Chemotherapy for malignant germ cell ovarian cancer in adult patients with early stage, advanced and recurrent disease

Abstract

Background

Malignant germ cell tumour of the ovary occurs in up to 0.07% of woman globally. Due to its rarity, evidence for treatment is lacking and often extrapolates clinical trial results of testicular germ cell cancers. The investigation on this rare tumour is further compounded by the fact that its occurrence in the adult population is even less compared to their paediatric counterpart. At present, the effectiveness of chemotherapy, regardless of stage in malignant germ cell tumour of the ovary is not entirely clear.

Objectives

To evaluate the effectiveness and safety of chemotherapy in adult women with early stage, advanced and recurrent malignant germ cell ovarian cancers.

Search methods

We searched the Cochrane Central Register of Controlled Trials (CENTRAL), Issue 1, 2010, Cochrane Gynaecological Cancer Group Trials Register, MEDLINE and EMBASE up to April 2010. We also searched registers of clinical trials, abstracts of scientific meetings and reference lists of included studies.

Selection criteria

We searched for randomised controlled trials (RCTs), quasi‐RCTs and non‐randomised studies that compared systemic therapy in adult women diagnosed with germ cell ovarian cancer who have confirmed pathological diagnoses.

Data collection and analysis

Two review authors independently assessed whether potentially relevant studies met the inclusion criteria, abstracted data and assessed risk of bias.

Main results

We found one RCT and one retrospective study that met our inclusion criteria. The data from these studies were too sparse to adequately assess the effectiveness and safety of adjuvant chemotherapy in the treatment of malignant germ cell ovarian cancer. All comparisons were restricted to single study analyses and this review was only based on 32 women, so it was not adequately powered to detect differences in survival. Adverse effects of treatment and recurrence‐free survival were incompletely documented and QoL was not reported in any of the studies.We did not find any studies that reported specifically on adults so there were problems in separating data on adults and children in many of the potentially relevant studies.

Authors' conclusions

We found only low quality evidence on the use of chemotherapy in malignant germ cell tumours of the ovaries. Therefore we are unable to reach definite conclusions about the relative benefits and harms of chemotherapy use in this disease regardless of disease stage. Due to the benefit of chemotherapy in germ cell cancer of the testis, a trial of chemotherapy versus best supportive care is unlikely to be feasible. Despite this, good quality randomised studies are warranted in this disease to define the role of chemotherapy (type of chemotherapy, duration of treatment, benefit, short and long term toxicities). Given the rarity of this disease, we feel a trans‐global approach would be essential in order to perform such trials.

Plain language summary

Chemotherapy for adult women diagnosed with a rare type of ovarian cancer of all stages (malignant germ cell cancer of the ovary)

Malignant germ cell cancer of the ovary (type of ovarian cancer) is a very rare type of cancer. Malignant ovarian germ cell cancer is a term used to describe a group of heterogeneous rare tumours affecting the ovaries. These tumours start in the egg (ovum) producing cells of the ovary, whereas the more common epithelial ovarian cancers start in the cells that cover the surface of the ovary. Unlike epithelial ovarian cancers, these tumours are often diagnosed early and a combination of surgery and chemotherapy usually results in favourable long term overall survival. Due to its rarity, this review is based on only one very small RCT and one small retrospective study. The data from these studies were too sparse to adequately assess the effectiveness and safety of chemotherapy after surgery (adjuvant chemotherapy) in the treatment of malignant germ cell ovarian cancer. All comparisons were restricted to single study analyses and this review was only based on 32 women, so it was not adequately powered to detect differences in survival. Adverse effects of treatment and recurrence‐free survival were incompletely documented and QoL was not reported in any of the studies.We did not find any studies that reported specifically on adults as this disease usually afflicts younger people as opposed to the older population, so there were problems in separating data on adults and children in many of the studies. Many of the treatments used were taken from experiences of treating patients with testicular cancer, as they look similar under the microscope and behave similar clinically. Due to the small number of patients with malignant germ cell cancer in the two studies, our review shows that there were no good quality studies assessing the role of chemotherapy in this disease, be it in early or late stages. There was insufficient evidence to conclude that any form of chemotherapy or best supportive care is superior over the other. This review highlights the need for future good quality, well designed studies.

Background

Description of the condition

Malignant ovarian germ cell cancer is a term used to describe a group of heterogeneous rare tumours affecting the ovaries. These tumours start in the egg (ovum) producing cells of the ovary, whereas the more common epithelial ovarian cancers start in the cells that cover the surface of the ovary. They include dysgerminomas, immature teratomas, mature teratomas with malignant degeneration, mixed germ cell tumours, yolk sac tumours, embryonal carcinomas and choriocarcinomas.

These cancers are rare, accounting for around two to three percent of all malignant ovarian cancers (Quirk 2005). A woman's risk of developing an ovarian germ cell tumour by age 75 years varies between countries, ranging from 0% to 0.07% (IARC 2002).

Unlike epithelial ovarian cancers, these tumours are often diagnosed early and a combination of surgery and chemotherapy usually results in favourable long term overall survival (OS) (Gershenshon 1985).However, the evidence for the effectiveness of each intervention remains unclear.

Description of the intervention

Surgery is used both in diagnosing the stage of this disease and in its treatment. In early stage, the amount of residual disease following initial surgery is inversely correlated with outcome (Nawa 2001).

Despite apparently successful surgery, women who receive only surgical treatment can still relapse (Tewari 2000). It has been reported that the employment of chemotherapy after surgery has resulted in lowering the rates of relapse, hence improving OS (Williams 1994).A modest, but significant effect on survival has also been reported when chemotherapy is used in women with advanced disease. However, evaluations of the effectiveness of chemotherapy are often based on small numbers of patients in non‐randomised series. In addition, the drugs used in the management of this condition are selected from treatments that were successful for testicular cancers. 

Despite the potential benefits of chemotherapy in this disease, a number of side effects may be associated with this treatment. Short term effects include hair loss (alopecia), suppression of bone marrow function (which can result in leucopenia, anaemia and thrombocytopenia) and derangement of organ function (which can result in nausea, vomiting, anorexia and neurological effects). Long term side effects include problems such as fertility and development of secondary cancers.

