Molar Research: Persistent trophoblast disease following pmp

Persistent trophoblast disease following partial molar pregnancy
Sabien WIESMA,1 Linda KERKMEIJER,1 Ruud BEKKERS,1 Jan PYMAN,2 Jeffrey TAN3 and
Michael QUINN4
1Department of Obstetrics and Gynaecology, Radboud University, Nijmigen, The Netherlands, 2Department of Pathology, 3Dysplasia Unit, and
4Oncology Unit and Melbourne University Department of Obstetrics and Gynaecology, Royal Women’s Hospital, Melbourne, Australia

Abstract
Objective: Human chorionic gonadotrophin (hCG) follow-up data were analysed retrospectively in all patients
registered in the Hydatidiform Mole Registry at the Royal Women’s Hospital, Melbourne from January 1992 to January 2001 to determine the risk of persistent trophoblast disease following partial molar pregnancy and to review the present follow-up protocol of patients suffering from partial hydatidiform molar pregnancy (PHM). Methods: Demographic factors were determined for all 344 cases with a review diagnosis of PHM, included age, history of previous hydatidiform mole, gestation length, hCG levels and compliance with follow-up. Findings: Six of the 344 patients diagnosed with PHM required treatment with single-agent methotrexate and folinic acid rescue. All six patients achieved and maintained a complete biochemical remission after chemotherapy.  hCG regression assays were analysed for 235 patients: 225 patients had at least one normal hCG measurement during follow-up, of whom 152 (64.7%) patients obtained normal values within 2 months after evacuation. All patients obtained normal levels within 32 weeks after evacuation of the partial hydatidiform mole. Only 63 (25.6%) patients completed the recommended follow-up program. No patient who achieved normal hCG levels required chemotherapy because of a recurrent gestational trophoblastic tumour.
Recommendations: This study indicates that 1.7% of all partial mole pregnancy patients needed treatment for
malignant sequelae. In contrast, no patient diagnosed with partial mole had a biochemical or clinical relapse after achieving normal levels of hCG, consistent with previous studies. Patients who have had a partial hydatidiform mole should be followed by hCG assays until normal levels are achieved and then follow-up can be safely discontinued.
Introduction
Gestational trophoblastic disease (GTD) defines a continuum of a neoplastic process that arises from the fetal chorion of the placenta. This continuum consists of diseases with a varying propensity for local invasion and metastasis. The WHO has classified GTD as two premalignant diseases consisting of complete or classical hydatidiform mole and partial hydatidiform mole (PHM), and as three malignant disorders consisting of invasive mole, choriocarcinoma and placentalsite trophoblastic tumour. In 2001 miscellaneous trophoblastic
tumour and unclassified trophoblastic lesions were added to this classification.  In an attempt to determine a high-risk group for malignant trophoblastic neoplasma, Vassilakos and Kajii2 distinguished two mole entities in 1976. They described the complete (classical) mole without a traceable embryo and the partial
(incomplete) mole with an ascertainable embryo/fetus.  The two entities can be distinguished on gross morphology, histopathology and karyotype, but many of their clinical features and their basic management are similar.3–6 Patients with PHM usually do not exhibit the dramatic clinical features of complete moles. They most often present in the late first trimester or early second trimester with missed or incomplete abortion. The volume of tissue is generally not excessive and clinical features like theca lutein cysts, hyperthyroidism or
respiratory insufficiency are seldom present. Preevacuation hCG levels are usually in the low to normal range and the diagnosis is often made after histological review of curettage specimens.6–8 Ultrasound can be of value in predicting a high likelihood of partial mole prior to curettage with the ratio of the transverse to the antero-posterior dimension of the gestational sac being greater than 1.5 and cystic changes, irregularity, or increased echogenicity in the decidual reaction/placenta or myometrium being found. The gross appearance of partial mole is variable and dependent on the gestational timing of evacuation. Hydropic change may not develop until the second trimester and is characteristically focal, but invariably vesicles are few or absent
macroscopically. Recognisable fetus or fetal parts including placental membranes and cord are frequently found. Placental tissue is not generally as bulky as that found with complete mole. Microscopically, excess trophoblast is usually only focal, rarely prominent and not generally circumferential. Characteristically, there are two distinct populations of villi identified, large oedematous villi with, when changes are well developed, central cisterns, and a second population of small villi that usually show some degree of stromal fibrosis. Villi
