European Review for Medical and Pharmacological Sciences 2013; 17: 3381-3384

The impact of ultrasonographic lesion size

and initial human chorionic gonadotropin
values on treatment success in cases with
complete hydatidiform mole
K.D. SECKIN, E. BASER, I. YERAL, C. TOGRUL, B. OZDAL, T. GUNGOR

Department of Gynecologic Oncology, Zekai Tahir Burak Women's Health Training and
ResearchHospital, Ankara, Turkey

Abstract. – OBJECTIVE: Our aim to assess with the availability of highly sensitive human
the impact of sonographically measured lesion chorionic gonadotropin (hCG) assays and ultra-
size and initial human chorionic gonadotropin
levels on treatment success in cases of com-
sound, many cases of CHM are diagnosed and
plete hydatidiform mole (CHM). treated earlier2. This results in a great decrease of
PATIENTS AND METHODS: Patients with CHM the aforementioned complications.
diagnosed between January 2007 and January CHM incidence varies among different coun-
2012 were included in the study. Clinical parameters tries across the world3-5. The rarity of the disease
such as patient age, fertility history, smoking, alco- has resulted in inadequacy of evidence-based da-
hol consumption, presenting symptom, pregnancy
duration, ultrasonographic mean lesion size, beta- ta for development of appropriate management
hCG level on admission, primary treatment method, strategies. Most of our clinical practice is based
beta-hCG normalization time (NT) and adjuvant upon data from single or multi-center descriptive
treatments were recorded and analyzed. reports. Especially optimal method of post-treat-
RESULTS: One hundred-twelve cases of CHM ment surveillance is controversial at the time, due
were identified in the study period. Mean patient to lack of randomized controlled trials.
age was 27.3 ± 8.2 years. Suction curettage was
employed as the primary treatment in all of the In this paper, we present the clinical analysis of
study cases. No perioperative complications CHM cases that were diagnosed and treated in a
were encountered. None of the patients were tertiary referral hospital in Ankara, Turkey within a
treated with prophylactic adjuvant chemothera- five-year period, in an effort contribute to the limit-
py. Twelve patients (10.7%) required adjuvant ed data in the literature on this rare disease.
chemotherapy. Beta-hCG NT did not have an as-
sociation with patient age and initial beta-hCG
levels (p > 0.05). Also, patient age, gravidity, pari-
ty, smoking, initial beta-hCG and ultrasono- Patients and Methods
graphic mean lesion size did not predict adju-
vant chemotherapy requirement (p > 0.05). This retrospective study was performed in the
CONCLUSIONS: Early detection and treatment Gynecologic Oncology Department of Zekai
of CHM is associated with a favorable clinical
outcome. Tahir Burak Women’s Health Education and Re-
search Hospital in Ankara, Turkey, after gaining
Key Words: approval from the Institutional Review Board
Complete hydatidiform mole, Gestational tro- (Date 27.03.2012/ No: 12). Hospital pathology
phoblastic disease, Suction curettage, Chemotherapy. database was searched for CHM cases that were
diagnosed between January 2007 and January
2012. For each case, patient age, fertility history,
Introduction smoking, alcohol consumption, presenting symp-
tom on admission, pregnancy duration upon ad-
Complete hydatidiform mole (CHM) is associ- mission, ultrasonographic mean lesion size, quan-
ated with many complications such as severe titative beta sub-unit of human chorionic go-
vaginal bleeding, formation of theca lutein cysts, nadotropin (beta-hCG) level on admission, prima-
early development of preeclampsia, hyperthy- ry treatment method, beta-hCG normalization
roidism and malignant transformation1. Today, time (NT) and adjuvant treatments were recorded.

