Abstract
Background
This is an update of the original review published in the Cochrane Database of Systematic Reviews 2011, Issue 11, and updated in 2015, Issue 4.
Chemotherapy has significantly improved prognosis for women with malignant and some non‐malignant conditions. This treatment, however, is associated with ovarian toxicity. The use of gonadotropin‐releasing hormone (GnRH) analogues, both agonists and antagonists, may have a protective effect on the ovaries. The primary mechanism of action of GnRH analogues is to suppress the gonadotropin levels to simulate pre‐pubertal hormonal milieu and subsequently prevent primordial follicles from maturation and therefore decrease the number of follicles that are more vulnerable to chemotherapy.
Objectives
To assess the efficacy and safety of GnRH analogues given before or in parallel to chemotherapy to prevent chemotherapy‐related ovarian damage in premenopausal women with malignant or non‐malignant conditions.
Search methods
The search was run for the original review in July 2011, and for the first update in July 2014. For this update we searched the following databases in November 2018: the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, Embase, and the Chinese Biomedicine Database (CBM).
Selection criteria
Randomised controlled trials (RCTs), in all languages, which examined the effect of GnRH analogues for chemotherapy‐induced ovarian failure in premenopausal women, were eligible for inclusion in the review.
Data collection and analysis
Two review authors independently extracted data and assessed trial quality using the Cochrane 'Risk of bias' tool. We analysed binary data using risk ratios (RRs) with 95% confidence intervals (CI) and for continuous data, we used the standardized mean difference (SMD) to combine trials. We applied the random‐effects model in our analyses. We used the GRADE approach to produce a 'Summary of findings' table for our main outcomes of interest.
Main results
We included 12 RCTs involving 1369 women between the ages of 12 and 51.1 years. Participants were diagnosed with breast malignancy, ovarian malignancy, or Hodgkin's lymphoma, and most of them received alkylating, or platinum complexes, based chemotherapy. The included studies were funded by a university (n = 1), research centres (n = 4), and pharmaceutical companies (n = 1). Trials were at high or unclear risk of bias. Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone The incidence of menstruation recovery or maintenance was 178 of 239 (74.5%) in the GnRH agonist group and 110 of 221 (50.0%) in the control group during a follow‐up period no longer than 12 months (RR 1.60, 95% CI 1.14 to 2.24; 5 studies, 460 participants; I2 = 79%; low‐certainty evidence), with an overall effect favouring treatment with GnRH agonist (P = 0.006). However, we observed no difference during a follow‐up period longer than 12 months between these two groups (P = 0.24). In the GnRH agonist group, 326 of 447 participants had menstruation recovery or maintenance (72.9%) in comparison to the control group, in which 276 of 422 participants had menstruation recovery or maintenance (65.4%) during a follow‐up period longer than 12 months (RR 1.08, 95% CI 0.95 to 1.22; 8 studies, 869 participants; I2 = 56%; low‐certainty evidence).
The incidence of premature ovarian failure was 43 of 401 (10.7%) in the GnRH agonist group and 96 of 379 (25.3%) in the control group (RR 0.44, 95% CI 0.31 to 0.61; 4 studies, 780 participants; I2 = 0%; moderate‐certainty evidence), with an overall effect favouring treatment with GnRH agonist (P < 0.00001).
The incidence of pregnancy was 32 of 356 (9.0%) in the GnRH agonist group and 22 of 347 (6.3%) in the control group (RR 1.59, 95% CI 0.93 to 2.70; 7 studies, 703 participants; I2 = 0%; low‐certainty evidence), with no difference between groups (P = 0.09). However, we are cautious about this conclusion because there were insufficient data about whether the participants intended to become pregnant.
The incidence of ovulation was 29 of 47 (61.7%) in the GnRH agonist group and 12 of 48 (25.0%) in the control group (RR 2.47, 95% CI 1.43 to 4.26; 2 studies, 95 participants; I2 = 0%; low‐certainty evidence) with an overall effect favouring treatment with GnRH (P = 0.001).
The most common adverse effects of GnRH analogues included hot flushes, vaginal dryness, urogenital symptoms, and mood swings. The pooled analysis of safety data showed no difference in adverse effects between GnRH agonist group and control group. Comparison 2: GnRH agonist‐antagonist cotreatment plus chemotherapy versus chemotherapy alone Only one RCT discussed GnRH agonist‐antagonist cotreatment. The limited evidence showed the incidence of menstruation recovery or maintenance was 20 of 25 (80%) in both cotreatment group and control group during a 12‐month follow‐up period (RR 1.00, 95% CI 0.76 to 1.32; 50 participants; very low‐certainty evidence), with no difference between groups (P = 1.00). In the cotreatment group, 13 of 25 participants had menstruation recovery or maintenance (52.0%) in comparison to the control group, in which 14 of 25 participants had menstruation recovery or maintenance (56.0%) during a follow‐up period longer than 12 months (RR 0.93, 95% CI 0.56 to 1.55; 50 participants; very low‐certainty evidence), with no difference between groups (P = 0.78). The incidence of pregnancy was 1 of 25 (4.0%) in the cotreatment group and 0 of 25 (0%) in the control group (RR 3.00, 95% CI 0.13 to 70.30; 50 participants; very low‐certainty evidence), with no difference between groups (P = 0.49).
Authors' conclusions
GnRH agonist appears to be effective in protecting the ovaries during chemotherapy, in terms of maintenance and resumption of menstruation, treatment‐related premature ovarian failure and ovulation. Evidence for protection of fertility was insufficient and needs further investigation. Evidence was also insufficient to assess the effect of GnRH agonist and GnRH antagonist cotreatment on ovarian protection against chemotherapy. The included studies differed in some important aspects of design, and most of these studies had no age‐determined subgroup analysis. Large and well‐designed RCTs with longer follow‐up duration should be conducted to clarify the effects of GnRH analogues in preventing chemotherapy‐induced ovarian failure, especially on different age groups or different chemotherapy regimens. Furthermore, studies should address the effects on pregnancy rates and anti‐tumour therapy.
Plain language summary
Gonadotropin‐releasing hormone analogues for women with ovarian cancer undergoing chemotherapy
Review question This is the first update of a review published in the Cochrane Database of Systematic Reviews (2011, Issue 11). Chemotherapy has improved the prognosis for people with cancer and some non‐cancerous conditions, however, this treatment, in women can be associated with ovarian function failure. A hormone called gonadotropin‐releasing hormone (GnRH), both agonists and antagonists, may make ovaries less sensitive to the effects of chemotherapy drugs. We conducted this review to establish whether GnRH analogues can prevent damage to ovaries caused by chemotherapy in premenopausal women undergoing chemotherapy treatment for cancer or other diseases.
