.
60. McCorriston CC. (1968) Racial incidence of hydatidiform mole: A
study in a contained polyracial community. Am J Obstet Gynecol
101: 377-82.
61. Natoli WJ, Rashad NM. (1972) Hawaiian moles. Am J Roentgenol
114: 142-4.
62. Matsuura J, Chiu D, Jacobs PA, Szulman AE. (1984) Complete
hydatidiform mole in Hawaii: an epidemiological study. Genetic
Epidemiol 1:271-84.
63. Jacobs PA, Szulman AE, Funkhouser, J. et al. (1982) Human
triploidy: relationship between parental origin of the additional haploid
complement and development of partial hydatidiform mole. Ann Hum
Genet 46: 223-31.
64.Kim SJ, Bae SN, Kim JH, Kim CJ, Han KT, Chung JK, Lee JM.
(1998) Epidemiology and time trends of gestational trophoblastic
disease in Korea. Int J Gynaecol Obstet 60: S33-8.
65. Hando T, Ohno M, Kurose T. (1998) Recent aspects of gestational
trophoblastic disease in Japan. Int J Gynaecol Obstet 60: S71-6.
66. The Joint Project for Study of Choriocarcinoma and Hydatidiform
Mole in Asia. (1954) Geographic variation in the occurrence of
hydatidiform mole and choriocarcinoma. Ann N Y Acad Sci 80:
178-96.

79
 Epidemiology

66. Nakano R, Sasaki K, Yamoto M, Hata H. (1980) Trophoblastic

disease: analysis of 342 patients. Gynecol. Obstet. Invest., 11, 237-42.
68. Sakumoto K, Kigashi M, Kanazawa K. (1999) Hydatidiform mole in
Okinawa Islands and mainland Japan. Int J Gynaecol Obstet 64:
309-10.
69. Llewellen-Jones D. (1965) Trophoblastic tumours: Geographical
variations in incidence and possible aetiological factors. J Obstet
Gynaecol 72: 242-8
70. Wei P-Y, Ouyang P-C. (1963) Trophoblastic disease in Taiwan: A
review of 157 cases in a 10 year period. Am J Obstet Gynecol 85:
844-9.
71. Poen HT, Djojopranoto M. (1965) The possible etiologic factors of
hydatidiform mole and choriocarcinoma: preliminary report. Am J
Obstet Gynecol 92: 510-13.
72. Acosta-Sison, H. (1966) Trophoblastic or chorionic tumors as
observed in the Philippines. Choriocarcinoma from July 4, 1963 to June
30, 1965. Philipp J Surg Spec 21: 41-2.
73. Westerhout Jr FC, Morel ES, Slate WG. (1969) Observations on 138
molar pregnancies. Am J Obstet Gynecol 103: 56-9.
74. Yen S, MacMahon B. (1968) Epidemiologic features of
trophoblastic disease. Am J Obstet Gynecol 101: 126-32.
75. Agboola A, Abudu OO. (1984) Epidemiology of trophoblast
disease in Africa–Lagos. In Human Trophoblastic Neoplasms (ed. RA
Patillo, RO Hussa), Plenum Press, NY, pp. 187-195.
76. Agboola A. (1979) Trophoblastic neoplasia in an African urban
population. J Natl Med Assoc 71: 935-7.
77. Ogunbode O. (1978) Benign Hydatidiform mole in Ibadan, Nigeria.
Int J Gynaecol Obstet 15: 387-90.
78. Junaid TA. (1984) Epidemiological aspects of gestational
trophoblastic disease in Ibadan, Nigeria. In Human Trophoblastic
Neoplasms (ed. RA Patillo, RO Hussa), Plenum Press, NY, pp. 177-85.
79. Akinkugbe A. (1976) Incidence and malignancy rate in hydatidiform
molar pregnancies in Ile-Ife. Nigerian Med J 6: 310-4.
80. Ayangade O. (1979) Gestational trophoblastic disease in Nigeria —
A 10 year review. East Afr Med J 56: 278-82.
81. Osamor JO, Oluwasola AO, Adewole IF (2002) A clinico-
pathological study of complete and partial hydatidiform moles in a
Nigerian population. J Obstet Gynaecol 22:423-5
82. Leighton PC. (1973) Trophoblastic disease in Uganda. Am J Obstet
Gynecol 117: 341-4.
83. Egwuatu VE, Ozumba BC. (1989) Observations on molar
pregnancy in Enugu, Nigeria. Int J Gynaecol Obstet 29: 219-25.
84. Ragland KE, Selvin S, Merrill DW. (1991) White-black differences
in stage-specific cancer survival: Analysis of seven selected sites. Am J
Epidemiol 133: 672-82.
85. Massad LS, Abu-Rustum NR, Lee SS, Renta V. (2000) Poor
compliance with postmolar surveillance and treatment protocols by
indigent women. Obstet Gynecol 96: 940-4.
86. Bracken MB, Brinton LA, Hayashi K. (1984) Epidemiology of
hydatidiform mole and choriocarcinoma. Epidemiol Rev 6: 52-75.
87. Bracken MB. (1987) Incidence and aetiology of hydatidiform mole:

80
 Epidemiology

An epidemiological review. Br J Obstet Gynaecol 94: 1123-35.

88. Yuen BH, Cannon W. (1981) Molar pregnancy in British Columbia:
estimated incidence and post-evacuation regression patterns of the beta
subunit of human chorionic gonadotrophin Am J Obstet Gynecol
139: 316-9.
89. Giwa-Osagie MO, Okwerekwu G. (1999) Epidemiology of molar
pregnancies in Northern Ireland. Int J Gynaecol Obstet 66: 175-7.
90. Nielsen NH, Hansen JPH. (1979) Trophoblastic tumors in
Greenland. J Cancer Res Clin Oncol 95: 177-86.
91. Franke HR., Risse EK, Kenemans P, Vooijs GP, Stolk JG. (1983)
Epidemiologic features of hydatidiform mole in the Netherlands.
Obstet Gynecol 62: 613-6.
92. Bagshawe KD, Dent J, Webb J. (1986) Hydatidiform moles in
England and Wales 1973-1983. Lancet 11: 673-7.
93. Ringertz, N. (1970) Hydatidiform mole, invasive mole and
choriocarcinoma in Sweden 1958-1965. Acta Obstet Gynecol Scand
49: 195-203.
94. Flam F, Lundström-Lindsted V, Rutqvist LE. (1992) Incidence of
gestational trophoblastic disease in Stockholm county, 1975-1988. Eur J
Epidemiol 8: 173-7.
95. Loukovaara M, Pukkala E, Lehtovirta P, Leminen A (2005)
Epidemiology of hydatidiform mole in Finland, 1975 to 2001. Eur J
Gynaecol Oncol 26:207-8.
96. Steigrad SJ. (1969) The incidence of neoplastic trophoblastic disease
in Australia. Aust N Z J Obstet Gynaecol 9: 100-2.
97. Beischer NA, Bettinger HF, Fortune DW, Pepperell R. (1970)
Hydatidiform mole and its complications in the state of Victoria. J
Obstet Gynaecol Br Commonw 77: 263-76.
98. Olesnicky G, Quinn M. (1984) Molar pregnancy after artificial
insemination (donor). Lancet 1(8389): 1296.
99. Di Fabio C., De Aloysio D. (1976) Valutazione statistica di alcuni
dati epidemiologici concernenti la mola idatdea. Riv Ital Ginecol 57:
187—93.
100. Mazzanti P, La Vecchia C., Parazzini F, Bolis G. (1986) Frequency
of hydatidiform mole in Lombardy, Northern Italy. Gynecol Oncol
24: 337-42.
101. Matalon M, Paz B, Modan M, Modan B. (1972) Malignant
trophoblastic disorders: Epidemiologic aspects and relationship to
hydatidiform mole. Am J Obstet Gynecol 112: 101-6.
102. Matalon M, Modan B. (1972) Epidemiologic aspects of
hydatidiform mole in Israel. Am J Obstet Gynecol 112: 107-12.
103. Javey H, Sajadi H. (1978) Hydatidiform mole in Nigeria, Iran, and
Alaska. II. Hydatidiform mole in southern Iran: a statistical survey of
113 cases. Int J Gynaecol Obstet 15: 390-5.
104. Graham IH, Fajardo AM. (1988) The incidence and morphology of
hydatidiform mole in Abu Dhabi, United Arab Emirates 1978—1986.
Br J Obstet Gynaecol 95: 391-2.
105. Ozalp SS, Yalçin OT, Tanir HM. (2001) Hydatidiform mole in
Turkey from 1932 to 2000. Int J Gynaecol Obstet 73: 257-8.
106. Mungan T, Kusçu E, Dabakoglu T, Senöz S, Ugur M, Cobanoglu
O. (1996) Hydatidiform mole: Clinical analysis of 310 cases. Int J

81
 Epidemiology

Gynaecol Obstet 52: 233-6.

