Gestational trophoblastic disease: is intensive follow up essential in all women?

Authors


Professor B. Hancock, YCR Department of Clinical Oncology, Weston Park Hospital, Sheffield, S10 2SJ UK.

Abstract

Objective  To determine the timescale of the registration process for gestational trophoblastic disease and its impact on hCG level at registration and subsequent need for chemotherapy.

Design  A prospective observational study using a standardised protocol for registration, assessment and treatment for molar pregnancy.

Setting  A supra-regional tertiary referral centre for gestational trophoblastic disease.

Participants  A total of 2046 consecutive women registered between January 1994 and December 1998 with a diagnosis of molar pregnancy.

Methods  Data at and after registration, collected prospectively on a computerised database, were statistically analysed (by multiple logistic regression and ANOVA).

Main outcome measures  Relationship between length of time to and hCG value at registration; also the subsequent need for chemotherapy.

Results  A total of 2046 women with a diagnosis of molar pregnancy were registered in the study period. The mean time interval between first evacuation and registration at the referral centre was 47 days (median 37, range 0–594). One hundred and five out of 2046 (5.1%) women needed chemotherapy. Sixty-three precent of the women (1296 out of 2046) had a normal level of urinary hCG (less than 40 IU/24 hours) at the time of registration and only one (0.08%) needed chemotherapy. Binary logistic regression analysis showed a statistically significant relationship between time to registration, hCG value, histology, pretreatment risk score and decision to administer chemotherapy.

Conclusion  Women with gestational trophoblastic disease who were registered late were significantly more likely to have normal levels of hCG and were less likely to need chemotherapy. A less intensive follow up may be justified in women with gestational trophoblastic disease who are registered with a normal hCG level.

INTRODUCTION

The incidence of gestational trophoblastic disease in the UK is around 1.5 per 1000 live births.1,2 A large proportion of molar pregnancies undergo spontaneous resolution but 7–8% of women have persistent disease and need chemotherapy.3 Hence, on establishing a diagnosis of a molar pregnancy, women are registered at a regional centre to facilitate follow up in order to identify cases with persistent disease. There are three centres in the UK for registration, follow up and treatment, Charing Cross Hospital, London; Ninewells Hospital, Dundee; and Weston Park Hospital, Sheffield. This programme of registration has achieved impressive results, with high cure and low chemotherapy rates.4 Currently, hCG level after one or two evacuations is an important determinant of treatment,5 which in itself is a function of time. It is possible that with time, the proportion of women needing chemotherapy may decrease. We therefore hypothesised that in women who registered late, there is more time to allow spontaneous resolution of molar tissue and therefore a reduced need for chemotherapy.

This study was carried out to determine the timescale of the registration process for gestational trophoblastic disease and its effect on hCG level at registration and subsequent need for chemotherapy.

METHODS

Women referred to Weston Park Hospital with a diagnosis of molar pregnancy were registered as per clinical guidelines by the Royal College of Obstetrics and Gynaecologists.5 A computerised database was used to record information including histological diagnosis, date of first evacuation, date of registration, date of initial urinary hCG estimation, initial hCG level, date of treatment, pretreatment risk score and chemotherapy regimen. Data were collected prospectively over five years between January 1994 and December 1998. MS Access 2000, Excel 2000 and SPSS 7.5 were used for analysing the data. Multiple logistic regression analysis and ANOVA were used for statistical analysis.

The participants were new cases of gestational trophoblastic disease registered at the Weston Park Hospital, Sheffield between January 1994 and December 1998.

RESULTS

A total of 2046 women were registered in the study period. The age distribution is shown in Table 1. Twenty-three women (1.1%) did not complete the prescribed follow up. Of the 2046 cases, 105 (5.1%) needed chemotherapy. Eighty-seven (83%) had a risk score of 7 or less and 18 women (17%) had a risk score of 8 or more according to the Charing Cross Hospital scoring system.6 Twenty-six women (24%) needed second line chemotherapy (20 had resistant disease, 6 developed side effects to the first line treatment).

Table 1.  Age distribution of women with gestational trophoblastic disease at registration.
Age (years)No. of womenWomen requiring chemotherapy
<2029514 (4.8%)
20–39164582 (5.0%)
>391059 (8.8%)

Initial urinary hCG estimation at registration was normal (less than 40 IU/24 hours) in 1296 out of 2046 women (63%) and only one out of these 1296 cases (0.08%) subsequently needed chemotherapy. The initial urinary hCG level was elevated in 630 out of 2046 women (31%), and 55 of these needed chemotherapy (8.7%).

The mean interval between first evacuation of the molar pregnancy and registration was 47 days (range 0–594, median 37). The range of intervals is shown in Table 2. We noted that the need for chemotherapy decreased as the time interval from evacuation of the molar pregnancy to registration increased, except in women who were registered later than 80 days after the first evacuation (Table 2).