Study characteristics We searched the medical literature up to November 2018 and selected randomised controlled trials (RCTs), where women were randomly assigned to two (or more) groups, to test if GnRH analogues given before or alongside chemotherapy could prevent the damage to women's ovaries caused by chemotherapy. Included studies were funded by universities, research centres, or pharmaceutical companies. However, the trials were of low methodological quality.
Key results We included 12 RCTs involving 1369 women undergoing chemotherapy. The studies examined women given GnRH agonist plus chemotherapy compared with chemotherapy alone (Group 1) or women given GnRH agonist‐antagonist cotreatment plus chemotherapy compared with chemotherapy alone (Group 2).
For Group 1, we found GnRH agonist had a protective effect on ovarian function, which could reduce the rate of premature ovarian failure (moderate‐certainty evidence) and increase the rate of ovulation (low‐certainty evidence). The incidence of menstruation (periods) recovery or maintenance during 12‐month follow‐up period in the GnRH agonist group was higher than that in the control group, but we observed no difference when women were followed up for more than 12 months (low‐certainty evidence). There was no difference in pregnancy rates between groups (low‐certainty evidence); however, it was not clear if the women were trying to get pregnant. There was no difference in side effects including hot flushes, vaginal dryness, headaches, and depression between groups (very low‐ to moderate‐certainty evidence).
One RCT gave limited evidence for Group 2 and showed that GnRH agonist‐antagonist cotreatment had no protective effect on the ovaries in respect of menstruation recovery or maintenance and pregnancy rate (very low‐certainty evidence).
Conclusions GnRH agonist appears to be effective in protecting the ovaries during chemotherapy, in terms of menstruation recovery or maintenance, premature ovarian failure and ovulation. Evidence for rates of pregnancy was insufficient and needs further investigation. Evidence was also insufficient to assess the effect of GnRH agonist‐antagonist cotreatment on ovarian function with chemotherapy.
Summary of findings
Background
This review is an update of a previously published review in the Cochrane Database of Systematic Reviews (2011, Issue 11) on 'Adjuvant gonadotropin‐releasing hormone analogues for the prevention of chemotherapy induced premature ovarian failure in premenopausal women' (Chen 2011).
Description of the condition
Chemotherapy has significantly improved the prognosis in people with malignancies, including breast cancer, lymphoma and gynaecological cancers, and in some non‐malignant conditions, such as rheumatoid arthritis and systemic lupus erythematosus. This treatment, however, is associated with significant physical and psychological effects as more and more young people become long‐term survivors. Of these, ovarian toxicity has been under wider scrutiny. Animal studies have shown that chemotherapy agents, such as cyclophosphamide, busulphan, and 5‐fluorouracil, cause follicular atresia in the ovaries (Tan 2014). Oocyte death by apoptosis is one of the main mechanisms responsible for loss of germ cells (Soleimani 2011). Focal fibrosis of the ovarian cortex and damage to blood vessels are other mechanisms involved in the direct chemotherapy‐related ovarian injury (Bedoschi 2016). In addition, chemotherapy causes an increase in follicular recruitment, resulting in indirect depletion of primordial follicles and consequently ovarian failure (Kalich‐Philosoph 2013; Morgan 2012). Ovarian damage caused by chemotherapy is not an 'all or none' phenomenon. The impact ranges from partial damage resulting in irregular menses and reduced fertility up to full damage with total loss of primordial follicles and ovarian atrophy, which ultimately lead to amenorrhoea, permanent infertility and premature ovarian failure (Dayangan Sayan 2018; Kano 2017; Levine 2015; Trivers 2014). Regular menses shortly after chemotherapy does not rule out the development of premature ovarian failure, which may occur subsequently, and restoration of menstruation after chemotherapy‐related amenorrhoea is also possible (Lutchman‐Singh 2005). Women might have regular menses for several years after chemotherapy, but may have reduced fertility due to a significant reduction in ovarian reserve. Ultrasonographic (ovarian volume, antral follicle count) and biochemical markers (serum levels of follicle‐stimulating hormone (FSH), luteinizing hormone (LH), anti‐Mullerian hormone (AMH), inhibin, estradiol, and progesterone) might suggest varying levels of ovarian reserve damage before complete ovarian failure in women with preserved menstrual cycles (Bedoschi 2016). Other premature ovarian failure‐related symptoms such as increased cardiovascular risk and osteoporosis may also significantly affect a woman's overall well‐being and quality of life.
Factors that may affect the risk of chemotherapy‐induced ovarian damage include age, type and dose of chemotherapy agents, chemotherapy agents' association and length of treatment. Leonard and colleagues found a higher incidence of amenorrhoea after chemotherapy in women aged over 40 years diagnosed with breast cancer when compared with younger women (Leonard 2017). Similar results were also found by D'avila 2015, indicating that ovaries become more sensitive to chemotherapeutic agents in older women.
Description of the intervention
Gonadotropin‐releasing hormone (GnRH), also known as luteinizing‐hormone‐releasing hormone (LHRH), is a peptide hormone responsible for the release of gonadotropins (FSH and LH). GnRH is synthesized and released from the hypothalamus, and carried by the hypophyseal portal system to the pituitary gland, where it activates GnRH receptors and stimulates synthesis and secretion of LH and FSH. GnRH activity, critical for reproductive function, is very low during childhood and is activated at puberty. The frequency of the pulses varies during the menstrual cycle. Low‐frequency GnRH pulses lead to FSH release, whereas high‐frequency GnRH pulses stimulate LH release.
GnRH analogues are synthetic peptide drugs modelled on GnRH that are designed to interact with the GnRH receptors and modify the release of gonadotropins. Two types of analogues have been developed: GnRH agonists and GnRH antagonists, with specific amino acid substitutions, typically in positions 6 and 10 (goserelin and leuprorelin), or only a single substitution at position 6 (triptorelin). GnRH agonists do not degrade rapidly and do not dissociate from the GnRH receptor quickly. As a result, they trigger an initial increase in FSH and LH secretion ('flare‐up' effect), followed by a profound hypogonadal effect through receptor down‐regulation (Maggi 2016). The 'flare‐up' effect lasts for about one week (Guenther 2017), and down‐regulation of gonadotrophins can take up to two weeks after administration of GnRH agonists (Hasky 2015). Triptorelin, leuprolide, goserelin and buserelin are among the mostly used GnRH agonists. GnRH antagonists are derivatives of the natural GnRH decapeptide with multiple amino acid substitutions. Competing with GnRH for its receptor, GnRH antagonists decrease or block GnRH action to shut down the output of FSH and LH rapidly without triggering the undesirable 'flare‐up' effect (Maggi 2016). The combination of GnRH antagonist and GnRH agonist combines the quick onset of action of GnRH antagonist with the long‐lasting effects of GnRH agonist. Side effects of the GnRH analogues are similar to those of hypoestrogenaemia, including hot flushes, headaches and bone loss.