107. Gul T, Yilmazturk A, Erden AC. (1997) A review of trophoblastic
diseases at the medical school of Dicle University. Eur J Obstet
Gynecol Reprod Bio 74: 37-40.
108. Felemban AA, Bakri YN, Alkharif HA, Altuwaijri SM, Shalhoub J,
Berkowitz RS. (1998) Complete molar pregnancy. Clinical trends at
King Fahad Hospital, Riyadh, Kingdom of Saudi Arabia. J Reprod
Med 43: 11-3.
109. Khashoggi TY. Prevalence of gestational trophoblastic disease. A
single institution experience (2003) Saudi Med J 24:1329-33.
110. Bugti QA, Baloch N, Baloch MA (2005) Gestational Trophoblastic
Disease in Quetta. Pak J Med Res 44:92-5.
111. Yasin ZT, Chaied HM (2007) Gestational Trophoblastic Disease
in Basrah. MJBU 25:52-6.
112. Anfinan N, Sait K, Sait H (2014) Gestational trophoblastic disease
in the western region of Saudi Arabia (single-institute experience). Eur J
Obstet Gynecol Reprod Biol 17;180C:8-11.
113. Acosta-Sison, H. (1959) Observation which may indicate the
etiology of hydatidiform mole and explain its high incidence in the
Philippines and Asiatic countries. Phil J Surg 14: 290-3.
114. Teoh ES, Dawood MY, Ratnam SS. (1971) Epidemiology of
hydatidiform mole in Singapore. Am J Obstet Gynecol 110: 415-20.
115. Song H, Wu B, Tang M, Wang Y. (1981) Trophoblastic Tumors:
Diagnosis and Management, Capital Hospital, Beijing, China.
116. Nakano R., Sasaki K, Yamoto M, Hata H. (1980) Trophoblastic
disease: analysis of 342 patients. Gynecol Obstet Invest 11: 237-42.
117. Kanazawa K. (1998) Trophoblastic disease: Twenty years’
experience at Niigata University. Int J Gynecol Obstet 60: S97-03.
118. Takeuchi S. (1987) Incidence of gestational trophoblastic disease by
regional registration in Japan. Hum Reprod 729-34.
119. Sivanesaratnam V. (1998) The management of gestational
trophoblastic disease in developing countries such as Malaysia. Int J
Gynaecol Obstet 60: S105-9.
120. Sengupta, B.S., Persaud, V. and Wynter, RN. (1977)
Choriocarcinoma in Jamaica. Int Surg 62: 84-7.
121. Aranda JM, Martinez I. (1968) Choriocarcinoma in Puerto Rico.
A 15-year study. Bol Asoc Med P R 60: 523-35
122. Brisson J, Fabia J. (1976) Incidence de la mole hydatidiforme et des
maladies trophoblastiques malignes l’est du Québec, 1969-1973.
[Translated title: Occurrence of hydatiform mole and malignant
trophoblastic disease in eastern Quebec, 1969-1973]. Union Med Can
105: 1238-40.
123. Chan, DPC. (1962) Chorion epithelioma: a study of 42 cases. Br
Med J 2: 853-7.
124. Pai KN. (1967) A study of choriocarcinoma: its incidence in India
and its aetiopathogenesis. In Choriocarcinoma: Transactions of a
Conference of the International Union Against Cancer (Eds JM.
Holland, M Hreshchyshyn), Springer-Verlag, Berlin, p.54—7.
125. Ong HC, Lee, PYA, Ng TKF, Chong CH. (1978) Clinical
observation of hydatidiform mole in a Malaysian hospital. Singapore
Med J 19: 33-6.

