Article Text

Download PDFPDF

Gestational trophoblastic neoplasia: a meta-analysis evaluating reproductive and obstetrical outcomes after administration of chemotherapy
  1. Anastasios Tranoulis1,
  2. Dimitra Georgiou2,
  3. Ahmad Sayasneh1,3 and
  4. John Tidy4
  1. 1 Department of Gynaecological Oncology, Guy's and St Thomas' NHS Foundation Trust, King's College, London, UK
  2. 2 Department of Obstetrics and Gynaecology, Chelsea and Westminster NHS Foundation Trust, Imperial College, London, UK
  3. 3 School of Life Course Sciences, Faculty of Life Sciences and Medicine, King's College, London, UK
  4. 4 Sheffield Trophoblastic Centre, Weston Park Hospital, Sheffield, UK
  1. Correspondence to Dr Anastasios Tranoulis, Department of Gynaecological Oncology, Guy's and St Thomas' NHS Foundation Trust, King's College, London SE1 7EH, UK; tasostranoulis{at}yahoo.com

Abstract

Introduction Gestational trophoblastic neoplasia represents a rare placental malignancy spectrum that is treated with single- or multi-agent chemotherapy. This disease often impacts women of childbearing age, making post-chemotherapy fertility and obstetrical outcomes an important consideration. We aimed to ascertain the pregnancy rates and obstetric outcomes in women with gestational trophoblastic neoplasia after undergoing treatment with chemotherapy.

Methods A systematic literature review was conducted to identify studies that reported post-chemotherapy fertility and obstetric outcomes among women with gestational trophoblastic neoplasia. We performed a single-proportion meta-analysis for the outcomes of conception/pregnancy rate, term live birth rate, first and second trimester spontaneous abortions rate, stillbirth rate, premature delivery rate, and fetal/neonatal malformation rate.

Results A total of 27 studies were included in the analysis. The median age ranged between 25.5 and 33.1 years. The pregnancy rate among women with a desire to conceive, comprising a total of 1329 women and 1192 pregnancies, was 86.7% (95% CI 80.8% to 91.6%). The term live birth rate in 6752 pregnancies was 75.84% (95% CI 73.4% to 78.2%). The adverse pregnancy outcomes were seemingly comparable to those of the general population apart from a minor increase in the stillbirth rate. The pooled proportion for the outcome of malformation rate was 1.76% (95% CI 1.3% to 2.2%). The repeat mole rate in 6384 pregnancies was 1.28% (95% CI 0.95% to 1.66%). Subsequent sub-group analysis indicated that neither multi-agent chemotherapy nor conception within 12 months post-chemotherapy increased the adverse obstetric events risk or fetal malformations.

Conclusions Nearly 90% of patients desiring future fertility after chemotherapy for gestational trophoblastic disease were able to conceive. In addition, adverse pregnancy outcomes were similar to that in the general population. Multi-agent chemotherapy does not seemingly increase the malformation rate.

  • gestational trophoblastic neoplasia
  • fertility
  • obstetric outcomes

Statistics from Altmetric.com

Highlights

  • The pregnancy rate for patients with gestational trophoblastic neoplasia after undergoing treatment with chemotherapy was 87%.

  • The term live birth rate after chemotherapy treatment was 75.8% in patients with gestational trophoblastic neoplasia.

  • Conception within 12 months post-chemotherapy does not seemingly increase the risk of adverse sequelae.

Introduction

Gestational trophoblastic neoplasia represents a spectrum of rare placental malignancies comprising persistent gestational trophoblastic disease, invasive mole, choriocarcinoma, placental site trophoblastic tumor, and epithelioid trophoblastic tumor.1 Given that many women who develop gestational trophoblastic neoplasia are of childbearing age, preservation of fertility becomes an important consideration.1 Chemotherapy is the treatment of choice for gestational trophoblastic neoplasia and the option between single- and multiple-agent chemotherapy is based on the International Federation of Gynecology and Obstetrics (FIGO) prognostic score.2 Women scoring <7 are classified as low-risk and are generally treated with single-agent chemotherapy, usually methotrexate or actinomycin-D, with an overall cure rate of nearly 100%.2 In cases of persistent disease, treatment may be switched to actinomycin-D or multi-agent chemotherapy.2 3 High-risk disease is usually treated with a chemotherapy regimen comprising etoposide, methotrexate, actinomycin-D, cyclophosphamide, and vincristine. For resistant or relapsing disease, second-line chemotherapy consisting of paclitaxel/etoposide alternating with paclitaxel/cisplatin or bleomycin has been used.4 The overall cure rates for high-risk disease have been reported up to 94%.5