How the intervention might work
Several proposed mechanisms have been put forward on how GnRH analogues may work.
Suppression of gonadotropin levels to stimulate pre‐pubertal hormonal milieu and subsequently prevent primordial follicles from maturation and therefore decrease the number of follicles that are more vulnerable to chemotherapy (Blumenfeld 2008b; Hasky 2015)
Decrease in utero‐ovarian perfusion, resulting in a decreased exposure of the ovaries to the chemotherapeutic agents (Blumenfeld 2007; Meirow 2004)
Direct activation of GnRH receptors on ovaries (Blumenfeld 2007; Imai 2007)
Regulation of the expression of intra‐gonadal anti‐apoptotic molecules, such as sphingosine‐1‐phosphate, Bcl‐2 and caspases‐3 (Blumenfeld 2007; Huang 2009)
Protection of the undifferentiated germline stem cells (Blumenfeld 2007)
Why it is important to do this review
For adolescents and young adults taking chemotherapeutic agents, premature ovarian failure causing oestrogen deficiency symptoms and loss of fertility can significantly affect quality of life and self‐esteem. GnRH analogues are easily accessible drugs, administration is simple and there are few side effects. The possibility of administering an adjuvant GnRH analogue, which may minimize the ovarian damage caused by an otherwise successful chemotherapy, is therefore a good clinical option.
Animal studies have been conducted to investigate the protective effect of GnRH analogues on the ovaries. Li 2013 found that the percentage of normal cyclicity in rats treated with or without GnRH analogues increased from 0% with cisplatin alone to 25.0%, 33.3%, 66.7% and 41.7%, in the GnRH agonist plus cisplatin, GnRH antagonist plus cisplatin, GnRH agonist and GnRH antagonist cotreatment plus cisplatin groups, respectively. Pretreatment with GnRH agonist, GnRH antagonist and combination of GnRH agonist and antagonist significantly protected the primordial follicles from destruction by preserving 57.6%, 63.4%, 87.1% and 60.4% of the follicles, respectively. GnRH antagonists have also been found to decrease ovarian damage caused by cyclophosphamide in mice (Meirow 2004), and to deplete primordial follicles in a murine model in a prospective primary research study (Danforth 2005). However, some other studies found contradictory results. Horicks 2018 found that cyclophosphamide induced a significant follicular loss of more than 50% in mice, regardless of previous treatment with gonadotropins, and observed no difference in cell proliferation or apoptosis. Horicks 2015 found that GnRH agonist and antagonist failed to inhibit follicular development, irrespective of the doses and injection sites. They still observed a depletion of primordial follicles after chemotherapy. Peng 2007 showed that GnRH antagonists had no protective effect on ovaries from cyclophosphamide in rats.
In clinical studies, the protection effect of GnRH analogues on ovaries is still under debate. Although some guidelines and expert opinions support the use of GnRH analogues to preserve ovarian function for women with breast cancer (AGO 2017; Lambertini 2016), the administration of GnRH analogues during chemotherapy is considered an experimental strategy to preserve ovarian function and fertility as a whole due to inconsistency of results in the literature.
Objectives
To assess the efficacy and safety of GnRH analogues given before or in parallel to chemotherapy to prevent chemotherapy‐related ovarian damage in premenopausal women with malignant or non‐malignant conditions.
Methods
Criteria for considering studies for this review
Types of studies
Randomised controlled trials (RCTs) in all languages that examined the effect of GnRH analogues for chemotherapy‐induced ovarian failure in premenopausal women were eligible for inclusion in the review. We included studies published in abstract form, if extraction of data was possible, and treated these as being at high risk of bias, if insufficient information was available. Where data extraction was not possible, we included studies as awaiting classification.
Types of participants
Inclusion criteria
Premenopausal women (confirmed by age, menstrual history, ovarian sonography and/or hormone levels, such as FSH and estradiol) with malignant or non‐malignant conditions, receiving GnRH analogues supplementation before or in parallel to chemotherapy.
Exclusion criteria
Premenopausal women who had taken oral contraceptive pills or hormone replacement therapy after chemotherapy.
Types of interventions
GnRH agonists or antagonists or both versus placebo or other interventions, before or in parallel to chemotherapy
Combination of GnRH agonists and antagonists versus GnRH agonists or antagonists alone, before or in parallel to chemotherapy
GnRH agonists or antagonists or both plus other interventions versus GnRH agonists or antagonists or both, or other methods alone, before or in parallel to chemotherapy (where 'other interventions' indicate alternatives to GnRH analogues to preserve ovarian function, such as oral contraceptive pills or cryopreservation of embryo etc., and are the same in each arm of the RCT comparison)
Types of outcome measures
Primary outcomes
Menstruation recovery or maintenance: maintenance or recovery of menstruation (normal or oligomenorrhoea) with or without clear evidence of ovulation
Premature ovarian failure: no resumption of menstrual activity with postmenopausal FSH levels at least one year after the end of chemotherapy
Pregnancy
Secondary outcomes
Ovulation (confirmed by, for example, ovarian sonography or serum progesterone levels)
Ovarian sonography (antral follicle count, size of ovary, etc.)
Hormone levels (FSH, LH, AMH, inhibin, estradiol, progesterone, etc.)
Anti‐cancer efficacy (five‐year survival rate, etc.)
Adverse effects (symptoms of hypoestrogenaemia, such as hot flashes, headaches and osteoporosis, etc.)
Search methods for identification of studies
We sought papers in all languages and carried out translations where necessary.
Electronic searches
The search was run for the original review in July 2011 and subsequent searches were run in July 2014.
For this update we searched the following databases in November 2018:
the Cochrane Central Register of Controlled Trials (CENTRAL; 2018, Issue 11), in the Cochrane Library;
MEDLINE via Ovid (July 2014 to November week 2 2018);
Embase via Ovid (July 2014 to 2018 week 47);
Chinese Biomedicine Database (CBM), (July 2014 to November 2018).
We identified all relevant articles found on PubMed and using the 'related articles' feature, carried out a further search for newly published articles.
We formulated comprehensive search strategies in order to identify all relevant studies regardless of language or publication status. We have provided the search strategies in, Appendix 1 for CENTRAL, Appendix 2 for MEDLINE, Appendix 3 for Embase, and Appendix 4 for CBM.
Searching other resources
We searched the following databases for ongoing studies and conference abstracts in November 2018: www.who.int/ictrp/en/; www.clinicaltrials.gov; www.controlled-trials.com/; www.trialscentral.org/; www.asco.org/meetings; www.isiwebofknowledge.com.