Table 2.  Interval between first evacuation of the molar pregnancy and registration, and need for chemotherapy.
Days from evacuationNo. of casesCases requiring chemotherapy
0–2031431 (9.9%)
20–4078239 (5.0%)
40–6046919 (4.1%)
60–802003 (1.5%)
80 or more24811 (4.4%)

As the time to registration increased, the urinary hCG level was significantly more likely to be normal (ANOVA; P < 0.0001).

Multiple logistic regression analysis was carried out to determine the factors that influence the decision to administer chemotherapy. Age, histological diagnosis, hCG value, pretreatment risk score and number of days from diagnosis to registration were entered into the regression equation as explanatory variables. The dependent variable was decision to prescribe chemotherapy. The Wald Statistic and P value are summarised in Table 3. This regression equation accounted for 63% of the variance in the dependent variable (R2= 0.633). This analysis showed that there was a statistically significant negative relationship between time to registration and need for chemotherapy. Histology was a predictor of need for chemotherapy; complete moles and choriocarcinoma (P= 0.043, P= 0.035, respectively) were significantly associated, whereas partial moles and hydropic pregnancies were not (P= 0.270, P= 0.203).

Table 3.  Logistic regression analysis of factors influencing the decision to administer chemotherapy.
VariableWald statisticP
  1. Logistic regression analysis (SPSS).

Age0.170.89
Time to registration3.810.0449
hCG level19.300.00001
Histology16.740.0022
Risk score152.560.00001

DISCUSSION

Gestational trophoblastic disease includes a spectrum of conditions ranging from benign disease (i.e. partial or complete hydatidiform mole) to choriocarcinoma. Benign molar tissue resolves spontaneously in majority of cases. However, in 5% to 20% of cases, it may persist for up to several months and may need chemotherapy.7–9 Hence, most cases are managed conservatively, after initial evacuation, with careful observation and monitoring of hCG levels.

Human chorionic gonadotrophin level is a reliable estimate of molar disease activity and a valuable tumour marker. With spontaneous resolution of molar tissue, hCG levels decline, whereas elevated levels reflect persistent trophoblastic tissue. Currently, hCG level after one or two evacuations is an important determinant of treatment9 which in itself is a function of time. It is logical that with time the proportion of women needing chemotherapy may decrease. Careful clinical supervision is however mandatory.

In the UK, there is a national service for registration of women with molar pregnancy. This registration process can take a variable length of time (0 to 594 days from first evacuation of the molar pregnancy in this study).

The mean interval between first evacuation of the molar pregnancy and registration was 47 days. This interval may be influenced by several factors: delay in initial diagnosis, delay in obtaining histopathological assessment, delay or failure in communication, non-compliance and lack of awareness about the registration process.

Sixty-three percent of the women (1296) had a normal hCG level at the time of registration. Of these, only one subsequently required chemotherapy. A less intensive follow up programme may be suitable for women who have a normal level of hCG at the time of registration. However, long term follow up should not be abandoned because of the risk of reactivation of gestational trophoblastic disease or the transformation of molar disease to choriocarcinoma.10–12

Women who were registered later were more likely to have a normal initial level of hCG. This is probably due to the longer time interval allowing spontaneous resolution. Multiple regression analysis showed a statistically significant relationship between number of days to registration and decision to administer chemotherapy. Although it is tempting to suggest that the delay in registration was responsible for this, clinical bias in the referral pattern is also an important factor. Women with problematic disease (high initial hCG, presence of metastasis, persistent symptoms, choriocarcinoma, disease even after two evacuations) are more likely to be referred earlier. This may explain the higher incidence of persistent disease in these women. Indeed, hCG value, histological diagnosis and risk score were also shown to be statistically significant independent predictors of the decision to administer chemotherapy.

Women older than 39 years were also more likely to need chemotherapy; although this difference was not statistically significant. It is generally acknowledged that age over 39 years is a predisposing factor for persistent disease9 and age is still included in the new FIGO 2000 staging/scoring system.13

Interestingly, women who registered later than 80 days were also more likely to need chemotherapy. This group was not significantly different from other women in mean age, mean risk scores and histology. It is possible that in some of these women, the diagnosis was initially missed and only came to attention due to persistence of disease causing symptoms. However, all women in this group who required chemotherapy had elevated levels of hCG at registration.

In the UK, with close surveillance, we have been prepared to wait for up to six months14 for hCG levels to return to normal before initiating chemotherapy. The present study adds support to this conservative approach and suggests also that a less intensive follow up may be justified in women registered late with a normal hCG level.

Accepted 3 March 2004

Ancillary