Corresponding Author: Kerem Doga Seckin, MD; e-mail: [email protected] 3381

 K.D. Seckin, E. Baser, I. Yeral, C. Togrul, B. Ozdal, T. Gungor

Mean ultrasonographic lesion sizes were cal- Forty-four cases (39.3%) were asymptomatic
culated with the formula: largest antero-posterior on admission. Sixty cases (53.6%) presented
(AP) + superior-inferior (SI) lesion dimension di- with vaginal bleeding, and 8 cases (7.1%) pre-
vided by two. Serum beta-hCG measurements sented with pelvic pain. Mean pregnancy dura-
were performed using a commercial kit (ADVIA tion according to last menstrual period (LMP)
Centaur, Siemens Medical Solutions Diagnostics, upon admission was 72.4±10.4 days. Mean ultra-
Tarry Town, NY, USA). NT was defined as the sonographic lesion size was 65.3±22.7 mm.
time period in days, starting from the day of first Mean and median beta-hCG levels upon admis-
intervention until detection of beta-hCG below 5 sion were 149,349 and 101,579 mIU/ml (mini-
mIU/ml. Patients were divided into three groups mum 551- maximum 844,441), respectively.
according to age (Group 1: < 20 years, Group 2: Suction curettage was employed as the primary
20-35 years, Group 3 > 35 years). Mean present- treatment in all of the study cases. No perioperative
ing hCG, pregnancy duration, ultrasonographic complications such as uterine perforation, excessive
lesion size, and hCG normalization time values bleeding or infection were encountered in these
were compared between these groups. procedures. None of the patients were treated with
prophylactic adjuvant chemotherapy. Mean and me-
Statistical Analysis dian NT of beta-hCG were 60.6 ± 23.4 and 60 days
Study data were analyzed using SPSS version (minimum 27 – maximum 158), respectively. Mean
17.0 (SPSS Inc., Chicago, IL, USA). Categorical Beta-hCG NT in cases that required adjuvant
variables were presented numerically and by per- chemotherapy versus cases that did not were 82 ±
centage, and continuous variables were presented 48.3 and 58 ± 17.6 days, respectively.
as mean ± standard deviation (SD). Correlation Cases were divided into three groups accord-
analysis (Pearson) was used to determine the asso- ing to patient age. Mean presenting hCG, preg-
ciation between patient age, initial beta-hCG, nancy duration, ultrasonographic lesion size, and
pregnancy duration and NT. Logistic regression hCG normalization time values were similar be-
analysis was used to determine the effect of pa- tween the three groups (p > 0.05) (Table I).
tient age, gravidity, parity, smoking, initial beta- Twelve patients (10.7%) required adjuvant
hCG and ultrasonographic mean lesion size on chemotherapy. Ten of these received single agent
post-treatment chemotherapy for treatment failure (weekly intramuscular methotrexate (Mtx) – 50
or relapse. Kaplan-Meier analysis was performed mg/m2) for persistent disease and two cases with
to demonstrate cumulative beta-hCG normaliza- subsequent development of gestational trophoblas-
tion curve. One-way ANOVA method was used to tic neoplasia (GTN) and lung metastasis required
compare mean presenting hCG, pregnancy dura- multi-agent (EMA/CO-etoposide, Mtx, dactino-
tion, ultrasonographic lesion size, and hCG nor- mycin, cyclophosphamide and vincristine)
malization time values between age groups. p < chemotherapy. Mean number of Mtx cycles need-
0.05 was considered statistically significant. ed for remission was 6.3 ± 2.5 (minimum 2-maxi-
mum 9). EMA/CO regimen was instituted as the
Results frontline chemotherapy after the detection of lung
metastasis in a 24 years-old and 28 years-old pa-
In total, 112 cases of CHM were identified with- tients. In this cases, beta-hCG was normalized af-
in the study period. None of the study patients ter two cycles of EMA/CO. Additional three cy-
were lost to follow-up. Mean patient age was cles were administered for consolidation. There
27.3±8.2 years. Mean gravidity, parity, and abor- were no major chemotherapy related complica-
tion frequencies were 2.9, 1.5 and 0.3, respectively. tions and none of the cases had vaginal metastasis.
Forty-four cases (39.3%) were primigravidas, None of the cases relapsed after chemotherapy.
whereas 68 cases (60.7%) had a previous pregnan- Beta-hCG NT did not have an association with
cy. Sixty-four cases (57.1%) had at least one prior patient age, pregnancy duration and initial beta-
live birth. Twenty-two cases (19.6%) had a prior hCG levels (p > 0.05) (Table II). Also, patient
abortion and 14 of these had dilatation and curet- age, gravidity, parity, smoking, pregnancy dura-
tage (D&C) for treatment. None of the cases had tion, initial beta-hCG and ultrasonographic mean
history of a previous gestational trophoblastic dis- lesion size did not predict adjuvant chemotherapy
ease (GTD). Only six patients in the study popula- requirement (p > 0.05). Cumulative beta-HCG
tion were smokers (5.4%). None of the patients normalization curve of the study cases is present-
had regular alcohol consumption. ed in Figure 1.

3382
 Treatment success in complete hydatidiform mole

Table I. Clinical parameters according to patient age groups.