Efficacious chemotherapy regimens have substantially improved the prognosis for gestational trophoblastic neoplasia over the last decades. Women treated with chemotherapy, in the vast majority of cases, are expected to be long-term survivors. In light of their age and ability to preserve fertility, a high proportion of women express the desire to conceive post-chemotherapy. However, there remains concern about the reproductive success rate, in addition to the potential obstetric complications and fetal anomalies in future pregnancies. Despite a growing body of reassuring data,6–32 there is a paucity of standardized evidence for clinicians seeking guidance to inform this subgroup of women concerning their fertility rates and obstetric sequelae from chemotherapy treatment. To address this, we performed a meta-analysis of the pregnancy rate and obstetrical outcomes in women with gestational trophoblastic neoplasia treated with chemotherapy.

Methods

We carried out a systematic review and meta-analysis of the literature complying with PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines to identify English language studies. We searched MEDLINE, Embase, Web of Science, and Cochrane Database from inception until October 2018. Eligible studies were retrieved based on the Boolean search strategy to describe patient population (P: women with gestational trophoblastic neoplasia), the intervention (I: single- or multi-agent chemotherapy), the outcome (O: pregnancy rate OR miscarriage OR ectopic pregnancy OR stillbirth OR repeat mole OR premature deliveries OR live births OR fetal anomalies) and comparison (C: single-agent vs multi-agent chemotherapy) without defining any study design (S:/). In searching PubMed the following MeSH terms and keywords were used: ‘gestational trophoblastic disease’; ‘gestational trophoblastic neoplasia’; ’pregnancies’; ’fertility’; ’obstetric outcomes’; ’fetal anomalies’. References of selected studies were searched to retrieve articles not identified by the electronic search.

The study selection was processed in two stages. Full titles and abstracts were examined, and relevant articles were obtained. All the articles that met the inclusion criteria were retrieved as full texts. The methodological index for non-randomized studies (0–16 scale) was used to evaluate the quality of the included studies. Study selection, data extraction and quality assessment were carried out in parallel by two co-authors (AT, DG). Inconsistencies were discussed and resolved by consensus by four authors (AT, DG, AS, JT). We extracted data on the following: name of authors, year of publication, study design, study population, age, histology, fertility rate, and obstetric outcomes.

The primary endpoints consisted of conception/pregnancy rate and term live birth rate. Secondary outcomes included rates for each of the following: first and second trimester spontaneous abortions, ectopic pregnancy, stillbirth, repeat mole, premature delivery, and fetal/neonatal malformation. The pregnancy rate was defined as the ratio of patients who had at least one pregnancy to the total number of women. The term live birth was defined as the ratio of live birth deliveries at term to the total number of pregnancies. A premature delivery was defined as a delivery at >24 weeks and <37 weeks of gestation. Miscarriage was defined as the loss of pregnancy during the first 24 weeks of gestation. Stillbirth was defined as fetal death after 24 weeks of gestation. Congenital anomalies were defined as structural or functional anomalies that occurred during pregnancy or were identified prenatally, at birth, or later in infancy.

The inclusion criteria consisted of the following: confirmed diagnosis of gestational trophoblastic neoplasia; studies on the treatment of gestational trophoblastic neoplasia with chemotherapy; randomized controlled trials and observational studies; reported data concerning pregnancy rate and obstetric outcomes; sufficient published data for estimating the pooled proportions and 95% confidence intervals (95% CI); English language. Exclusion criteria consisted of the following: studies irrelevant to treatment of gestational trophoblastic neoplasia with chemotherapy; patients treated with hysterectomy; insufficient data for pregnancy rate and obstetric outcomes; case reports, reviews, and small case series of five or fewer patients; repeated or overlapped data.

We carried out a single proportion meta-analysis with inverse-variance weighting using a random-effects model owing to the significant heterogeneity in the methodological characteristics of included studies. The pooled proportions were presented with exact binomial 95% CI. The variance between the studies was established with the DerSimonian-Laird estimator, and we applied the arcsine transformation to raw proportions.33 We assessed the percentage of total variation across studies due to heterogeneity by the I2 index, and the Cochran’s Q test was used to establish the significance of heterogeneity. We performed a single proportion stratified analysis for pregnancies achieved within 12 months of chemotherapy completion. We also performed a single proportion sub-group analysis to explore the variation of effect of single- and multi-agent chemotherapy on the aforementioned outcomes. Direct comparison meta-analysis was performed between single- and multi-agent chemotherapy by calculating the pooled odd ratios (OR) and 95% CI. The p value was set at 0.05. Statistical meta-analysis was performed using MedCalc (MedCalc software, Ostend, Belgium).