Why it is important to do this review

We were not aware of any systematic reviews assessing the benefits of chemotherapy in the management of malignant germ cell cancer of the ovary. Current evidence in using chemotherapy in the management of this disease is based on studies of small numbers of patients; pooling the evidence from studies of similar treatments could provide more precise evidence of their effectiveness. Furthermore, a critical assessment of the quality of the studies was needed, since biases are likely to be greater for non‐randomised studies than for randomised trials.

Objectives

To evaluate the effectiveness and safety of chemotherapy in adult women with early stage, advanced and recurrent malignant germ cell ovarian cancers.

Methods

Criteria for considering studies for this review

Types of studies

• Randomised controlled trials (RCTs)

Since germ cell ovarian cancer is rare, we expected to find few (if any) RCTs, so non‐randomised studies non‐randomised studies with concurrent comparison groups were also included:

• Quasi‐randomised trials, non‐randomised trials, prospective and retrospective cohort studies, and case series of 20 or more patients.

Case‐control studies and case series of fewer than 20 patients were excluded.

In order to minimise selection bias, we decided to include only studies that used statistical adjustment for baseline case mix using multivariable analyses (e.g. age, performance status, grade, etc) if any constraints were placed on treatment allocation (e.g. women with poor performance status would not be given chemotherapy, etc) or it was based on clinician preference.

Types of participants

Adult women (over age 18 years) diagnosed with germ cell ovarian cancer who had confirmed pathological diagnoses. Patients with other concurrent malignancies will be excluded.

Types of interventions

Intervention:

• Chemotherapy (single agents or combination regimen, e.g. carboplatin or bleomycin, etoposide and platinum (BEP) or platinum, vinblastine and bleomycin (PVB) regimens).

Control:

• Best supportive care

We included comparisons of chemotherapy and best supportive care as well as direct comparisons of different types of chemotherapy.

Types of outcome measures

Primary outcomes

1. Overall survival (survival until death from any cause). Survival was assessed from the time when women were enrolled in the study.

Secondary outcomes

1. Progression‐free survival (PFS)

2. Quality of life (QoL), measured using a scale that has been validated through reporting of norms in a peer‐reviewed publication.

3. Adverse events:

   1. immediate and intermediate term systemic toxicities

      1. haematological (leucopenia, anaemia, thrombocytopenia, neutropenia, haemorrhage)

      2. gastrointestinal (nausea, vomiting, anorexia)

      3. skin (stomatitis, mucositis, alopecia, allergy)

      4. neurological (peripheral and central)

      5. pulmonary (pneumonitic/fibrosis)

   2. long term systemic treatment toxicities

      1. development of secondary cancers

      2. infertility in patients who underwent fertility sparing surgery.

Search methods for identification of studies

Papers in all languages were sought and translations were carried out when necessary.

Electronic searches

See: Cochrane Gynaecological Cancer Group methods used in reviews.
The following electronic databases were searched:

• The Cochrane Gynaecological Cancer Collaborative Review Group's Trial Register

• Cochrane Central Register of Controlled Trials (CENTRAL, Issue 1, 2010)

• MEDLINE

• EMBASE

The Medline, EMBASE and CENTRAL search strategies based on terms related to the review topic are presented in Appendix 1, Appendix 2 and Appendix 3 respectively.

Databases were searched from 1980 until April 2010.

All relevant articles found were identified on PubMed and using the 'related articles' feature, a further search was carried out for newly published articles.

Searching other resources

Unpublished and Grey literature

Metaregister, Physicians Data Query, www.controlled-trials.com/rct, www.clinicaltrials.gov and www.cancer.gov/clinicaltrials were searched for ongoing trials.

Handsearching

We performed a hand search of all available abstracts within the Proceedings of the American Society of Clinical Oncology from 2000 to April 2010.

The reference lists of all relevant trials obtained by this search were hand searched for further trials. 

Correspondence

Authors of relevant trials were contacted to ask if they knew of further data which may or may not have been published.

Data collection and analysis

Selection of studies

All titles and abstracts retrieved by electronic searching were downloaded to the reference management database Endnote. Duplicates were removed and the remaining references were examined by three review authors (AB, AR, LL) independently. Those studies which clearly did not meet the inclusion criteria were excluded and copies of the full text of potentially relevant references were obtained. The eligibility of retrieved papers was assessed independently by three review authors (AB, AR, LL). Disagreements were resolved by discussion between the three review authors. Reasons for exclusion were documented. 

Data extraction and management

For included studies, data were abstracted as recommended in chapter 7 of the Cochrane Handbook 2008. This included data on the following:

• Author, year (of publication if published) and journal citation (including language)

• Country

• Setting

• Inclusion and exclusion criteria

• Study design, methodology

• Study population

  • Total number enrolled in each group

  • Patient characteristics

  • Age

  • Co‐morbidities

  • Recurrent disease

• Malignant ovarian germ cell tumour details at diagnosis

  • FIGO stage

  • Histological cell type

  • Tumour grade

  • Extent of disease

• Intervention details

  • Details of chemotherapy

    • Dose

    • Cycle length

• Risk of bias in study (See Assessment of risk of bias in included studies)

• Duration of follow‐up

• Outcomes ‐

  • Data were extracted as below:

    • For dichotomous outcomes (e.g. adverse events or number of patients who died as it was not possible to use a hazard ratio), we extracted the number of patients in each group who experienced the outcome of interest and the number of patients assessed at endpoint, in order to estimate a risk ratio (RR).

Where possible, all data extracted were those relevant to an intention‐to‐treat (ITT) analysis, in which participants were analysed in groups to which they were assigned (to reduce bias).

Data were abstracted independently by three review authors (AB, AR, LL) onto a data abstraction form specially designed for the review. Differences between review authors were resolved by discussion.