Age < 20 years Age 20-35 years Age > 35 years

Parameters n = 20 n = 74 n = 18 p*

Initial beta-hCG ± SD†, (mIU/ml) 119952.5 ± 141776 166006.1 ± 187432 113538.4 ± 38857 > 0.05
Mean pregnancy duration (days) ± SD† 67.8 ± 8.9 72.6 ± 10.0 77.1 ± 12.4 > 0.05
Mean ultrasonographic lesion size (mm) ± SD† 63.5±26.4 64.0 ± 22.8 72.7 ± 18.4 > 0.05
Mean beta-hCG normalization time ± SD† 49.3±11.7 62.9 ± 25.4 63.5 ± 22.3 > 0.05

†
SD: Standard deviation.

Table II. The correlation between patient age, pregnancy duration, initial beta-hCG and beta-hCG normalization time (NT).

Variable Mean Min Max p*

Age (years) 27.4 17 52 0.907

Pregnancy duration (days) 72.4 54 91 0.461
Initial beta-hCG (mIU/ml) 149.349 551 844.441 0.066

*p value for correlation with beta-HCG normalization time (NT).

Discussion effect on treatment failure. A previous report dis-

cussing imaging practices in GTD has indicated
Complete hydatidiform mole is a localized and the recent advances in technology10,11. Although
non-invasive proliferative trophoblastic disorder these improvements may lead to increased diag-
with various incidences reported throughout the nostic sensitivity, cost-effectiveness and global
world3-5. In a previous study, the incidence of feasibility must be assessed before integrating
CHM was reported as 0.5 in 1000 pregnancies these technologies to routine clinical manage-
and 0.8 in 1000 deliveries6. ment. Considering the widespread availability of
Two main factors affecting GTD occurrence is ultrasound in many institutions, future studies on
maternal age and previous history of GTD. Mean sonographic characterization of GTDs may be
maternal age in our study was 27.3 years, which promising for increasing diagnostic accuracy.
is compatible with previous reports3,7. The risk of All of the cases in our report were primarily
CHM is increased significantly in extremes of treated with suction curettage, without any associ-
maternal age (i.e. < 20 and > 35)7,8. Nevertheless, ated complications. As mentioned earlier, this may
most cases of CHM are diagnosed in women be attributable to early intervention, which is most-
aged less than 35 years, due to increased birth ly preceded by early detection of a non-viable
rates in this period. In our study population, 20
patients (19.6%) were under 20 years and 18 pa-
tients (17.8%) were over 35 years of age. None
of the study patients had a history of previous
GTD. Due to the increased risk of GTD occur-
rence, patients with a history of CHM must be
observed closely during subsequent pregnancies.
Patients

A significant number of cases in our popula-

tion were diagnosed in the early pregnancy peri-
od (approximately 10 weeks according to LMP).
Similar with previous studies, most cases pre-
sented with vaginal bleeding9. However, 39.3%
of cases were asymptomatic at the time of diag-
nosis, which underlines the importance of first
trimester ultrasound to confirm fetal viability. In Days to BHCG normalization
this work, ultrasonographic mean lesion sizes
were determined, and were found not to have an Figure 1. Days to BHCG normalization.