characteristically have a ‘scalloped’ outline resulting in rounded trophoblast inclusions. Often fetal development is seen in the form of nucleated red cells in the villi. This finding is helpful in the exclusion of a complete mole.3,5,10,11 Differentiating early partial mole from some non-molar hydropic miscarriages and distinguishing partial mole from early complete mole in early gestation may be difficult because of histological overlap.12 As most partial moles are found to be triploid, hydropic abortuses diploid and complete moles
diploid or occasionally tetraploid, ploidy studies may be useful to demonstrate triploidy.  Immunohistochemistry for p57kip2 has recently also been useful in distinguishing partial (p57 positive) and complete (p57 negative) moles. Although gestational trophoblastic tumour can follow any kind of gestation, 55% of trophoblastic malignancies followhydatidiform mole.14 Both complete and partial moles may develop gestational trophoblastic tumour with local uterine invasion and dissemination.15 All GTDs produce hCG, its
monitoring can be used to assure sustained remission and to prevent the development of undetected persistent GTD.  As suggested by Morrow et al.17 in 1977, normal hCG titre regression curves were drafted as an aid to early recognition of postmolar trophoblastic tumour.18–20 These studies suggest there are two groups of patients consisting of one group of patients with a ‘normal’ regression curve that declines spontaneously to normal levels over a short time, remaining well and free of disease and the other group consisting
of patients developing persistent GTD with plateauing or slowly declining hCG levels or relapse of GTD with
levels rising after a normal level has been achieved. Following the current FigO criteria, postmolar GTD is
diagnosed, if four or more hCG values plateau over at least 3 weeks, if there is a rise of hCG of 10% or greater for three values or longer if present over at least 2 weeks or if there is histopathological evidence of choriocarcinoma.The Hydatidiform Mole Registry of Victoria was established at the Royal Women’s Hospital, Melbourne in September 1973. All patients diagnosed with hydatidiform mole in Victoria are referred for registration and pathology review, and a free postal 24-h urinary chorionic gonadotropin
(UCG) excretion follow-up program is commenced. Because of the low incidence of continued trophoblast
disease after PHM, it was decided in July 1998 that, following confirmation of the diagnosis through the Registry, followup was not required and patients were advised that no special precautions were necessary. It was, however, considered appropriate to reevaluate these follow-up recommendations to ensure that no adverse outcomes nor unnecessary surgery resulted from this policy. As this study conformed to the standards established by the National Health and Medical Research Council for ethical quality review, ethics approval was not sought.
Results
Between 1 January 1992 and 1 January 2001, 843 patients with 846 pregnancies were registered. The histological diagnosis of complete mole was confirmed in 432 (51.1%) cases, 344 (40.7%) were diagnosed as partial, 11 (1.3%) were diagnosed as GTD without any further specification, 48 (5.7%) were misdiagnosed, in eight (0.9%) cases the initial diagnosis of choriocarcinoma was made, in two (0.2%) cases the diagnosis of placental site trophoblastic tumour was made and one (0.1%) patient was diagnosed with a twin pregnancy consisting of one partial mole and one complete mole. Four patients with a PHM had an obstetric history of more than one molar pregnancy, one with a previous partial mole, one with two previous unspecified moles and the other two with one previous unspecified mole. During the study period there were 568 279 live births recorded in Victoria, giving an incidence of 1.37 hydatidiform mole pregnancies per 1000 live births and an incidence of 0.61 PHM pregnancies per 1000 live births. The range of gestational age was 5–27 weeks with a mean of 13 weeks and the maternal age range between 14 and 44 years with a mean of 29.9 years. No significant differences between age and gestation of patients with uncomplicated partial molar pregnancy and complicated partial moles could be found (10.8 vs. 12.8 weeks, 29.8 vs. 29.5 years). Six (1.7%) of the 344 patients with partial molar pregnancy required chemotherapy with methotrexate and folinic acid. All patients were treated successfully. None have relapsed.  The time for hCG to reach normal was analysed in 235 uncomplicated cases. One hundred and fifty-two (64.7%) patients had normal hCG levels within 2 months, 61 (25.9%) patients normalised after 2 months and 11 (4.7%) patients had their first hCG value measured after 2 months and were found to be normal. Eleven (4.7%) patients failed to continue surveillance before normal levels were obtained. All patients with a spontaneous regression curve obtained
normal levels by 32 weeks. Only 63 (26%) of the 235 uncomplicated pregnancies were followed according to the protocol used in the Royal Women’s Hospital between 1992 and July 1998. Twenty-three (13.4%) patients among the 172 patients with uncompleted follow-up became pregnant before the 6 months they were advised not to become pregnant. No patient developed gestational trophoblastic tumour after obtaining normal hCG levels.