Assessment of risk of bias in included studies

The risk of bias in the included RCT was assessed using the Cochrane Collaboration's tool and the criteria specified in chapter 8 of the Cochrane Handbook 2008. This included assessment of:

Was the allocation sequence adequately generated?

• Yes e.g. participants assigned to treatments on basis of a computer‐generated random sequence or a table of random numbers

• No e.g. participants assigned to treatments on basis of date of birth, clinic id‐number or surname, or no attempt to randomise participants

• Unclear e.g. not reported, information not available

Was allocation adequately concealed?

• Yes e.g. where the allocation sequence could not be foretold

• No e.g. open random number lists or quasi randomisation such as alternate days, odd/even date of birth, or hospital number

• Unclear e.g. not reported.

Were outcome assessors adequately prevented from knowing the allocated interventions during the study?

• Yes

• No

• Unclear.

Was loss to follow‐up less than 20% and were the reasons for loss to follow‐up similar in both arms?

• Yes

• No

• Unclear

Were reports of the study free of suggestion of selective outcome reporting?

• Yes, e.g. review adheres to protocol

• No

• Unclear.

Was the study apparently free of other problems that could put it at a high risk of bias?

• Yes

• No

• Unclear

The risk of bias in non‐randomised studies was assessed in accordance with four additional criteria:

Were details of criteria for assignment of patients to treatments provided?

• Yes

• No

• Unclear

Was the group of women who received the experimental intervention representative?

• Yes, if they were representative of women with germ cell tumour of the ovary

• No, if group of patients was selected

• Unclear, if selection of group was not described

Was the group of women who received the comparison intervention representative?

• Yes, if drawn from the same population as the experimental cohort

• No, if drawn from a different source

• Unclear, if selection of group not described

Were there no differences between the two groups or were differences controlled for, in particular with reference to age, FIGO stage, recurrent disease, histology, dose and duration of chemotherapy and type and extent of surgery?

• Yes, if at least three of these characteristics were reported and any reported differences were controlled for.

• No, if the two groups differed and differences were not controlled for.

• Unclear, fewer than three of these characteristics were reported, even if there were no other differences between the groups, and other characteristics were controlled for.

The risk of bias tool was applied independently by three review authors (AB, AR, LL) and differences were resolved by discussion. Results are presented in the risk of bias table and also in both a risk of bias graph and a risk of bias summary. Results of meta‐analyses were interpreted in light of the findings with respect to risk of bias.

Measures of treatment effect

We used the following measures of the effect of treatment:

• For dichotomous outcomes (e.g. adverse events, or time‐to‐event data if it was not possible to use a HR), we used the risk ratio.

Dealing with missing data

We did not impute missing outcome data for any of the outcomes.

Data synthesis

We identified two included studies but comparisons differed so it was not possible to perform meta‐analyses. Therefore it was not relevant to assess heterogeneity between results of studies and we were unable to assess reporting biases using funnel plots or conduct any subgroup analyses or sensitivity analyses.

Results

Description of studies

Results of the search

The search strategy identified 1815 unique references.The abstracts of these were read independently by three review authors and articles which obviously did not meet the inclusion criteria were excluded at this stage. Eighteen articles were retrieved in full and translated into English where appropriate and up‐dated versions of relevant studies were identified. The full text screening of these 18 references excluded 16 of them for the reasons described in the table Characteristics of excluded studies. However, one RCT and one non‐randomised study were identified that met our inclusion criteria and are described in the table Characteristics of included studies.

Included studies

The two included studies (Germa 1992; Sumi 2000) included 32 eligible women (20 from the Germa 1992 study and further 12 from the Sumi 2000 study), of whom all 32 were assessed at the end of the studies.

Design

The Germa 1992 study was a retrospective cohort study while the Sumi 2000 study was a randomised control trial, both of which were single institutional trials. In the Germa 1992 study, the main outcomes investigated included overall survival, toxicities as well as menstrual and reproductive outcomes; while the Sumi 2000 trial investigated parameters such as overall survival and adverse drug reaction.

Participant characteristics

We limited the Germa 1992 study population to 20 adult patients (out of 32 patients, as per our eligibility criteria). The median age of patients was 27.5 years with 15 patients (75%) with stage I/II disease while the rest had advanced disease stageIII/IV (n = 5, 25%). Histology included three patients with dysgerminoma (15%), 11 patients with immature teratoma (55%) and six patients with yolk sac tumour (30%). Sixty‐five percent of patients had a unilateral salpingo‐oophorectomy while 35% of patients had both ovaries removed. In this study, patients received best supportive care (n = 4) or chemotherapy (BEP n = 3, PVB n = 1, POMB ACE PVB n = 12). In patients that were treated with platinum based chemotherapy, the number of treatment cycles ranged from 4 to 9.

In the Sumi 2000 trial, we included 12 (out of 16 patients) who were randomly divided into two chemotherapy groups. The 12 included patients were adults and were suitable for inclusion in this review. The median age of the 12 women in this trial was 23.5 years. Eight (67%) women had early stage disease (stage I/II) and 4 (33%) had advanced stage disease (stage III). Histology included three patients with dysgerminoma  (25%), two patients with immature teratoma (16.5%) and seven patients with yolk sac tumours (58.5%). All patients received unilateral salpingo‐oophorectomy and were randomised to receive adjuvant bleomycin, actinomycin‐D and cisplatin (BAP n = 6) or bleomycin, etoposide and cisplatin (BEP n = 6).

Outcomes

In the Germa 1992 study, one death and one case of disease recurrence were reported, therefore the median survival was not reached at a median follow up of 46.5 months (range 1 to 130 months). Both disease relapse and death was in the same patient who received platinum based chemotherapy. Adverse events were incompletely reported and in most cases not possible to distinguish between adults and children. Menstrual and reproductive outcomes were not reported separately for adults and children.

In the Sumi 2000 trial, we calculated a risk ratio at five years. This approach was necessary, even though censored data could not be considered in our calculation. The calculation of hazard ratios from the published study was not possible due to the inclusion of four children. Adverse events were not reported by adults and intervention groups and individual patient data was not presented for this outcome.