82
 Epidemiology

126. Srivannaboon S, Vatananusara C, Boonyanit S. (1974) The

incidence of trophoblastic disease in Siriraj Hospital. J Med Assoc
Thai 57: 537-42.
127. Brewer JI, Gerbie AB. (1967) Early development of
choriocarcinoma, In: Choriocarcinoma: Transactions of a Conference of
the International Union Against Cancer (Eds JM Holland, M
Hreshchyshyn), Springer-Verlag, Berlin, pp. 45-53.
128. Hertig AT, Mansell H. (1956) Tumors of the Female Sex Organs
Part I Hydatiform Mole and Choriocarcinoma, In: Atlas of Tumor
Pathology, Section IX, Fascicle 33, Washington, D.C, Armed Forces
Institute of Pathology.
129. Shanmugaratnan K, Muir CS, Tow SH, Cheng WC, Christine B,
Peuersen E. (1971) Rates per 100,000 births and incidence of
choriocarcinoma and malignant mole in Singapore Chinese and Malays:
Comparison with Connecticut, Norway, and Sweden. Int J Cancer 8:
165-75.
130. Song HZ, Yang XY, Xiang Y. (1996) Forty-five years’ experience
of the treatment of choriocarcinoma and invasive mole. Int J Gynecol
Obstet 60: S77-83.
131. Park WW, Lee JC. (1950) Choriocarcinoma: A general review, with
analysis of 516 cases. AMA Arch Pathol 49: 204-41.
132. Wang T-H, Wang H-S. (1995) Gestational trophoblastic diseases:
Current trends and perspectives. J Formos Med Assoc 94: 449-57.
133. Okudaira, Y. and Strauss, L. (1967) Ultrastructure of molar
trophoblast: observations on hydatidiform mole and chorioadenoma
destruens. Obstet Gynecol 30: 172-87.
134. Notake, Y. (1963) Atlas of Tissue Culture in Obstetrics and
Gynaecology, Japan Obstetrics and Gynaecology Society, Tokyo.
135. Chattopadhyay SK, Sengupt BS, al-Ghreimil M, Edrees YB,
Lambourne A. (1988) Epidemiologic study of gestational trophoblastic
diseases in Saudi Arabia. Surg Gynecol Obstet 167: 393-8.
136. Stone M., Dent J, Kardana A, Bagshawe KD. (1976) Relationship
of oral contraception to development of trophoblastic tumor after
evacuation of a hydatidiform mole. Br J Obstet Gynaecol 83: 913-6.
137. Bagshawe KD, Lawler SD. (1982) Choriocarcinoma. In: Cancer
Epidemiology and Prevention (eds D Schottenfeld, JF Fraumeni),
Saunders, Philadelphia, pp. 909—24.
138. Martin BH, Kim JH. (1998) Changes in gestational trophoblastic
tumors over four decades: A Korean experience. J Reprod Med 43:
60-8
139. Messerli ML, Lilienfeld AM, Parmley I, Woodruff JD, Rosenshein
NB. (1985) Risk factors for gestational trophoblastic neoplasia. Am J
Obstet Gynecol 153: 294-300.
140. Sasaki K, Yamoto M, Hata H and, Nakano R. (1979) The
relationship between maternal aging and the incidence of hydatidiform
mole and malignant changes. Acta Obstet Gynecol Jpn 31: 292-6.
141. Acosta-Sison J. (1959) The chance of malignancy in a repeated
hydatidiform mole. Am J Obstet Gynecol 78: 876-7.
142. La Vecchia C, Francheschi S, Parazzini F. Fasoli M, Decarli A,
Gallus G, Tognoni G. (1985) Risk factors for gestational trophoblastic
disease in Italy. Am J Epidemiol 121: 457-64.