We searched clinical study results for clinical trial results of marked pharmaceuticals (www.clinicalstudyresults.org). For grey literature, we searched OpenGrey database (http://www.opengrey.eu/) and Google (www.google.com).
We also searched reference lists of relevant trials and reviews.
Data collection and analysis
Selection of studies
Two review authors (HC and JL) independently examined titles and abstracts from the initial search in order to identify studies that met the inclusion criteria. We retrieved the full text of those studies thought to fulfil the inclusion criteria and those without abstracts. We have detailed reasons for excluding trials in Characteristics of excluded studies.
Data extraction and management
Two review authors (HC and JL) independently extracted data on study characteristics including methods, study quality, participants, interventions, outcomes and duration of follow‐up (Appendix 5). We resolved differences by discussion between the review authors. Review authors were blinded to the trial authors, their institutions, the source of funding and acknowledgments. HC conducted a double data entry.
Assessment of risk of bias in included studies
Two review authors (HC and JL) independently assessed risk of bias according to the guidelines stated in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2017). The 'Risk of bias' tables in the Characteristics of included studies consist of six domains, including sequence generation; allocation concealment; blinding; incomplete outcome data addressed; free of selective reporting bias; and free of other bias, with a judgement of 'Yes' indicating low risk of bias, 'No' indicating high risk of bias and 'Unclear' indicating unclear or unknown risk of bias (Appendix 6). We resolved differences by discussion among the review authors or by consulting Cochrane Gynaecological, Neuro‐oncology and Orphan Cancers.
Measures of treatment effect
We analysed data using Review Manager 5 (Review Manager 2014). We compared outcome measures for binary data using risk ratios (RRs) with 95% confidence intervals (CI). For continuous data, we used the standardized mean difference (SMD) to combine trials because all trials used different scales. Due to substantial heterogeneity in terms of different chemotherapy agents, we used random‐effects models in our analyses.
Unit of analysis issues
The included studies recruited and analysed all of the participants individually.
Dealing with missing data
We tried to contact study authors for additional or missing data, but we received no response.
Assessment of heterogeneity
We tested heterogeneity using the Cochrane Q test (Deeks 2017), with significance at a P value of less than 0.10. We used the I2 statistic (Higgins 2003), to estimate the percentage of heterogeneity between trials that could not be ascribed to sampling variation.
0% to 40%: might not be important;
30% to 60%: may represent moderate heterogeneity;
50% to 90%: may represent substantial heterogeneity;
75% to 100%: considerable heterogeneity.
If there was evidence of substantial heterogeneity, we investigated and reported the possible reasons for this.
Assessment of reporting biases
We assessed study reports, and protocols where available, to assess whether study authors reported all prespecified outcomes. In future updates of this review, if there are 10 or more studies in the meta‐analysis, we will investigate reporting biases using funnel plots.
Data synthesis
We combined the data using the random‐effects model. For dichotomous outcomes, we calculated the RR for each trial and then pooled these RRs. When there were no events in one of the groups for RR, we used the default zero‐cell correction. For continuous data, we calculated and then pooled the SMD.
Assessing the certainty of the evidence
We used the GRADE system to rate the certainty of the evidence for the seven most important outcomes. We downgraded certainty for inconsistency, design limitations (risk of bias), imprecision, indirectness and other factors, such as publication bias, where appropriate (Higgins 2011a). Two review authors (HC and JL) conducted the grading, resolving differences by discussion and, if necessary, by involving a third review author (WH).
We downgraded the evidence from 'high' certainty by one level for serious (or by two for very serious) concerns for each limitation:
High‐certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate‐certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low‐certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low‐certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
Subgroup analysis and investigation of heterogeneity
We intended to explore the following potential sources of heterogeneity using subgroup analyses. Subgroup analyses may be based on the following: the participant's age at treatment; types of diseases; types of chemotherapy agents; chemotherapy regimen (types, dose, and duration); onset of GnRH analogues supplementation; types of GnRH analogues; regimen of GnRH analogues; and follow‐up periods.
Sensitivity analysis
We performed sensitivity analysis to assess the robustness of the meta‐analyses by comparing the results using all trials and then excluding trials of lower methodological quality or those we considered to be at a higher risk of bias.
Results
Description of studies
Results of the search
The original literature search identified 2122 articles, of which we excluded 2113 as they were animal studies, reviews, and repeated reports etc. Of the remaining nine studies, we excluded another two studies with reasons (Bernhard 2007; Sverrisdottir 2009). In the end, we included four studies; two studies are still awaiting classification (Behringer 2007; Ismail‐Khan 2008), and one is still ongoing (Manger 2006).
For this update the Cochrane Gynaecological, Neuro‐Oncology and Orphan Cancer's Information Specialist ran electronic searches in November 2018.
The November 2018 search produced a list of 462 references. This list was reduced to 417 by removing duplicates. Two review authors (HC and JL) independently screened the 417 references by title and abstract, leading to the identification of 16 references for classification. We excluded four studies with reasons (Demeestere 2016; Mina 2013; Regan 2017; Zhang 2018) and included nine references; these were references related to eight RCTs. Another three studies are awaiting classification (Anderson 2016; Karimi‐Zarchi 2011; Rossi 2015).
The Internet databases search produced a list of 247 references. Two review authors (HC and JL) independently screened the references by title, abstract, or protocol, leading to the identification of five ongoing RCTs.
The review therefore comprises 12 included studies (involving 13 articles) (Characteristics of included studies), six ongoing studies (Characteristics of ongoing studies), and five studies awaiting classification (Characteristics of studies awaiting classification). We have included the study selection as a flow chart (Figure 1).
1.
Flow diagram
Included studies
We attempted to contact authors of 10 included studies by email for additional or missing data, but received no replies (Badawy 2009; Elgindy 2013; Gerber 2011; Giuseppe 2007; Karimi‐Zarchi 2014; Lambertini 2015a; Leonard 2017; Moore 2015; Munster 2012; Song 2013).
Population
The total number of randomised participants was 1369. Lambertini 2015a contributed the largest sample of 281, while Waxman 1987 recruited only 18 participants. All included studies recruited women menstruating normally and between the ages of 12 and 51.1 years. Women were diagnosed with breast cancer (Badawy 2009; Elgindy 2013; Gerber 2011; Karimi‐Zarchi 2014; Lambertini 2015a; Leonard 2017; Moore 2015; Munster 2012; Song 2013), ovarian cancer (Gilani 2007), or Hodgkin's lymphoma (Giuseppe 2007; Waxman 1987). Women in most included studies received alkylating‐ or platinum complexes‐based chemotherapy (Badawy 2009; Elgindy 2013; Gerber 2011; Gilani 2007; Giuseppe 2007; Karimi‐Zarchi 2014; Lambertini 2015a; Leonard 2017; Munster 2012; Song 2013).