Excluded studies

Sixteen references were excluded after obtaining the full text for the following reasons:

• Nine references (Chang 1994; Dimopoulos 1998; Hong 1998; Lian 1996; Mayordomo 1994; Mitchell 1999; Nawa 2001; Schwartz 1992; Segelov 1994 ) were studies that included females less than the age of 18 years where it was not possible to distinguish between adults and children. Hence, these studies did not fulfil our age criteria for inclusion.

• Two references (Culine 1995; Lokey 1981) were correspondence letters and did not include any studies which met our inclusion criteria.

• Four references (Dara 1982; Goto 1987; Linasmita 1998; Singh 1988) were case series of less than 20 patients (5, 2, 18, and 10 respectively) so did not meet our criteria for inclusion.

• One reference (Zhang 1998) did not have a suitable comparison group.

For further details of all the excluded studies see the table Characteristics of excluded studies.

Risk of bias in included studies

Both studies were at high risk of bias as they only satisfied one of the six core criteria (see Figure 1; Figure 2).

1.

Methodological quality graph: review authors' judgements about each methodological quality item presented as percentages across all included studies.

2.

Methodological quality summary: review authors' judgements about each methodological quality item for each included study.

It was unclear whether an adequate sequence had been generated to assign women to treatment groups and whether the allocation was adequately concealed in the trial of Sumi 2000. These criteria were not relevant to the non‐randomised study (Germa 1992) so this study was at high risk of bias for these individual quality items. Blinding of the outcome assessor was not reported in either study and there was insufficient information to make judgement on whether any additional risk factor for bias existed or whether there had been selective reporting of outcomes. Both studies did not analyse recurrence‐free survival, but the main outcome was overall survival and this was reported and was not statistically significant so it seems unlikely recurrence‐free survival was not reported because it was not significant. Both studies assessed an adequate proportion of their recruited women, as 100% of eligible women were assessed at the endpoint for all outcomes.

The trial of Sumi 2000 was also additionally assessed using the risk of bias criteria for non‐randomised studies because it included very sparse data and was at high risk of bias. The trial of Sumi 2000 reported a random assignment of women to the two methods of treatment and consequently we can assume that the two groups were representative of the wider population of women with malignant germ cell ovarian tumours. Details of assignment were not reported in the Germa 1992 study and the two groups were not representative as interventions were decided depending on particular disease characteristics. The data were too sparse in both studies for an accurate assessment of comparability of treatment groups as there were only 12 and 20 women in Sumi 2000 and Germa 1992 respectively who met the inclusion criteria.

Effects of interventions

See Table 1 for summary of outcomes for comparisons of interventions in included studies.

1. Summary of study comparisons, outcomes and follow‐up.

Trial ID, design and comparison	Deaths	Recurrences	Adverse events	Duration of follow‐up
Germa 1992 Retrospective study Adjuvant platinum based chemotherapy versus Best supportive care	Death at 5 years (same figures at end of study) 1/16 (6%) versus 0/4 (0%) see: Effects of interventions	Disease recurrence at 5 years (same figures at end of study) 1/16 (6%) versus 0/4 (0%)	Reversible alopecia: 16/16 (100%) versus 0/4 (0%) Nausea and vomiting: 16/16 (100%) versus 0/4 (0%) Hematologic toxicity: 2/16 (12%) versus 0/4 (0%)	Mean and median follow‐up was 58.6 and 46.5 months respectively (range: 1 to 130 months)
Sumi 2000 RCT BAP versus BEP	Death at 5 years: 1/8 (12.5%) versus 1/4 (25%)	Not reported	Not reported in adult women only.	Mean and median follow‐up was 75.3 and 50 months respectively (range from 9 to 204 months)

For time‐to‐event we were unable to estimate an adjusted HR because the data were too sparse.

For dichotomous outcomes, we were unable to estimate a RR for comparisons of treatments if one or both treatment groups experienced no events.

BAP versus BEP (Adjuvant setting)

Death at five years

The trial of Sumi 2000, found no statistically significant difference in the risk of death at 5 years in adult women with malignant germ cell ovarian cancer who received BAP and those who received BEP (RR = 0.50, 95% CI: 0.04 to 6.08). There were only two deaths in the trial; 1/8 (12.5%) and 1/4 (25%) in the BAP and BEP groups respectively.

Adverse drug reactions

Adverse drug reactions in the trial of Sumi 2000 were not reported separately for adults and children so the conclusions in the trial are based on 12 adults and four children. The trial authors reported adverse drug reactions, such as nausea and vomiting, alopecia, low haemoglobin, leukopenia, neutropenia, and thrombopenia, to be very frequently observed in both therapy groups, however, the grades of the alopecia, low haemoglobin, and neutropenia were significantly lower in the BAP group than in the BEP group.

Adjuvant platinum based chemotherapy versus best supportive care (no further treatment after surgery)

For survival outcomes we did not use statistical adjustment as specified in the methods section as there was only one observed event.

Death at five years

The Germa 1992 study did not find any statistically significant difference in the risk of death at five years between women who received adjuvant platinum based chemotherapy and those who received no further treatment after surgery. The maximum length of follow up was 130 months (mean: 58.6 months) and the study reported only one death, which was of a woman who received chemotherapy (POMB/ACE/PAV). This woman died of the disease one month after the start of treatment.

Disease recurrence at five years

The Germa 1992 study, found no statistically significant difference in the risk of disease recurrence at five years in adult women with malignant germ cell ovarian cancer who received BAP and those who received BEP. There was only one case of disease recurrence in the study and this patient died of the disease (see above).

Toxicity

All patients in the Germa 1992 study who received adjuvant platinum based chemotherapy experienced reversible alopecia, nausea and vomiting. Hematologic toxicities were observed in two patients; one who received POMB‐ACE‐PAV and the other who received BEP. Toxicity was incompletely reported and the authors did not distinguish between adults and children or it was not possible to deduce this for most of the outcomes.