Results

The search strategy yielded 2567 citations, of which 2468 were excluded. The remaining 99 full articles were reviewed and a further 52 citations were excluded. A total of 27 studies encompassing 6752 pregnancies were included in the review6–32 (Figure 1). Studies were published between 1976 and 2018. Most of them originated from the UK (n=9),9 13 16 17 21 22 26 29 31 USA (n=7),7 8 11 20 28 30 32 and Asia (n=7).12 14 18 19 24 25 27 All the studies were retrospective. The methodological index for non-randomized studies ranged between 7 and 12 (online supplementary Table 1). The data from these studies6–32 were tabulated in the structured Table 1. The median age ranged between 25.5 and 33.1 years. The sample size per study varied across the reports and ranged from 22 to 1460 women and from 21 to 1313 pregnancies. One study assessed the effect of single-agent chemotherapy,13 three of multi-agent chemotherapy,15 21 29 and 18 studies of both single- and multi-agent chemotherapy, respectively.6 7 9 10 12 14 17–20 22 23 25–27 30–32 Five studies did not provide data for the type of chemotherapy.8 11 16 24 28 We identified 14 studies examining pregnancy rates, which included a total of 3774 women and 3846 pregnancies.6 7 9 12–15 19 21 22 25 26 29 30 Six studies reported on the pregnancy rate among women with a desire to conceive.15 19 22 25 28 31 All studies evaluated the obstetric outcomes of subsequent pregnancies, while 24 of these provided data for the malformation rate.6–8 12–32

Supplemental material

Figure 1

Flowchart of the literature selection process. Full titles and abstracts were examined, and relevant articles were obtained. All articles that met the inclusion criteria were retrieved as full texts. Twenty-seven studies were included in the meta-analysis for data synthesis.

Table 1

Studies selected for inclusion into the meta-analysis of the chemotherapy impact on fertility and obstetric outcomes of women with gestational trophoblastic neoplasia

Pregnancy Rate

In total, 14 studies6 7 9 12–15 19 21 22 25 26 29 30 assessed the pregnancy rate after chemotherapy, accounting for 3764 women and 3846 pregnancies. The pooled proportion was 67.42% (95% CI 48.2% to 83.9%; Figure 2A) with substantial inter-study heterogeneity (I2=99.25%, p<0.0001; online supplementary Figure 1A). The pooled proportion from six studies15 19 22 25 28 31 assessing the pregnancy rate among women with a desire to conceive, comprising a total of 1329 women and 1192 pregnancies, was 86.7% (95% CI 80.8% to 91.6%; Figure 2B) with high inter-study heterogeneity (I2=82.62%, p<0.0001; online supplementary Figure 1B).

Supplemental material

Figure 2

(A) Pregnancy rate following chemotherapy treatment for gestational trophoblastic neoplasia. Forest plots demonstrate the proportions of women (with confidence intervals (CIs)) with subsequent pregnancy. (B) Pregnancy rate among women with strong desire to conceive following chemotherapy treatment for gestational trophoblastic neoplasia. Forest plots demonstrate the proportions of women (with CIs) with subsequent pregnancy. High heterogeneity was observed for both outcomes.

Term Live Birth Rate

The pooled proportion from 26 studies6–20 22–32 assessing the term live birth rate in 6752 pregnancies was 75.84% (95% CI 73.4% to 78.2%; Figure 3A) with high inter-study heterogeneity (I2=76.62%, p<0.0001).

Figure 3

(A) Live term birth rate following chemotherapy treatment for gestational trophoblastic neoplasia. Forest plots demonstrate the proportions of women (with confidence intervals (CIs)) with term live birth. (B) Premature delivery rate following chemotherapy treatment for gestational trophoblastic neoplasia. Forest plots demonstrate the proportions of women (with CIs) with premature delivery. (C) Stillbirth rate following chemotherapy treatment for gestational trophoblastic neoplasia. Forest plots demonstrate the proportions of women (with CIs) with stillbirth. (D) First and second trimester spontaneous abortion rate following chemotherapy treatment for gestational trophoblastic neoplasia. Forest plots demonstrate the proportions of women (with CIs) with first and second trimester spontaneous abortion. (E) Ectopic pregnancy rate following chemotherapy treatment for gestational trophoblastic neoplasia. Forest plots demonstrate the proportions of women (with CIs) with ectopic pregnancy. (F) Repeat mole rate following chemotherapy treatment for gestational trophoblastic neoplasia. Forest plots demonstrate the proportions of women (with CIs) with repeat mole. (G) Malformation rate following chemotherapy treatment for gestational trophoblastic neoplasia. Forest plots demonstrate the proportions of structural or functional anomalies (with CIs) that occurred during intra-uterine life or were identified prenatally, at birth, or later in infancy. Moderate to high heterogeneity was observed for all outcomes.