Settings
Of the 12 studies identified for inclusion in the review, two studies were conducted in Egypt (Badawy 2009; Elgindy 2013), two in England (UK) (Leonard 2017; Waxman 1987), two in Italy (Giuseppe 2007; Lambertini 2015a), and two in Iran (Gilani 2007; Karimi‐Zarchi 2014); one each in China (Song 2013), Germany (Gerber 2011), and the USA (Munster 2012); and one international, multi‐centre trial in Australia, Switzerland, Peru and USA (Moore 2015).
Intervention and comparators
All included studies compared GnRH agonist plus chemotherapy versus chemotherapy alone. Elgindy 2013 also compared combined GnRH agonist and GnRH antagonist plus chemotherapy versus chemotherapy alone.
Follow‐up
Women in all studies were followed up for more than 12 months except in three studies (Badawy 2009; Gilani 2007; Karimi‐Zarchi 2014). Participants in seven studies were followed up for no less than two years (Gerber 2011; Giuseppe 2007; Lambertini 2015a; Leonard 2017; Moore 2015; Munster 2012; Waxman 1987). The mean time of follow‐up in Giuseppe 2007 was 4.2 years and the median long‐term follow‐up was 7.3 and 4.1 years in Lambertini 2015a and Moore 2015 respectively.
Outcomes
Primary outcome
Eleven studies reported the rates of menstruation recovery or maintenance (Badawy 2009; Elgindy 2013; Gerber 2011; Gilani 2007; Giuseppe 2007; Karimi‐Zarchi 2014; Lambertini 2015a; Leonard 2017; Munster 2012; Song 2013; Waxman 1987). Four studies measured treatment‐related premature ovarian failure using combination of amenorrhoea and postmenopausal FSH levels (Lambertini 2015a; Leonard 2017; Moore 2015; Song 2013 ). Seven studies reported the rates of pregnancy (Elgindy 2013; Gerber 2011; Giuseppe 2007; Lambertini 2015a; Moore 2015; Munster 2012; Waxman 1987).
Secondary outcomes
Badawy 2009 and Waxman 1987 reported ovulation rates. Elgindy 2013, Gerber 2011 and Giuseppe 2007 reported ultrasound antral follicle count. Hormone levels were reported outcomes in ten listed studies (Badawy 2009; Elgindy 2013; Gerber 2011; Gilani 2007; Giuseppe 2007; Karimi‐Zarchi 2014; Lambertini 2015a; Leonard 2017; Munster 2012; Song 2013). Lambertini 2015a and Moore 2015 reported disease‐free survival. Five studies reported adverse events (Elgindy 2013; Gerber 2011; Lambertini 2015a; Moore 2015; Song 2013).
Funding sources
One included study was funded by a university (Elgindy 2013), four by research centres (Lambertini 2015a; Leonard 2017; Moore 2015; Munster 2012), and one by a pharmaceutical company (Gerber 2011). Six of the included studies did not mention the funding sources (Badawy 2009; Gilani 2007; Giuseppe 2007; Karimi‐Zarchi 2014; Song 2013; Waxman 1987).
Further information is available in the Characteristics of included studies table.
Excluded studies
We excluded six studies. Bernhard 2007 gave the intervention (goserelin) after chemotherapy; the women received hormone replacement therapy during chemotherapy in Demeestere 2016; Mina 2013 did not discuss the primary outcomes of this review; not all women in treatment group received GnRH analogues in Regan 2017; not all women received chemotherapy in Sverrisdottir 2009; and all included women received sequential or simultaneous GnRH analogues in Zhang 2018. For further details please refer to the Characteristics of excluded studies table.
Risk of bias in included studies
We judged seven included RCTs as unclear risk of bias due to one or more criteria assessed being unclear (Badawy 2009; Gilani 2007; Giuseppe 2007; Karimi‐Zarchi 2014; Lambertini 2015a; Leonard 2017; Song 2013), while another five included RCTs were at high risk of bias due to one or more criteria assessed being high (Elgindy 2013; Gerber 2011; Moore 2015; Munster 2012; Waxman 1987). Random sequence generation, allocation concealment and blinding in most studies were unclear. Please see Figure 2 and Figure 3. Detailed information is available in the Characteristics of included studies table.
2.
'Risk of bias' summary: review authors' judgements about each risk of bias item for each included study
3.
'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies
Allocation
We rated four studies at low risk of bias for randomisation because they used computer‐generated lists of random numbers (Elgindy 2013; Lambertini 2015a; Leonard 2017; Munster 2012).
We rated eight studies at unclear risk of bias for randomisation; none provided details of the generation of the randomisation sequence (Badawy 2009; Gerber 2011; Gilani 2007; Giuseppe 2007; Karimi‐Zarchi 2014; Moore 2015; Song 2013; Waxman 1987).
We rated four studies at low risk of bias for allocation concealment because they reported using unbiased methods of concealment. Methods included: using a clinical trials centre to keep the randomisation code, or sealed envelopes (Badawy 2009; Elgindy 2013; Lambertini 2015a; Leonard 2017).
We rated eight studies at unclear risk of bias for allocation concealment; none provided information on how they had achieved allocation concealment (Gerber 2011; Gilani 2007; Giuseppe 2007; Karimi‐Zarchi 2014; Moore 2015; Munster 2012; Song 2013; Waxman 1987).
Blinding
We rated nine studies at unclear risk of bias for blinding; none mentioned whether participant/personnel/outcome assessment was blinded or not (Badawy 2009; Gilani 2007; Giuseppe 2007; Karimi‐Zarchi 2014; Lambertini 2015a; Leonard 2017; Munster 2012; Song 2013; Waxman 1987).
We rated three studies at high risk of bias for blinding; these studies were open‐label trials without blinding (Elgindy 2013; Gerber 2011; Moore 2015).
Although most studies did not indicate whether they had implemented blinding, the impact on primary outcomes, including menstruation recovery or maintenance, premature ovarian failure and pregnancy was limited, because the primary outcomes were objective indicators.