Discussion

Summary of main results

We found one RCT and one retrospective study that met our inclusion criteria. The data from these studies were too sparse to adequately assess the effectiveness and safety of adjuvant chemotherapy in the treatment of malignant germ cell ovarian cancer.

All comparisons were restricted to single study analyses. We found no evidence that any form of chemotherapy prolonged the survival of women or was associated with greater safety in terms of adverse events in women with malignant germ cell ovarian cancer. However this review was only based on 32 women and meta analyses were not possible. The review was not adequately powered to detect differences in survival. We were also unable to use hazard ratios for survival as planned as they were not reported and the data were too sparse.

Given the fact that the prognosis for women with malignant germ cell ovarian cancer is good, the RCT and retrospective study both had a relatively short follow‐up period, with the RCT (Sumi 2000) having a mean follow‐up of 58.6 months and the Germa 1992 study having mean follow‐up of 75.3 months (median follow‐up was significantly shorter). Adverse effects of treatment and recurrence‐free survival were incompletely documented and QoL was not reported in either study.We did not find any studies that reported specifically on adults so there were problems in separating data on adults and children in many of the potentially relevant studies.

At present, this is still no good evidence to support the use of chemotherapy in the adjuvant or metastatic setting for ovarian germ cell tumours. There are numerous papers reporting good clinical outcomes of ovarian germ cell tumours patients who were treated with chemotherapy, many of them are retrospective in design and contain heterogenous patient population. Although the use of chemotherapy is highly logical in this disease, the exact benefit still remains unknown. Given the rarity of this disease, the only way of accruing patients satisfactorily to a clinical trial is by conducting multi centre, multi‐national studies.

Overall completeness and applicability of evidence

Overall, the quality of the evidence was low (GRADE Working Group), as the review only included a small number of women with malignant germ cell ovarian cancer (n = 32). This review included studies on adjuvant chemotherapy but does not report many of our pre‐defined outcomes or only partially reports them. Data were inadequate on PFS, QoL assessments were not reported and adverse events were incompletely reported and consequently, no meta analyses were possible. Despite the absence of high quality studies in this area, current practice in most places still include chemotherapy such as the BEP regimen, given its role in germ cell tumour of the testis. Further evaluation of the role of systemic therapy (including the use of molecularly targeted therapies) are warranted in this disease, possibly utilising a trans‐global co‐operative effort given the rarity of the disease.

Quality of the evidence

We reviewed one RCT assessing only 12 adult participants and a non‐randomised study that assessed 20 adult women. Both studies evaluated adjuvant chemotherapy for the treatment of malignant germ cell ovarian cancer; the trial of Sumi 2000 directly compared two different forms of adjuvant combination chemotherapy, while the Germa 1992 study compared adjuvant platinum based chemotherapy (single and combination) and best supportive care. The studies were not adequately powered to detect differences in survival or adverse events, especially as overall survival was assessed at five years which is insufficient follow‐up for a disease with relatively good prognosis. The Germa 1992 study also included only four control women so it is very difficult to make inferences. QoL, adverse events and progression‐free survival were inadequately documented. Therefore, from the included studies, we cannot reach any definitive conclusions about the benefit of adjuvant chemotherapy or if best supportive care is preferable.

The studies were at high risk of bias. It was unclear whether the RCT was truly randomised and both studies were at high risk of bias in all core individual risk of bias items, with the exception of loss to follow up which was deemed satisfactory as all eligible women were assessed. The authors did not estimate a hazard ratio which is the best statistic to summarise the difference in risk in two treatment groups over the duration of a study, when there is "censoring" i.e. the time to death is unknown for some women as they were still alive at the end of the study. The trial of Sumi 2000 presented a Kaplan Meier plot for overall survival and a log rank P value, but the methods of Parmar 1998 could not be used since the plot included data from adults and children.

Few adult women were included in both studies and of these there were few deaths and outcomes were incompletely reported, so there is a great need for more data to ensure higher quality evidence.

Potential biases in the review process

A comprehensive search was performed, including a thorough search of the grey literature and all studies were sifted and data extracted by two reviewers independently. We were not restrictive in our inclusion criteria with regards to types of studies as we included non‐randomised studies with concurrent comparison groups that used multivariate analyses as we suspected that we would not find any relevant RCTs. The single RCT that was found was very small and it was questionable whether it was truly randomised. Therefore we attempted to ensure that we did not overlook any relevant evidence by searching a wide range of reasonable quality non‐randomised study designs (case‐control studies and case series of fewer than 20 patients were excluded).

The greatest threat to the validity of the review is likely to be publication bias i.e. studies that did not find the treatment to have been effective may not have been published. We were unable to assess this possibility as we did not find an adequate number of studies that met the inclusion criteria.

Agreements and disagreements with other studies or reviews

There are currently no systematic reviews with similar eligibility criteria as in this review, therefore comparison is difficult. Current reviews in this area often include paediatric populations as well as testicular germ cell cancers, making direct comparisons impossible. In our review, we excluded studies which did not report or it was not possible to deduce individual results distinguishing between the paediatric and adult populations.

Authors' conclusions

Implications for practice.

At present, there is no direct evidence that the use of chemotherapy is best for the treatment of malignant germ cell cancer of the ovary. Given the successes of chemotherapy in germ cell cancers of the testis, it is unlikely that current practices will change. The included studies lacked statistical power due to the small number of women in each group and the low number of observed events. The investigation of the role of chemotherapy and the potential best combination, sequence and its short and long term toxicities is still clearly warranted.

Implications for research.

Given the findings of our review, there is a clear need for good quality RCTs or at the very least comparative non‐randomised studies that include multivariate analyses or use some kind of statistical adjustment to assess the role of chemotherapy in malignant germ cell tumours. It is essential that these studies are adequately powered to allow for a satisfactory comparison of outcomes. Future studies should report long term outcomes to allow for assessment of overall and progression‐free survival. Quality of life using an appropriate validated scale should be considered as well as severe immediate and longer term adverse events as outcomes in future studies.

What's new

Date	Event	Description
5 January 2022	Amended	No longer for update as any future update will require the development of a new protocol reflecting current Cochrane methodological criteria.