83
 Epidemiology

143. Berkowitz RS, Cramer DW, Bernstein MR, Cassells S, Driscoll SG,
Goldstein DP. (1985) Risk factors for complete molar pregnancy from a
case control study. Am J Obstet Gynecol 152: 1016-20.
144. Sand PK, Lurain JR, Brewer JI. (1984) Repeat gestational
trophoblastic disease. Obstet Gynecol 63: 140-4.
145. Federschneider JM, Goldstein DP, Berkowitz RS, Marean AR,
Bernstein MR. (1980) Natural history of recurrent molar pregnancy.
Obstet Gynecol 55: 457-9
146. Kim SJ. (1982) Difficult cases of trophoblastic disease and its
classification and management. Semin Oncol 9: 234-8.
147. Be George FV. (1981) Hydatidiform mole in other pregnancies of
mothers of twins. Am J Obstet Gynecol 108: 369-71.
148. Parazzini F, La Vecchia C, Pampallona S, Francheschi S. (1985)
Reproductive patterns and the risk of gestational trophoblastic disease.
Am J Obstet Gynecol 152: 866-70.
149. Tuncer ZS, Bernstein MR, Wang J, Goldstein DP, Berkowitz RS.
(1999) Repetitive hydatidiform mole with different male partners.
Gynecol Oncol 75: 224-6.
150. Pai L, Toth TL, Leykin L, Isaacson KB. (1996) High incidence of
triploidy in in-vitro fertilized oocytes from a patient with a previous
history of recurrent gestational trophoblastic disease. Hum Reprod 11:
1529-32.
151. Ho HN, Gill TJ 3rd, Klionsky B, Ouyang PC, Hsieh CY, Seski J,
Kunschner A. (1989) Differences between white and Chinese
populations in human leukocyte sharing and gestational trophoblastic
tumors. Am J Obstet Gynecol 161: 942-8.
152. Lorigan PC, Sharma S, Bright N, Coleman RE, Hancock BW.
(2000) Characteristics of women with recurrent molar pregnancies.
Gynecol Oncol 78: 288-92.
153. Parsons, L. and Sommers, S.C. (1978) Genetics of Hydatidiform
Mole, In: Trophoblastic Disease Gynecology, 2nd Edn., Saunders,
Philadelphia, pp. 534—5.
154. Iliya FA, William S, Azar HA. (1967) Choriocarcinoma in the near
east: Consanguinity as a possible etiologic factor. Cancer 20: 144-9.
155. Chun D, Braga C, Chow C, Lok L. (1964) Clinical observations on
some aspects of hydatidiform moles. J Obstet Gynaecol Br Commonw
71: 180-4.
156. Ambani LM, Vaidya RA, Rao CS., Daftary SD, Motawshaw ND.
(1980) Familial occurrence of trophoblastic disease: report of recurrent
molar pregnancies in sisters in three families. Clin Genet 18: 27-9.
157. Bagshawe, KD, Rawlins G, Pike MC, Lawler SD. (1971) ABO
blood-groups in trophoblastic neoplasia. Lancet 1:(7699) 553-6.
158. Tomoda Y. Kaseki S, Goto S, Nishi H, Hara T, Naruki M. (1981)
Rh-D factor in trophoblastic tumor: A possible cause of high incidence
in Asia. Am J Obstet Gynecol 139: 742-3.
159. Lawler SD, Fisher RA, Pickthall VJ, Povey S, Evans MW. (1982)
Genetic studies on hydatidiform moles. I. The origin of partial moles.
Cancer Genet Cytogenet 5: 309-20.
160. Vassilakos P, Riotton G, Kajii T. (1977) Hydatidiform mole: two
entities. A morphologic and cytogenetic study with some clinical
considerations. Am J Obstet Gynecol 127: 167-70.