Incomplete outcome data
We assessed all studies at low risk of attrition bias because loss to follow‐up was balanced across these groups with rational reasons. Badawy 2009 reported that only one woman (2.5%) in each group dropped out. In Elgindy 2013, five women (5%) died during the follow‐up period: one woman in each group except the GnRH agonist plus delayed chemotherapy group, in which two women died. In Gerber 2011, the risk of bias in the primary outcomes is low as only two women (6.7%) in each group withdrew due to adverse events, progression of disease and being unwilling to continue to participate. All women were followed up in Gilani 2007, Giuseppe 2007 and Karimi‐Zarchi 2014. In the 12‐month follow‐up of Lambertini 2015a, six women (4.5%) in the chemotherapy alone group did not receive chemotherapy and six (4.5%) were lost to follow‐up, while two women (1.4%) in the GnRH agonist plus chemotherapy group did not receive chemotherapy and seven (4.7%) were lost to follow‐up. At the study cut‐off date for long follow‐up (median follow‐up was 7.3 years), 18 (13.5%) and 17 (11.5%) women were lost to follow‐up in the chemotherapy‐alone group and chemotherapy plus triptorelin group respectively. In Leonard 2017, three women (2.8% in GnRH group and 2.4% in control group) in each arm had died within 24 months of randomisation. For a further 19 women (11 in the control arm and 8 in the intervention arm), menstrual status during the interval between the 12‐month follow‐up visit and the 24‐month follow‐up visit could not be determined from the data available. Moore 2015 collected data on pregnancy and survival in 93.6% women. Attrition was mainly due to withdrawal, loss to follow‐up and non‐response. Munster 2012 reported that one woman (3.7%) in the GnRH agonist group withdrew consent before the first treatment. Another woman (4.5%), in the control arm, was found to have residual disease after neoadjuvant chemotherapy and was treated with additional chemotherapy not permitted by protocol. Song 2013 reported that 16 (14.5%) in the chemotherapy‐alone group (two women failed to complete at least one cycle of chemotherapy and 14 women were lost to follow‐up), and 21 (19.1%) in the chemotherapy plus GnRH agonist group (four women failed to complete at least one cycle of chemotherapy and 17 women were lost to follow‐up), were not included in the evaluation. Waxman 1987 reported that one woman (5.6%) was not included in the evaluation because she died during treatment.
Selective reporting
We found protocols for five studies (Gerber 2011; Lambertini 2015a; Leonard 2017; Moore 2015; Munster 2012) and all these studies except Munster 2012 reported all the outcomes mentioned in their protocol. Protocols were not available for seven included studies. We compared the methods and results sections of each study and found consistent reporting between sections in six studies (Badawy 2009; Gilani 2007; Giuseppe 2007; Karimi‐Zarchi 2014; Song 2013; Waxman 1987). Elgindy 2013 mentioned adverse events as an outcome measurement but did not report the results. Thus, we assessed ten studies in this review at low risk of reporting bias (Badawy 2009; Gerber 2011; Gilani 2007; Giuseppe 2007; Lambertini 2015a; Karimi‐Zarchi 2014; Leonard 2017; Moore 2015; Song 2013; Waxman 1987) and two studies at high risk (Elgindy 2013; Munster 2012).
Other potential sources of bias
We assessed Lambertini 2015a at low risk of other potential bias because there was no clear violation of the protocol and no significant difference in baseline characteristics between groups. The original protocol of Leonard 2017 restricted the entry of women to those with oestrogen receptor (ER)‐negative tumours only, but women with ER‐positive tumours, for whom the investigator did not deem ovarian suppression necessary as part of the treatment, were subsequently allowed entry to the study after a protocol amendment. There was no other violation of the protocol and no significant difference in baseline characteristics between groups. We also assessed this study at low risk of other potential bias.
There was no significant difference in baseline characteristics between groups in another seven studies (Badawy 2009; Elgindy 2013; Gerber 2011; Gilani 2007; Giuseppe 2007; Karimi‐Zarchi 2014; Song 2013). However, the protocols were not mentioned. Thus, we assessed these studies at unclear risk of other potential bias.
We assessed three studies at high risk of other potential bias (Moore 2015; Munster 2012; Waxman 1987). Moore 2015 closed early owing to loss of funding for study‐drug distribution. Munster 2012 initially planned to enrol 124 women, but the trial was stopped for futility after 49 women were enrolled. The planned follow‐up time was five years but was stopped early because no difference between the two groups was identified. In Waxman 1987, women in the treatment groups tended to be younger than those in the control groups and the low dose of buserelin used could have led to incomplete pituitary downregulation, which might have affected the results.
Effects of interventions
Summary of findings 1. GnRH agonist plus chemotherapy versus chemotherapy alone for the prevention of chemotherapy‐induced premature ovarian failure in premenopausal women.
| GnRH agonist plus chemotherapy versus chemotherapy alone for the prevention of chemotherapy‐induced premature ovarian failure in premenopausal women | ||||||
|
Patient or population: premenopausal women with malignant conditions, receiving GnRH agonist before or in parallel to chemotherapy
Settings: hospital
Intervention: GnRH agonist plus chemotherapy Comparison: chemotherapy alone | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect (95% CI) | No of participants (studies) | Certainty of the evidence (GRADE) | Comments | |
| Assumed risk | Corresponding risk | |||||
| Control: chemotherapy alone | GnRH agonist plus chemotherapy | |||||
|
Menstruation recovery or maintenance: Follow‐up ≤ 12 months |
Study population | RR 1.60 (1.14 to 2.24) | 460 (5 studies) | ⊕⊕⊝⊝ Low1,2 | ||
| 498 per 1000 | 796 per 1000 (567 to 1000) | |||||
| Moderate | ||||||
| 496 per 1000 | 794 per 1000 (565 to 1000) | |||||
|
Menstruation recovery or maintenance: Follow‐up > 12 months |
Study population | RR 1.08 (0.95 to 1.22) | 869 (8 studies) |
⊕⊕⊝⊝ Low3,4 | ||
| 654 per 1000 | 706 per 1000 (621 to 798) | |||||
| Moderate | ||||||
| 681 per 1000 | 735 per 1000 (647 to 831) | |||||
| Premature ovarian failure | Study population | RR 0.44 (0.31 to 0.61) | 780 (4 studies) | ⊕⊕⊝⊝ Moderate5 | We defined premature ovarian failure as amenorrhoea after chemotherapy with postmenopausal FSH levels. Follow‐up periods were ≥ 12 months | |
| 253 per 1000 | 111 per 1000 (79 to 155) | |||||
| Moderate | ||||||
| 249 per 1000 | 110 per 1000 (77 to 152) | |||||
| Pregnancy | Study population | RR 1.59 (0.93 to 2.70) | 703 (7 studies) | ⊕⊕⊝⊝ Low6,7 | Follow‐up periods were ≥ 18 months | |
| 63 per 1000 | 101 per 1000 (59 to 171) | |||||
| Moderate | ||||||
| 95 per 1000 | 151 per 1000 (88 to 257) | |||||
| Ovulation | Study population |
RR 2.47 (1.43 to 4.26) |
95 (2 studies) |
⊕⊕⊝⊝ Low8,9 | Follow‐up periods ranged from 8 months to > 3 years | |