History

Protocol first published: Issue 1, 2009
Review first published: Issue 3, 2011

Appendices

Appendix 1. MEDLINE search strategy

Ovid Medline 1980 to April 2010

1   exp Ovarian Neoplasms/
2   exp Ovary/
3   ovar*.mp.
4   1 or 2 or 3
5   exp Ovum/
6   exp "Neoplasms, Germ Cell and Embryonal"/
7   "germ cell".mp.
8   dysgerminoma.mp.
9   "endodermal sinus tumor".mp.
10  "embryonal carcinoma".mp.
11  "yolk sac tumor*".mp.
12  "yolk sac tumour*".mp.
13  polyembryoma.mp.
14  choriocarcinoma.mp.
15  teratoma.mp.
16  or/5‐15
17  exp Chemotherapy, Adjuvant/
18  exp Antineoplastic Combined Chemotherapy Protocols/
19  exp Antineoplastic Agents/
20  chemother*.mp.
21  exp Bleomycin/
22  bleomycin.mp.
23  exp Etoposide/
24  etoposide.mp.
25  exp Platinum/
26  platinum.mp.
27  BEP.mp.
28  exp Vinblastine/
29  vinblastine.mp.
30  PVB.mp.
31  exp Cisplatin/
32  cisplatin.mp.
33  exp Carboplatin/
34  carboplatin.mp.
35  exp Vincristine/
36  vincristine.mp.
37  adriamycin.mp.
38  exp Doxorubicin/
39  doxorubicin.mp.
40  exp Cyclophosphamide/
41  cyclophosphamide.mp.
42  VAC.mp.
43  CAV.mp.
44  exp Methotrexate/
45  methotrexate.mp.
46  or/17‐45
47  4 and 16 and 46
48  "randomized controlled trial".pt.
49  "controlled clinical trial".pt.
50  randomized.ab.
51  placebo.ab.
52  drug therapy.fs.
53  randomly.ab.
54  trial.ab.
55  groups.ab.
56  exp Cohort Studies/
57  cohort*.ti,ab.
58  (case adj series).mp.
59  or/48‐58
60  47 and 59

key:

mp=title, original title, abstract, name of substance word, subject heading word
pt=publication type
ab=abstract
fs=floating subheading

 

Appendix 2. EMBASE search strategy

Ovid Embase 1980 to April 2010

1   exp Ovary Tumor/
2   exp OVARY/
3   ovar*.mp.
4   1 or 2 or 3
5   exp Oocyte/
6   Germ Cell Tumor/
7   exp Germ Cell Tumor/
8   "germ cell".mp.
9   dysgerminoma.mp.
10  "endodermal sinus tumor".mp.
11  "embryonal carcinoma".mp.
12  "yolk sac tumor*".mp.
13  "yolk sac tumour*".mp.
14  polyembryoma.mp.
15  choriocarcinoma.mp.
16  teratoma.mp.
17  or/5‐16
18  exp Adjuvant Chemotherapy/
19  exp Antineoplastic Agent/
20  chemother*.mp.
21  exp BLEOMYCIN/
22  bleomycin.mp.
23  exp ETOPOSIDE/
24  etoposide.mp.
25  exp PLATINUM/
26  platinum.mp.
27  BEP.mp.
28  exp VINBLASTINE/
29  vinblastine.mp.
30  PVB.mp.
31  exp CISPLATIN/
32  cisplatin.mp.
33  exp CARBOPLATIN/
34  carboplatin.mp.
35  exp VINCRISTINE/
36  vincristine.mp.
37  adriamycin.mp.
38  exp DOXORUBICIN/
39  doxorubicin.mp.
40  exp CYCLOPHOSPHAMIDE/
41  cyclophosphamide.mp.
42  VAC.mp.
43  CAV.mp.
44  exp METHOTREXATE/
45  methotrexate.mp.
46  or/18‐45
47  4 and 17 and 46
48  exp Controlled Clinical Trial/
49  randomized.ab.
50  randomly.ab.
51  trial.ab.
52  groups.ab.
53  exp Cohort Analysis/
54  cohort*.mp.
55  (case adj series).mp.
56  or/48‐55
57  47 and 56

 key:

mp=title, abstract, subject headings, heading word, drug trade name, original title, device manufacturer, drug manufacturer name ab=abstract

 

Appendix 3. CENTRAL search strategy

CENTRAL Cochrane Library Issue 1 2010

1     MeSH descriptor Ovarian Neoplasms explode all trees
#2   MeSH descriptor Ovary explode all trees
#3   ovar*
#4   (#1 OR #2 OR #3)
#5   MeSH descriptor Ovum explode all trees
#6   MeSH descriptor Neoplasms, Germ Cell and Embryonal explode all trees
#7   "germ cell"
#8   dysgerminoma
#9   "endodermal sinus tumor"
#10  "embryonal carcinoma"
#11  "yolk sac tumor*"
#12  "yolk sac tumour*"
#13  polyembryoma
#14  choriocarcinoma
#15  teratoma
#16  (#5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14 OR #15)
#17  MeSH descriptor Chemotherapy, Adjuvant explode all trees
#18  MeSH descriptor Antineoplastic Combined Chemotherapy Protocols explode all trees
#19  MeSH descriptor Antineoplastic Agents explode all trees
#20  chemother*
#21  MeSH descriptor Bleomycin explode all trees
#22  bleomycin
#23  MeSH descriptor Etoposide explode all trees
#24  etoposide
#25  MeSH descriptor Platinum explode all trees
#26  platinum
#27  "BEP"
#28  MeSH descriptor Vinblastine explode all trees
#29  vinblastine
#30  MeSH descriptor Vinblastine explode all trees
#31  "PVB"
#32  MeSH descriptor Cisplatin explode all trees
#33  cisplatin
#34  MeSH descriptor Carboplatin explode all trees
#35  carboplatin
#36  MeSH descriptor Vincristine explode all trees
#37  vincristine
#38  adriamycin
#39  MeSH descriptor Doxorubicin explode all trees
#40  doxorubicin
#41  MeSH descriptor Cyclophosphamide explode all trees
#42  cyclophosphamide
#43  "VAC"
#44  "CAV"
#45  MeSH descriptor Methotrexate explode all trees
#46  methotrexate
#47  (#17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 OR #35 OR #36 OR #37 OR #38 OR #39 OR #40 OR #41 OR #42 OR #43 OR #44 OR #45 OR #46)
#48  (#4 AND #16 AND #47)

 

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Germa 1992.