84
 Epidemiology

161. Yamashita K, Wake N, Araki T, Ichinoe K, Makoto K. (1979)

Human lymphocyte antigen expression in hydatidiform mole:
Androgenesis following fertilization by a haploid sperm. Am J Obstet
Gynecol 135: 597-0.
162. Szulman AE, Surti U. (1982) The clinicopathologic profile of the
partial hydatidiform mole. Obstet Gynecol 59: 597-602.
163. Romero R, Horgan JG, Kohorn EI, Kadar N, Taylor KJW,
Hobbins JC. (1985) New criteria for the diagnosis of gestational
trophoblastic disease. Obstet Gynecol 66: 553-8.
164. Lage JM, Weinberg DS, Yavner DL, Bieber FR. (1989) The
biology of tetraploid hytatidiform moles: Histopathology, cytogenetics,
and flow cytometry. Hum Pathol 20: 419-25.
165. Fisher RA. Newlands ES. (1993) Rapid diagnosis and classification
of hydatidiform motes with polymerase chain reaction. Am J Obstet
Gynecol 168: 563-9.
166. Azuma C, Saji F, Tokugwa Y, Kimura T, Nobunaga T, Takemura
M, Kameda T, Tanizawa O. (1991) Application of gene amplification by
polymerase chain reaction to genetic analysis of molar mitochondrial
DNA: The detection of anuclear empty ovum as the cause of complete
mole. Gynecol Oncol 40: 29-33.
167. Fulop V, Colitti CV, Genest D, Berkowitz RS, Yiu GK, Ng S,
Szepesi J, Mok SC. (1998) DOC-2/hDab2, a candidate tumor
suppressor gene involved in the development of gestational
trophoblastic diseases. Oncogene 17: 419-24.
168. Dumur CI, Koritschoner NP, Flury A, Panzetta-Dutari G, Bocco
JL, Patrito LC. (1998) Differential expression of a tumor necrosis factor
receptor-related transcript in gestational trophoblastic diseases in
women. Biol Reprod 59: 621-5.
169. Walsh C, Miller SJ, Flam F, Fisher RA, Ohlsson R. (1995)
Paternally derived H19 is differentially expressed in malignant and
nonmalignant trophoblast. Cancer Research 55: 1111-6.
170. Kim SJ, Park SE, Lee C, Lee SY, Kim IH, Ahn HJ and Oh YK.
Altered imprinting, promoter usage, and expression of insulin-like
growth factor-II in gestational tropoblastic diseases. Gynecol Oncol
88: 411-418.
171. Fulop V, Mok SC, Genest DR, Gati I, Doszpod J, Berkowitz RS.
(1998) p53, p21, Rb and mdm2 oncoproteins: Expression in normal
placenta, partial and complete mole, and choriocarcinoma J Reprod
Med 43: 119-27.
172. Shaarawy M, Darwish NA, Abdel-AZIZ O. (1996). Serum
interleukin-2 and soluble interleukin-2 receptor in gestational
trophoblastic diseases. J Soc Gynecol Investig 3: 39-46.
173. Arima T, Imamura T, Sakuragi N, Higashi M, Kamura T, Fujimoto
S, Nakano H, Wake N. (1995) Malignant trophoblastic neoplasms with
different modes of origin. Cancer Genet Cytogenet 85: 5-15.
174. Martin DA, Sutton GP, Ulbright TM, Sledge Jr GW, Stehman FB,
Ehrlich C. (1989) DNA content as a prognostic index in gestational
trophoblastic neoplasia. Gynecol Oncol 34: 383-8.
175. van de Kaa CA, Schijf CPT, de Wilde PCM, Hanselaar AGJM,
Vooijs PG. (1997) The role of deoxyribonucleic acid image cytometric
and interphase cytogenetic analyses in the differential diagnosis,