| 250 per 1000 |
618 per 1000 (357 to 1000) |
|||||
| Moderate | ||||||
| 239 per 1000 |
590 per 1000 (342 to 1000) |
|||||
| Adverse effects: hot flush | Study population | RR 1.49 (0.16 to 13.60) | 731 (4 studies) | ⊕⊝⊝⊝ Very low10,11,12 | Follow‐up periods ranged from 1 year to > 7 years | |
| 390 per 1000 | 580 per 1000 (62 to 1000) | |||||
| Moderate | ||||||
| 245 per 1000 | 365 per 1000 (39 to 1000) | |||||
| Adverse effects: vaginal dryness | Study population | RR 1.18 (0.68 to 2.04) | 488 (2 studies) | ⊕⊕⊕⊝ Moderate13 | Follow‐up periods ranged from 5 months to > 7 years | |
| 88 per 1000 | 104 per 1000 (60 to 180) | |||||
| Moderate | ||||||
| 88 per 1000 | 104 per 1000 (60 to 180) | |||||
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; FSH: follicle‐stimulating hormone; RR: risk ratio | ||||||
| GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect. | ||||||
1Downgraded by one level due to concern about risk of bias (unclear random sequence generation in three out of five studies, unclear allocation concealment in two studies, unclear blinding in four studies, no blinding in one study, and selective reporting in one study). 2Downgraded one level due to substantial heterogeneity among results of included studies (I2= 79%). 3Downgraded by one level due to concern about risk of bias (unclear random sequence generation in four out of eight studies, unclear allocation concealment in five studies, unclear blinding in six studies, no blinding in two studies, selective reporting in two studies, and other biases in two studies). 4Downgraded one level due to substantial heterogeneity among results of included studies (I2= 56%). 5Downgraded by one level due to concern about risk of bias (unclear random sequence generation and unclear allocation concealment in one out of four studies, unclear blinding in all studies, and other bias in one study). 6Downgraded by one level due to concern about risk of bias (unclear random sequence generation in four out of seven studies, unclear allocation concealment in five studies, unclear blinding in all studies, incomplete outcome data in one study, selective reporting in two studies and other bias in four studies). 7Downgraded by one level due to concern about indirectness (insufficient data about the participants' intention of pregnancy). 8Downgraded by one level due to concern about risk of bias (unclear allocation concealment in one out of two studies, unclear random sequence generation and unclear blinding in all studies, and other bias in one study). 9Downgraded by one level due to imprecision (small sample size). 10Downgraded by one level due to concern about risk of bias (unclear random sequence generation, allocation concealment and other bias in three out of four studies; unclear blinding in all studies). 11Downgraded one level due to considerable heterogeneity among results of included studies (I2= 99%). 12Downgraded by one level due to imprecision (wide confidence interval crossing the line of no effect). 13Downgraded by one level due to concern about risk of bias (unclear random sequence generation and unclear allocation concealment in one out of two studies, unclear blinding in both two studies, and other bias in one study).
Summary of findings 2. GnRH agonist‐antagonist cotreatment plus chemotherapy versus chemotherapy alone for the prevention of chemotherapy‐induced premature ovarian failure in premenopausal women.
| GnRH agonist‐antagonist cotreatment plus chemotherapy versus chemotherapy alone for the prevention of chemotherapy induced premature ovarian failure in premenopausal women | ||||||
|
Patient or population: premenopausal women with malignant conditions, receiving GnRH agonist before or in parallel to chemotherapy
Settings: hospital
Intervention: GnRH agonist‐antagonist cotreatment plus chemotherapy Comparison: chemotherapy alone | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect (95% CI) | No of participants (studies) | Certainty of the evidence (GRADE) | Comments | |
| Assumed risk | Corresponding risk | |||||
| Control: chemotherapy alone | Agonist‐antagonist cotreatment plus chemotherapy | |||||
|
Menstruation recovery or maintenance: Follow‐up ≤ 12 months |
Study population | RR 1.00 (0.76 to 1.32) | 50 (1 study) | ⊕⊝⊝⊝ Very low1,2 | ||
| 800 per 1000 | 800 per 1000 (608 to 1000) | |||||
| Moderate | ||||||
| 800 per 1000 | 800 per 1000 (608 to 1000) | |||||
|
Menstruation recovery or maintenance: Follow‐up > 12 months |
Study population | RR 0.93 (0.56 to 1.55) | 50 (1 study) | ⊕⊝⊝⊝ Very low1,2 | ||
| 560 per 1000 | 521 per 1000 (314 to 868) | |||||
| Moderate | ||||||
| 560 per 1000 | 521 per 1000 (314 to 868) | |||||
| Premature ovarian failure | ‐ | ‐ | Not estimable | 0 | ‐ | No data. We defined premature ovarian failure as amenorrhoea after chemotherapy with postmenopausal FSH levels |
| Pregnancy | Study population | RR 3.00 (0.13 to 70.30) | 50 (1 study) | ⊕⊝⊝⊝ Very low1,2 | Follow‐up: 18 months | |
| 0 per 1000 | 0 per 1000 (0 to 0) | |||||
| Moderate | ||||||
| 0 per 1000 | 0 per 1000 (0 to 0) | |||||
| Ovulation | ‐ | ‐ | Not estimable | 0 | ‐ | No data |
| Adverse events | ‐ | ‐ | Not estimable | 0 | ‐ | No data |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; FSH: follicle‐stimulating hormone; RR: risk ratio | ||||||
| GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect. | ||||||
1 Downgraded by one level due to concern about risk of bias (unclear blinding, and high risk of selective reporting). 2 Downgraded by one level due to serious imprecision (single RCT with small sample size).
See Table 1 and Table 2. We obtained all data from published literature. We attempted to contact authors of included studies by email for additional or missing data, but received no replies (Badawy 2009; Elgindy 2013; Gerber 2011; Giuseppe 2007; Karimi‐Zarchi 2014; Lambertini 2015a; Leonard 2017; Moore 2015; Munster 2012; Song 2013). For those data that were incomplete and could not be analysed in the meta‐analysis, we tried to report them narratively or in additional tables wherever possible.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone
Primary outcomes
Menstruation recovery or maintenance
Subgroup analysis based on follow‐up period of 12 months or less
Based on results from five included studies (Badawy 2009; Elgindy 2013; Gilani 2007; Karimi‐Zarchi 2014; Lambertini 2015a), the incidence of menstruation recovery or maintenance was 178 of 239 (74.5%) in the GnRH agonist plus chemotherapy group and 110 of 221 (50.0%) in the chemotherapy alone group during a follow‐up period no longer than 12 months. The RR was 1.60 (95% CI 1.14 to 2.24; I2 = 79%; Analysis 1.1), with an overall effect favouring treatment with GnRH agonist (P = 0.006). We detected heterogeneity among results of included studies (I2 = 79%; P = 0.0007), then we performed sensitivity analysis excluding one study that was at high risk of bias (Elgindy 2013). When this study was excluded, the result of the meta‐analysis was unchanged (RR 1.82, 95% CI 1.23 to 2.67; 4 studies, 410 participants; P = 0.002; I2 = 75%). We downgraded the evidence for this outcome to low certainty, for risk of bias and inconsistency (Table 1).