3383
 K.D. Seckin, E. Baser, I. Yeral, C. Togrul, B. Ozdal, T. Gungor

pregnancy. None of the cases received prophylactic –––––––––––––––––-––––

Conflict of Interest
chemotherapy after primary treatment and only The Authors declare that there are no conflicts of interest.
10% of them required adjuvant treatment.
Median normalization period (NT) after treat-
ment was 60 days in our study. In a previous pa- References
per by Yuen et al12 , it was reported that 50% of
the cases had elevated levels of hCG 60 days af- 1) ALTIERI A, FRANCESCHI S, FERLAY J, SMITH J, LA VECCHIA C.
ter primary treatment. In women that did not un- Epidemiology and aetiology of gestational tro-
dergo hysterectomy, there even were elevated phoblastic diseases. Lancet Oncol 2003; 4: 670-678.
levels in a maximum of 219 days. In our study, a 2) ALTMAN AD, BENTLEY B, MURRAY S, BENTLEY JR. Mater-
nal age-related rates of gestational trophoblastic
maximum time interval after treatment was 158 disease. Obstet Gynecol 2008; 112: 244-250.
days. Kizaki et al13 referred that analysis of beta- 3) DI CINTIO E, PARAZZINI F, ROSA C, CHATENOUD L, BENZI
hCG regression curves is a possible method for G. The epidemiology of gestational trophoblastic
GTN risk assessment. In this report, the authors disease. Gen Diagn Pathol 1997; 143: 103-108.
reported approximate beta-hCG NTs in CHM as 4) E LIAS KM, S HONI M, B ERNSTEIN M, G OLDSTEIN DP,
9.3 weeks, which is very close to our data (8.5 B ERKOWITZ RS. Complete hydatidiform mole in
weeks). Also in our paper, NT was significantly women aged 40 to 49 years. J Reprod Med 2012;
longer in cases that required adjuvant chemother- 57: 254-258.
apy. Interestingly, beta-hCG NT did not have an 5) EMOTO M, SADAMORI R, HACHISUGA T, KAWARABAYASHI
T, MIYAMOTO S. Clinical usefulness of contrast-en-
association between the age of the initial beta- hanced color Doppler ultrasonography in invasive
hCG and pregnancy duration. The value of NT and noninvasive gestational trophoblastic dis-
and regression rates as a prognostic marker in eases: a preliminary study. J Reprod Med 2011;
GTD should be evaluated in future studies. 56: 224-234.
We grouped the cases according to patient age 6) KIZAKI S, MATSUI H, USUI H, SHOZU M, HANAWA S, YA-
(Table I) to further define our study population. MAMOTO E, KIKKAWA F. Normal human chorionic go-
nadotropin regression curves in uneventful post-
Initial beta-hCG, pregnancy duration, ultrasono- molar patients. J Reprod Med 2012; 57: 243-248.
graphic mean lesion size, and hCG NT was simi-
7) KOHORN EI. Imaging practices in the diagnosis and
lar in these three groups. Extremes of age (espe- management of gestational trophoblastic disease:
cially above 35 years) constitute an important an assessment. J Reprod Med 2012; 57: 207-210.
risk factor for GTD formation. Prognosis of 8) KOIRALA A, KHATIWADA P, GIRI A, KANDEL P, REGMI M,
CHM in these patients may also differ from the UPRETI D. The demographics of molar pregnancies
average aged group. In a study by Elias et al14, in BPKIHS. Kathmandu Univ Med J (KUMJ) 2011;
the authors stated that CHM had a distinct clini- 9: 298-300.
cal course in women aged 40 to 49 years. They 9) MOSHER R, GOLDSTEIN DP, BERKOWITZ R, BERNSTEIN M,
GENEST DR. Complete hydatidiform mole. Compari-
reported that aggressive primary therapy (hys- son of clinicopathologic features, current and
terectomy or prophylactic chemotherapy) was as- past. J Reprod Med 1998; 43: 21-27.
sociated with shorter hCG NTs, and decreased 10) NORMAN RJ, GREEN-THOMPSON RW, JIALAL I, SOUTTER
need of surgical or chemotherapeutic interven- WP, PILLAY NL, JOUBERT SM. Hyperthyroidism in ges-
tion. Although we could not detect a variable tational trophoblastic neoplasia. Clin Endocrinol
clinical course in women above 35 years, future (Oxf) 1981; 15: 395-401.
studies are required to establish management 11) OGE T, OZALP SS, GUNGOR T, YILDIRIM Y, SANCI M, DO-
GAN A, ERTAS IE, YETIMALAR H, DILEK S, CELIK C. Hyda-
guidelines in this special age group.
tidiform mole in Turkey: results from six centers. J
Considering the rare occurrence of this dis- Reprod Med 2012; 57: 259-261.
ease, we believe that accumulation of complete 12) S ALERNO A. The incidence of gestational tro-
moles by nation-wide investigation is required to phoblastic disease in Italy: a multicenter survey. J
improve the diagnosis and management of CHM. Reprod Med 2012; 57: 204-206.
13) S O T O -W R I G H T V, B E R N S T E I N M, G O L D S T E I N DP,
Conclusions BERKOWITZ RS. The changing clinical presentation
of complete molar pregnancy. Obstet Gynecol
1995; 86: 775-779.
Early detection and treatment of CHM is associ-
14) YUEN BH, CANNON W. Molar pregnancy in British
ated with a favorable clinical outcome. Initial beta- Columbia: estimated incidence and postevacua-
hCG levels and ultrasonographic mean lesion size tion regression patterns of the beta subunit of hu-
did not predict adjuvant chemotherapy requirement man chorionic gonadotropin. Am J Obstet Gy-
and did not change the type of treatment. necol 1981; 139: 316-319.

3384