85
 Epidemiology

prognosis, and clinical follow-up of hydatidiform moles: A report from

the central molar registration in the Netherlands. Am J Obstet
Gynecol 177: 1219-29.
176. Mark FL, Ahearn J, Lathrop JC. (1995) Constitutional trisomy 8
mosaicism and gestational trophoblastic disease. Cancer Genet
Cytogenet 80: 150-4.
177. Olsen JH, Mellemkjaer L, Gridley G, Brinton L, Johansen C, Kjaer
SK. (1999) Molar pregnancy and risk for cancer in women and their
male partners. Am J Obstet Gynecol 181: 630-4.
178. Marquez-Monter H, de la Vega GA, Robles M, Boljo-Cicero A.
(1963) Epidemiology and pathology of hydatidiform mole in the
General Hospital of Mexico: Study of 104 cases. Am J Obstet Gynecol
85: 856-64.
179. Hsu C, Chen TY, Chiu WH et a1. (1964) Some aspects of
trophoblastic disease peculiar to Taiwan. Am J Obstet Gynecol 92:
308-16.
180. Pour-Reza M, Ageli N, Vaghefi SB. (1974) Serum creatine, urea
and protein level changes in hydatidiform mote. JAMA 230: 580-1.
181. Parazzini, F., La Vecchia, C, Mangili, G. Caminiti C, Negri E,
Cecchetti G, Fasoli M. (1988) Dietary factors and the risk of
trophoblastic disease. Am J Obstet Gynecol 158: 93-100.
182. Brinton LA, Wu BZ, Wang W, Ershow AG, Song HZ, Li JY,
Bracken MB, Blot WJ. (1989) Gestational trophoblastic disease: A case-
control study from the People’s Republic of China. Am J Obstet
Gynecol 161: 121-7.
183. Palmer JR. (1991) Oral contraceptive use and gestational
choriocarcinoma. Cancer Detect Prev 15: 45-8.
184. Palmer JR, Driscol SG, Rosenberg L, Berkowitz RS, Lurain JR,
Soper J, Twoggs LB, Gershenson DM, Kohorn EI, Berman M, Shapiro
S, Rao RS. (1999) Oral contraceptive use and risk of gestational
trophoblastic tumors. J Natl Cancer Inst 91: 635-40.
185. Stone M, Bagshawe KD. (1979) An analysis of the influence of
maternal age, gestational age, contraceptive method, and the mode of
primary treatment of patients with hydatidiform moles on the incidence
of subsequent chemotherapy. Br J Obstet Gynaecol 86: 782-92.
186. Berkowitz RS, Goldstein DP, Marean AR, Bernstein M. (1981)
Oral contraceptives and postmolar trophoblastic disease. Obstet
Gynecol 58: 474-7.
187. Curry SL, Schlaerth JB, Kohorn EI, Boyce JB, Gore H, Twiggs LB,
Blessing JA. (1989) Hormonal contraception and trophoblastic sequelae
after hydatidiform mole (A Gynecologic Oncology Group study). Am J
Obstet Gynecol 160: 805-11.
188. Deicas RE, Miller DS, Rademaker AW, Lurain JR. (1991) The role
of contraception in the development of postmolar gestational
trophoblastic tumor. Obstet Gynecol 78: 221-6.
189. Schmidt G, Fowler WC, Talbert LM, et al. (1980) Reproductive
history of women exposed to diethylstilbestrol in utero. Fert Steril 33:
21-4.
190. Constable JD, Hatch, MC. (1985) Reproductive effects of herbicide
exposure in Vietnam: recent studies by the Vietnamese and others.
Teratog Carcinog Mutagen 5: 231-50.

86
 Epidemiology

191. Ha MC, Cordier S, Bard D, Le TB, Hoang AH, Hoang TQ, Le

CD, Abenhaim L, Nguyen TN. (1996). Agent orange and the risk of
gestational trophoblastic disease in Vietnam. Arch Environ Health 51:
368-74.
192. Sterling TD, Arundel A. Review of recent Vietnamese studies on
the carcinogenic and teratogenic effects of phenoxy herbicide exposure.
Int J Health Serv 16: 265-78.
193. Ujeno Y. (1985) Epidemiological studies of human fetal
development in area with various doses of natural background radiation.
Arch Environ Health 40: 181-4.
194. Jacobs PA, Hunt PA, Matsurra JS, Wilson CC. (1982) Complete
and partial hydatidiform mole in Hawaii: Cytogenetics, morphology, and
epidemiology. Br J Obstet Gynaecol 89: 258-66.
195. Parazzini F, La Vecchia C, Pampallona S. (1986) Parental age and
risk of complete and partial hydatidiform mole. Br J Obstet Gynaecol
93: 582-85.
196. Berkowitz RS, Bernstein MR, Harlow BL, Rice LW, Lage JM,
Goldstein DP, Cramer DW. (1995) Case-control study of risk factors
for partial molar pregnancy. Am J Obstet Gynecol 173: 788-94.
197. Steigrad SJ. (2003) Epidemiology of gestational trophoblastic
disease. Best Pract Res Clin Obstet Gynaecol 17: 837-47.
198. Altieri A, Rranceschi S, Ferlay J, Smith J, La Vecchiac. (2003)
Epidemiology and aetiology of gestational trophoblastic disease. Lancet
Oncol 4: 670-8.
199. Zalel Y, Dgani R. (1997) Gestational trophoblastic disease
following the evacuation of partial hydatidiform mole: a review of 66
cases. Eur J Obstet Gynecol Reprod Biol 71: 67-71.
200. Berkowitz RS, Goldstein DP, Bernstein MR. (1983) Natural history
of partial molar pregnancy. Obstet Gynecol 66: 677-81.
201. Chen R-Y, Huang S-C, chow S-N, Hsier C-Y, Hsu H-C. (1994)
Persistent gestational trophoblastic tumour with partial hydatidiform
mole as the antecedent pregnancy. Br J Obstet Gynaecol 101: 330-4.
202. Menczer J, Girtler O, Zajdel L, Glenzerman M. (1999) Metastatic
trophoblastic disease following partial hydatidiform mole: Case report
and literature review. Gynecol Oncol 74: 304-7.
203. Zalel Y. Dgani R. (1997) Gestational trophoblastic disease
following evacuation of partial hydatidiform mole: A review of 66 cases.
Eur J Obstet Gynecol 77: 67-71.
204. Seckl MJ, Fisher RA, Salerno G, Rees H, Paradinas FJ, Foskett M,
Newlands ES. (2000) Choriocarcinoma and partial hydatiform moles.
Lancet 356: 36-9.
205. Walden PA, Bagshawe KE. (1976) Reproductive performance of
women successfully treated for gestational trophoblastic tumors. Am J
Obstet Gynecol 125: 1108-14.
206. Baltazar JC. (1976) Epidemiological features of choriocarcinoma.
Bull. WHO 54: 523-32.
207. Buckley JB, Henderson BE, Morrow CP, Hammond CB, Kohorn
EI, Austin CF. (1988) Case-control study of gestational
choriocarcinoma. Cancer Res 48: 1004-10.
208. Tuncer ZS, Bernstein MR, Goldstein DP, Lu KH, Berkowitz RS.
(1999) Outcome of pregnancies occurring within 1 year of hydatidiform