Pre-term Birth Rate

The pooled proportion from 21 studies6–12 14–16 18–25 27 28 31 34 assessing the pre-term birth rate in 5781 pregnancies was 5.06% (95% CI 4.0% to 6.2%; Figure 3B) with high inter-study heterogeneity (I2=64.77%, p<0.0001).

Stillbirth Rate

The pooled proportion from 26 studies6–20 22–32 assessing the stillbirth rate in 6752 pregnancies was 1.32% (95% CI 0.9% to 1.7%; Figure 3C) with high inter-study heterogeneity (I2=40.83%, p=0.016).

First/Second Trimester Spontaneous Abortion Rate

The pooled proportion from 25 studies6–32 assessing the first/second trimester spontaneous abortion rate in 5439 pregnancies was 14.62% (95% CI 12.7% to 16.6%; Figure 3D) with high inter-study heterogeneity (I2=72.46%, p<0.0001).

Ectopic Pregnancy Rate

The pooled proportion from 24 studies6–10 12–20 23–32 assessing the ectopic pregnancy rate in 4808 was 0.61% (95% CI 0.38% to 0.88%; Figure 3E) with moderate inter-study heterogeneity (I2=15.36%, p=0.24).

Repeat Mole Rate

The pooled proportion from 25 studies6–20 22–25 27–32 assessing the repeat mole rate in 6384 pregnancies was 1.28% (95% CI 0.95% to 1.66%; Figure 3F) with moderate inter-study heterogeneity (I2=24.57%, p=0.13).

Malformation Rate

The pooled proportion from 24 studies6–8 12–32 assessing the malformation rate in 4685 pregnancies was 1.76% (95% CI 1.3% to 2.2%; Figure 3G) with moderate inter-study heterogeneity (I2=26.11%, p=0.12).

Sub-Group Analyses

The results of the sub-group analyses, including conception, after single- or multi-agent chemotherapy and conception within 12 months post-chemotherapy are shown in Table 2. Direct comparison meta-analysis between single- and multi-agent chemotherapy for the outcome of pregnancy rate was statistically significant (OR 0.54, 95% CI 0.38 to 0.77, p=0.001; I2=22.41%, p=0.27) (online supplementary Figure 2A). Direct comparison meta-analysis between single- and multi-agent chemotherapy for the outcome of premature delivery rate was not statistically significant (OR 1.12, 95% CI 0.54 to 2.35, p=0.75; I2=0%, p=0.96) (online supplementary Figure 2B). Direct comparison meta-analysis between single- and multi-agent chemotherapy for the outcome of malformation rate was not statistically significant (OR 1.81, 95% CI 0.47 to 7.01, p=0.38; I2=15.49%, p=0.3) (online supplementary Figure 2C).

Supplemental material

Table 2

Pooled proportions of the obstetric outcomes and malformation rate in women who conceived following chemotherapy treatment for gestational trophoblastic neoplasia by chemotherapy type (single- or multi-agent chemotherapy), and for conception within 12 months post-chemotherapy completion. The pooled proportions alongside the 95% CIs and heterogeneity are illustrated.

Discussion

Our meta-analysis outlines that women with gestational trophoblastic neoplasia may expect high post-chemotherapy pregnancy rates and reassuring obstetric outcomes. Early pregnancy within 12 months post-chemotherapy does not seemingly favor adverse obstetric outcomes, while the neonatal malformation rate appears to be low and comparable to that of the general population.

We demonstrated that nearly seven out of 10 women will conceive following chemotherapy treatment for gestational trophoblastic neoplasia. Notably, nearly nine out of 10 women with a desire for childbearing will get pregnant. The loss of primordial follicles caused by chemotherapy may theoretically result in decreased ovarian reserve. A proportion of women may thus experience a period of anovulation. Nonetheless, the vast majority of patients maintain a normal menstrual cycle through treatment or recover normal menstrual function without fertility compromise.6 17 We demonstrated a significantly lower pregnancy rate after multi-agent chemotherapy. However, we were unable to perform a sub-group analysis among women with a desire for future pregnancy. Therefore, this evidence should be interpreted with caution given that the pregnancy desire has a significant impact on the probability to attain future pregnancy.6 Of note, the recent MITO-9 study demonstrated a non-significant difference in pregnancy rate between the single- and multi-agent chemotherapy groups in women with a desire for pregnancy.6