Discussion
It cannot be assessed whether all the patients with hydatidiform moles in Victoria were registered during the time of the study and it is possible that the incidence of hydatidiform mole could be even higher than 1.37 hydatidiform moles and 0.61 partial moles per 1000 live births. Neither can it be assessed whether all patients diagnosed with PHM were diagnosed correctly. Despite improved morphological criteria for complete and PHM, many lesions are difficult to classify because the features of partial mole overlap with complete moles and hydropic abortions. The incidence found in this study is higher than that reported previously.22
This study indicates that 1.7% of all PHM pregnancies needed treatment for malignant sequelae. In all cases, proper treatment was instituted and cure resulted. A varying number of treated patients following a partial molar pregnancy has been reported (Table 2). These differences are largely due to variations in the total number of patients with PHM included, and variations in the exact indications for the diagnosis of partial molar pregnancy and for the diagnosis of persistent disease. Although previous studies showed that gestational trophoblastic tumours could follow partial molar pregnancies, none of the patients with PHM registered in the Hydatidiform Mole Registry after July 1998 were followed by hCG assays due to the very low risk of malignant sequelae, the resource intensiveness of the follow-up program, and the economic, social and emotional burden for patients. The absence of a follow-up protocol for partial mole leads to a watchful waiting policy. Four of the six patients were admitted for treatment after a second ultrasound was performed
because of heavy bleeding per vaginum and showed placental tissue, whereas the other two patients had rising
and plateaued levels of hCG, respectively. It may be argued that a further curettage and a continued expectant management policy may have been more appropriate for these patients. Furthermore, because of the fact that the diagnosis of gestational trophoblastic tumour is made and treatment is given without histological confirmation of tumour development, it cannot be assessed whether the two remaining patients were overtreated, whether they would have presented with clinical features shortly after the plateauing or rising levels of hCG, or whether their hCG levels may have dropped spontaneously. In 2003, Feltmate et al.33 found that none of the 320 patients with complete hydatidiform moles who achieved at least one undetectable hCG level had been diagnosed with gestational trophoblastic tumour relapse. Likewise, Wolfberg et al.35 analysed 1029 patients to determine the risk of relapse after achieving undetectable hCG levels in women with complete molar pregnancy. They found that the risk of recurrent neoplasm approached zero. Batorfi et al.36 analysed 150 randomly selected patients with molar pregnancy to determine if continuing follow-up of uncomplicated molar cases beyond attaining undetectable hCG levels was necessary for detecting relapse of GTD, and recently, a report of Lavie et al.32 found that none of the women who achieved hCG normalisation following evacuation of a PHM subsequently required chemotherapy. Both studies concluded that after
hCG values fall spontaneously within the normal range, patients can safely be discharged from follow-up.
In contrast, Bagshawe et al.28 analysed hCG values in 8095 patients to determine the risk of malignant sequelae after obtaining normal hCG levels. They found that none of thepatients with hCG levels falling to the normal range within 2 months after evacuation developed sequelae requiring treatment. When hCG levels fell to normal after 2 months after evacuation, the risk of recurrence was one in 96. When hCG values had stayed normal for 6 months, the risk of rising hCG values indicating recurrence was one in 286. None of their patients with a specified diagnosis of PHM developed malignant sequelae. Despite strong efforts, many patients with molar pregnancy do not complete the recommended hCG follow-up program. In this study, it was found that only 25.6% of all uncomplicated patients monitored by hCG measurements completed the follow-up program.
A recommendation that patients are followed until normal hCG levels are achieved would have resulted in 65% of patients being discharged from follow-up within 2 months and all patients would have been discharged from follow-up within 32 weeks. This approach seems a safe and reasonable compromise, which we recommend.
Source: Australian and New Zealand Journal of Obstetrics and Gynaecology 2006; 46: 119–123

To have this information sent to you in pdf form, please email me at isis715@yahoo.com