87
 Epidemiology

mole. Obstet Gynecol 94: 588-90.

209. Kim JH, Park DC, Bae SN, Namkoong SE, Kim SJ. (1998)
Subsequent reproductive experience after treatment for gestational
trophoblastic disease. Gynecol Oncol 71: 108-12.
210. Berkowitz RS, Tuncer ZS, Bernstein MR, Goldstein DP. (2000)
Management of gestational trophoblastic diseases: Subsequent
pregnancy experience. Semin Oncol 27: 678-85.
211. Cho S, Kim SJ. (1993) Genetic study of hydatidiform moles by
restriction fragment length polymorphisms (RFLPs) analysis. J Korean
Med Sci 8: 446-52.
212. Tsukamoto N, Iwasaka T, Kashimura Y., Uchino H, Kasimura M,
Matsuyama T. (1985) Gestational trophoblastic disease in women aged
50 or more. Gynecol Oncol 20: 53-61.
213. Davidson SA, Gottesfeld J, La Rosa FG. (1998) Molar pregnancy
in a 60-year-old woman. Int J Gynecol Obstet 56: 53-5.
214. Acaia B, Parazzini F, La Vecchia C, Ricciardiello O, Fedele L,
Candiani GB. (1988) Increased frequency of complete hydatidiform
mole in women with repeated abortion. Gynecol Oncol 31: 310-4.
215. Scott JS. (1962) Choriocarcinoma: Observation on the etiology.
Am J Obstet Gynecol 83: 185-93.
216. Lawler SD. (1978) HLA and trophoblastic tumors Brit Med Bull
34: 305-8.
217. Goto S, Yamada A, Ishizuka T. et al (1993) Development of
postmolar trophoblastic disease after partial molar pregnancy. Gynecol
Oncol 48: 165-170.
218. Berkowitz RS, Goldstein DP. (1987) Management of complete
molar pregnancy. J Reprod Med 32: 634-639.
219. Berkowitz RS, Goldstein DP. (1986) Management of molar
pregnancy and gestational trophoblastic tumors. Int Gynecol Oncol
425-443.
220. Kim SJ. (2003) Placental site trophoblastic tumour. Best Pract Res
Clin Obstet Gynaecol 17: 969-84.
221. Ngan HY, Odicino F, Maisonneuve P, Creasman WT, Beller U,
Quinn MA, Heintz AP, Pecorelli S, Benedet JL (2006) Gestational
trophoblastic neoplasia. FIGO 26th Annual Report on the Results of
Treatment in Gynecological Cancer. Int J Gynaecol Obstet 95(S1):S193-
203

88