Our meta-analysis showed that in three out of four women, pregnancies will result in live term birth. Furthermore, the adverse pregnancy outcomes appear to be comparable to those of the general population apart from a slightly increased stillbirth rate (1.32% vs 0.8%).8 22 The incidence of repeat mole pregnancy after a complete or partial mole pregnancy was <2% in most series.1 2 34 The repeat mole rate in the current review was 1.2%, postulating that chemotherapy does not increase the repeat mole rate. Interestingly, the malformation rate appears to be relatively low and comparable to that of the general population.35

Some early studies suggested that a greater exposure to multiple chemotherapy agents may lead to higher rates of adverse obstetric outcomes.24 26 Interestingly, our study suggests that multi-agent chemotherapy does not increase adverse obstetric outcome risk compared with single-agent. A slightly higher malformation rate was observed in the multi-agent group. Nonetheless, direct comparison meta-analysis highlighted a non-significant difference between the two groups. This comparison contributes meaningful information concerning the safety of multi-agent chemotherapy for the aforementioned outcomes.

The post-chemotherapy relapse risk is approximately 3%, and most relapses occur in the first 12 months of follow-up.1 3 Early pregnancy could compromise post-chemotherapy surveillance, posing challenges in trying to distinguish between a new pregnancy event and disease relapse. The cytotoxic agents have been associated with mutagenic and teratogenic effects.3 Chemotherapy can induce chromosomal aberrations during the pre-ovulatory oogenesis phase II oocytes.3 The maturation of the recruited oocytes may last >6 months.3 There should be a relatively long interval between chemotherapy completion and the first subsequent pregnancy to allow ovaries to repair or undergo regeneration, reducing thus the first/second trimester spontaneous abortion or malformation rate. In line with previously published data,6–8 12 13 15–32 the sub-group analysis, including solely studies reporting on obstetric outcomes of pregnancies within 12 months post-chemotherapy, demonstrated that early pregnancy does not increase the unfavorable obstetric outcome risk. The malformation rate was comparable to that of the general population.35 The EUROCAT organization recorded a total prevalence of major congenital anomalies of 2.05% for the period 1996 to 2009.36 Careful human chorionic gonadotropin monitoring is essential, and pregnancy should ideally be delayed until beyond this period using effective contraception.1–3 Since early pregnancies usually result in a favorable outcome, patients may be allowed to continue with their pregnancy under close surveillance. Ultrasonographic monitoring during the first trimester should be performed to confirm normal gestational development. In case of vaginal bleeding or any systematic symptoms, a thorough evaluation is essential.1 2

Our search strategy ensured a systematic literature review to minimize potential losses of relevant published articles. No date restriction was applied. We included a large number of studies and hence a large number of cases, resulting in more robust estimates. Sub-group analysis was performed to address heterogeneity. The pooled obstetric outcomes for subsequent pregnancies achieved within 12 months post-chemotherapy were also assessed.

Nonetheless, some limitations should be acknowledged. The studies included were of moderate quality. Moderate to high inter-study heterogeneity was detected for a few outcomes. Few studies reported on the pregnancy rate among women who strongly desired pregnancy, and fewer if the pregnancy was a result of an assisted reproductive technique or intercourse. Another limitation inherent to meta-analyses of aggregated data extracted from published data is the limited number of potentially influential covariates that were analyzed. We were unable to perform sub-group analysis or meta-regression that incorporated influential factors such as age, histology, chemotherapy agents, and population or hospital based studies to further address heterogeneity. Furthermore, we cannot exclude the fact that the obstetric outcomes are distributed differently between different geographical regions. Owing to the aforementioned limitations the evidence of our meta-analysis should be interpreted with caution.

The current meta-analysis demonstrated that nearly nine out of 10 women with a desire for childbearing will achieve future pregnancy. Favorable obstetric outcomes should be anticipated in subsequent pregnancies, while conception within 12 months post-chemotherapy or multi-agent chemotherapy does not seemingly increase the risk of adverse sequelae.

Acknowledgments

AT and DG would like to thank JT for his invaluable support and professional expertise.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.

Footnotes

  • Contributors Study conception and design belongs to AT. AT, DG carried out acquisition, analysis and interpretation of data as well as manuscript editing. AS carried out manuscript editing. JT critically revised and appraised the manuscript. The authors certify that there is no actual or potential conflict of interest in relation to this article. We also certify that no party has a direct interest in the results of the research and that no benefit will be conferred on us or any organization with which we are associated. This manuscript is not currently under consideration for publication in another journal.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.