Study characteristics
Methods	Retrospective study, Department of Oncology, Hospital de la Santa Creu i Sant Pau, Spain.
Participants	33 patients with pathologically confirmed malignant ovarian germ cell tumours were included in the study. 20 patients were adults and were suitable for inclusion in the review. Mean and median age of the 20 adult women in the study was 28 and 27.5 years respectively (SD=6.1, range: 19 to 39 years). 15 (75%) women had early stage disease (10 and 5 with stage I and II respectively) and 5 (25%) had advanced stage disease (4 and 1 with stage III and IV respectively). Histology was as follows: Dysgerminoma 3 patients (15%), Immature teratoma 11 patients (55%), Yolk sac tumour 6 patients (30%). 11 (55%) patients received unilateral salpingo‐oophorectomy surgery, 6 (30%) received total abdominal hysterectomy plus bilateral salpingo‐oophorectomy, 2 (10%) had total abdominal hysterectomy plus unilateral salpingo‐oophorectomy and 1 (5%) woman received BSO surgery.
Interventions	Intervention: Adjuvant platinum based chemotherapy (POMB/ACE/PAV (n = 12), PVB (n = 1), BEP (n = 3)) Control: Best supportive care (no further treatment after surgery (n = 4)) All patients received prior surgery. The mean and median length of follow‐up of the 20 adult women in the study was 58.6 and 46.5 months respectively (range: 1 to 130 months).
Outcomes	Overall survival Toxicity Menstrual and reproductive outcome
Notes	Number of cycles of platinum based chemotherapy ranged between 4 and 9 for the 20 included patients. Only one death and case of disease recurrence occurred in the study and this was in the same patient who received platinum based chemotherapy. Adverse events were incompletely reported and in most cases it was not possible to distinguish between adults and children. Menstrual and reproductive outcomes were not reported separately for adults and children.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Patients were not randomised
Allocation concealment (selection bias)	High risk	Allocation of concealment was not applicable to this study design.
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Not reported
Incomplete outcome data (attrition bias) All outcomes	Low risk	% analysed: 20/20 (100%) adult women eligible for the review. "No patient have been lost to follow‐up".
Selective reporting (reporting bias)	Unclear risk	Insufficient information to permit judgement.
Other bias	Unclear risk	Insufficient information to assess whether an additional risk of bias may be present.
Details of assignment of patients reported?	Unclear risk	Not reported
Representative intervention group?	High risk	Patients were given multiple agent chemotherapy if they did not have completely resected stage I disease.
Representative comparison group?	High risk	"Five patients with completely resected stage I disease were intentionally not given further chemotherapy".
Comparability of treatment groups?	Unclear risk	The data were sparse and there was only one death in the entire study so we did not perform a multivariate analysis for overall survival using the individual patient data available. All women who received best supportive care (no further treatment after surgery) had stage I disease (n = 4), whereas it varied for those receiving POMB/ACE/PAV (n = 12), BEP (n = 3) and PVB (n = 1). Woman who received BEP and PVB all had stage III disease. The groups are probably not comparable, but it is unclear due to small numbers in each group.

Sumi 2000.

Study characteristics
Methods	RCT, Osaka, City Hospital, Japan.
Participants	16 ovarian germ‐cell malignancy patients were randomly divided into two chemotherapy groups. 12 patients were adults and were suitable for inclusion in the review. Mean and median age of the 12 adult women in the trial was 23.3 and 23.5 years respectively (SD = 4.4, range: 18 to 34 years). Eight (67%) women had early stage disease (7 and 1 with stage I and II respectively) and 4 (33%) had advanced stage disease (all 4 had stage III disease). Histology was as follows: Dysgerminoma 3 patients (25%), Immature teratoma 2 patients (16.5%), Yolk sac tumour 7 patients (58.5%). All patients had received unilateral salpingo‐oophorectomy surgery.
Interventions	Interventions: Adjuvant bleomycin, actinomycin‐D and cisplatin (BAP) Adjuvant bleomycin, etoposide and cisplatin (BEP) Unilateral salpingo‐oophorectomy surgery was performed as the initial operation in order to preserve fertility. Chemotherapy was administered in 3 courses at 4 week intervals. The mean and median length of follow‐up of the 12 adult women in the trial was 75.3 and 50 months respectively (range: 9 to 204 months).
Outcomes	Overall survival Adverse drug reactions
Notes	A Kaplan Meier plot comparing the two combination regimens of chemotherapy and log rank p value would have enabled a HR to be calculated for overall survival using Parmar's method, but this was not possible due to the inclusion of four children. Instead we calculated a risk ratio at five years (censoring was therefore not taken into account). Adverse events were not reported by adults and intervention groups and individual patient data was not presented for this outcome.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not reported
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Not reported
Incomplete outcome data (attrition bias) All outcomes	Low risk	% analysed: 12/12 (100%) adult women eligible for the review.
Selective reporting (reporting bias)	Unclear risk	Insufficient information to permit judgement.
Other bias	Unclear risk	Insufficient information to assess whether an additional risk of bias may be present.
Details of assignment of patients reported?	Low risk	Authors state that patients were assigned randomly
Representative intervention group?	Low risk	Patients were randomised
Representative comparison group?	Low risk	Patients were randomised
Comparability of treatment groups?	Unclear risk	Data are too sparse to make accurate assessment

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Chang 1994	Comparison of different chemotherapy regimens was possible, but ages ranged from 6 to 35 years, with a median age of 20 years. 7/35 patients were under the age of 16 and from table 3 at least one further patient was 17. The analyses do not report by age or use statistical adjustment and it was not possible to perform our own analyses as only a subset of individual patient data is available.
Culine 1995	Comparison of AVAB versus VAB‐6 regimens in 21 patients was available as individual patient data were available which reported age, recurrent disease, treatment group and survival times. However, after exclusion of children there was only a comparison of 4 patients who received AVAB and two who received VAB for recurrent disease and 4 who were given AVAB versus five patients given VAB as primary adjuvant therapy.
Dara 1982	Only five patients in case series
Dimopoulos 1998	Study included children where the median age was 20 years (range 12 to 69 years) and could not be distinguished from adults
Goto 1987	Report comprises of only two patients
Hong 1998	Study included children where the median age was 23.3 years (range: 7 to 62 years).
Lian 1996	Study does not examine women with GCT
Linasmita 1998	Only 18 women were included in this case series
Lokey 1981	Letter of correspondence
Mayordomo 1994	Study included children where the median age was 20 years (range 6 to 68 years) and could not be distinguished from adults
Mitchell 1999	The study reports an adjusted hazard ratio for overall survival for platinum versus non‐platinum based chemotherapy, but does not report by age. The median age was 21 years (range, 4 to 44 years), so a high proportion of children will have been included in the study.
Nawa 2001	The study reports comparisons of different chemotherapy regimens for treatment of malignant ovarian germ cell tumours with yolk sac histology, but does not report by age. The median age was only 18 years (range, 7 to 47 years), so a high proportion of children will have been included in the study. Age was not a significant factor in multivariate analyses, but in general tumours in children should be reported separately from adults.
Schwartz 1992	A breakdown of age was not given and multivariate analyses were not carried out. the age distribution was from 6 to 47 years.
Segelov 1994	Study included children with a median age of 19 years (range 12 to 30 years) and could not be distinguished from adults
Singh 1988	A prospective study of only 10 patients and four were children.
Zhang 1998	The authors didn't report on baseline difference. Table 1 reported demographic information on intervention group (i.e. people receiving surgery), but no demographic information on control group was provided.

Differences between protocol and review

Searches

In the protocol, we stated:

"The main investigators of any relevant ongoing trials will be contacted for further information, as will any major co‐operative trials groups active in this area."

However, we did not find any relevant ongoing trials or active trials groups, so we did not make these contacts.

Time to event and continuous outcome data

Time to event and continuous outcome data were not reported for adults in either of the two studies so the following sections in the protocol which discussed the handling of data for survival and quality of life outcomes were removed as they were unnecessary:

"Data extraction and management

• For time to event (e.g. overall and progression‐free survival) data, we will extract the log of the hazard ratio [log(HR)] and its standard error from trial reports; if these are not reported, we will attempt to estimate them from other reported statistics using the methods of Parmar 1998. 

• For continuous outcomes (e.g. quality of life (QoL)), we will extract the final value and standard deviation of the outcome of interest and the number of patients assessed at endpoint in each treatment arm at the end of follow‐up, in order to estimate the mean difference (if trials measured outcomes on the same scale) or standardised mean differences (if trials measured outcomes on different scales) between treatment arms and its standard error.

Measures of treatment effect

• For time to event data, we will use the HR, if possible. The HR summarises the chances of survival in women who received one type of treatment compared to the chances of survival in women who received another type of treatment. However, the logarithm of the HR, rather than the HR itself, is generally used in meta‐analyses".

• For continuous outcomes (e.g. QoL), we will use the mean difference between treatment arms.

Data synthesis

We identified two included studies but comparisons differed so it was not possible to perform meta‐analyses. Therefore it was not relevant to assess heterogeneity between results of studies and we were unable to assess reporting biases using funnel plots or conduct any subgroup analyses or sensitivity analyses. The following sections of the protocol were therefore removed:

"Assessment of heterogeneity  

Heterogeneity between studies will be assessed by visual inspection of forest plots, by estimation of the percentage heterogeneity between trials which cannot be ascribed to sampling variation (Higgins 2003), by a formal statistical test of the significance of the heterogeneity (Deeks 2001) and, where possible, by sub‐group analyses (see below). If there was evidence of substantial heterogeneity, the possible reasons for this were investigated and reported.

Assessment of reporting biases  

Funnel plots corresponding to meta‐analysis of the primary outcome will be examined to assess the potential for small study effects such as publication bias. If these plots suggest that treatment effects may not be sampled from a symmetric distribution, as assumed by the random effects model, further meta‐analyses will be performed using fixed effects models.

Data synthesis

If sufficient, clinically similar studies are available, their results will be pooled in meta‐analyses.

• For time‐to‐event data, HRs will be pooled using the generic inverse variance facility of RevMan 5.

• For any dichotomous outcomes, the RR will be calculated for each study and these will then be pooled. 

• For continuous outcomes, the mean differences between the treatment arms at the end of follow‐up will be pooled if all trials measured the outcome on the same scale, otherwise standardised mean differences will be pooled. 

If any trials have multiple intervention groups, the control group will be divided between the intervention groups ― to prevent double counting of participants in the meta‐analysis ― and comparisons between each intervention and a split control group will be treated independently.

Random effects models with inverse variance weighting will be used for all meta‐analyses (DerSimonian 1986).

Subgroup analysis and investigation of heterogeneity

Sub‐group analyses will be performed, grouping the trials by:

• FIGO Staging

• Type of Surgery/debulking

• Types of chemotherapy

Factors such as age, length of follow‐up and adjusted/unadjusted analysis will be considered in interpretation of any heterogeneity.

Sensitivity analysis  

Sensitivity analyses will be performed (i) excluding studies at high risk of bias and (ii) using unadjusted results.

Contributions of authors

AR, LL and ID‐P drafted the clinical sections of the review. AB drafted the methodological sections of the review. All authors agreed the final version.

Sources of support

Internal sources

• No sources of support provided

External sources

• Department of Health, UK

  NHS Cochrane Collaboration Programme. Grant Scheme CPG‐506

Declarations of interest

None known.

Edited (no change to conclusions)