1.1. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 1: Menstruation recovery or maintenance
Subgroup analysis based on follow‐up period of more than 12 months
Based on results from eight studies (Elgindy 2013; Gerber 2011; Giuseppe 2007; Lambertini 2015a; Leonard 2017; Munster 2012; Song 2013; Waxman 1987), in the GnRH agonist plus chemotherapy group, 326 of 447 participants had menstruation recovery or maintenance (72.9%) in comparison to the chemotherapy alone group, in which 276 of 422 participants had menstruation recovery or maintenance (65.4%) during a follow‐up period longer than 12 months. The RR was 1.08 (95% CI 0.95 to 1.22; 8 studies, 869 participants; I2 = 56%; Analysis 1.1), with no difference between groups (P = 0.24). We detected statistical heterogeneity among the results of the included studies (I2 = 56%; P = 0.02), then we performed sensitivity analysis excluding four included studies that were at high risk of bias (Elgindy 2013; Gerber 2011; Munster 2012; Waxman 1987). When these four studies were excluded, the result of meta‐analysis was unchanged (RR 1.20, 95% CI 0.98 to 1.46; 4 studies, 695 participants; P = 0.08; I2 = 66%). We downgraded the evidence for this outcome to low certainty, for risk of bias and inconsistency (Table 1).
Subgroup analysis based on age
Two studies reported data regarding different age subgroups (Karimi‐Zarchi 2014; Leonard 2017), with inconsistent results. Leonard 2017 reported a protective effect of GnRH agonist on menses in women aged 40 years or older, while the effect was no different in women younger than 40 years (Table 3). However, Karimi‐Zarchi 2014 found that GnRH agonist had a protective effect regarding menstruation in women aged 35 years and older, and the effect was no different in women younger than 35 years (Table 4).
1. Leonard 2017: amenorrhoea and premature ovarian failure in age subgroups.
| Age (years) | Percentage of women with amenorrhoea/POF | P value | ||
| Chemotherapy+GnRH agonist group | Chemotherapy group | |||
| Amenorrhoea | ≤ 40 | 10% | 25.4% | 0.032 |
| > 40 | 42.9% | 54.2% | 0.376 | |
| POF | ≤ 40 | 2.6% | 20.0% | 0.038 |
| > 40 | 42.3% | 47.2% | 0.798 | |
| POF: premature ovarian failure; GnRH: gonadotropin‐releasing hormone | ||||
2. Karimi‐Zarchi 2014: menstruation recovery or maintenance at six months after chemotherapy in age subgroups.
| Age (years) | Number of participants with menstruation recovery or maintenance/total number of participants | RR | 95% CI | P value | |
| Chemotherapy+GnRH agonist group | Chemotherapy group | ||||
| < 35 | 13/13 | 3/15 | 1.5 | 0.15 to 15.46 | 0.73 |
| > 35 | 6/8 | 4/6 | 96.43 | 4.51 to 2059.53 | 0.003 |
| GnRH: gonadotropin‐releasing hormone; CI: confidence interval; RR: risk ratio | |||||
Premature ovarian failure
Combining results from four studies (Lambertini 2015a; Leonard 2017; Moore 2015; Song 2013), the incidence of treatment‐related premature ovarian failure was 43 of 401 (10.7%) in the GnRH agonist plus chemotherapy group and 96 of 379 (25.3%) in the chemotherapy alone group. The RR was 0.44 (95% CI 0.31 to 0.61; 4 studies, 780 participants; I2 = 0%; Analysis 1.2), with an overall effect favouring treatment with GnRH (P < 0.00001). We performed sensitivity analysis for this outcome as one included study was at high risk of bias (Moore 2015). When this study was excluded, the result of meta‐analysis was unchanged (RR 0.45, 95% CI 0.31 to 0.65; 3 studies, 645 participants; P < 0.0001; I2 = 9%). We downgraded the evidence for this outcome to moderate certainty for risk of bias (Table 1).
1.2. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 2: Premature ovarian failure (POF)
Subgroup analysis based on age
Leonard 2017 reported a protective effect of GnRH agonist on ovaries in respect of premature ovarian failure rates in women aged 40 years or older, while the effect was no different in women younger than 40 years (Table 3). However, Song 2013 reported no difference in premature ovarian failure rates in different age subgroups (women aged 35 years and older, and women younger than 35 years) between the chemotherapy plus GnRH agonist group and the chemotherapy alone group (Table 5).
3. Song 2013: premature ovarian failure in age subgroups.
| Age (years) | Number of participants with POF/ total number of participants | RR | 95% CI | P value | |
| Chemotherapy+GnRH agonist group | Chemotherapy group | ||||
| ≤ 35 | 5/43 | 10/41 | 0.48 | 0.18 to 1.28 | 0.14 |
| > 35 | 10/46 | 17/43 | 0.55 | 0.28 to 1.07 | 0.08 |
| POF: premature ovarian failure; GnRH: gonadotropin‐releasing hormone; CI: confidence interval; RR: risk ratio | |||||
Pregnancy
Based on results from seven included studies (Elgindy 2013; Gerber 2011; Giuseppe 2007; Lambertini 2015a; Moore 2015; Munster 2012; Waxman 1987), the incidence of pregnancy was 32 of 356 (9.0%) in the GnRH agonist plus chemotherapy group and 22 of 347 (6.3%) in the chemotherapy alone group. The RR was 1.59 (95% CI 0.93 to 2.70; 7 studies, 703 participants; I2 = 0%; Analysis 1.3), with no difference between groups (P = 0.09). The follow‐up periods were no less than 18 months. Five of the included studies were at high risk of bias (Elgindy 2013; Gerber 2011; Moore 2015; Munster 2012; Waxman 1987). Sensitivity analysis excluding these five studies gave the same result (RR 1.04, 95% CI 0.11 to 10.06; 310 participants; P = 0.97; I2 = 54%). We downgraded the evidence for this outcome to low certainty for risk of bias for risk of bias and indirectness (Table 1).
1.3. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 3: Pregnancy
Secondary outcomes
Ovulation
Based on results from two included studies (Badawy 2009; Waxman 1987), the incidence of ovulation was 29 of 47 (61.7%) in the GnRH agonist plus chemotherapy group and 12 of 48 (25.0%) in the chemotherapy alone group. The RR was 2.47 (95% CI 1.43 to 4.26; 2 studies, 95 participants; I2 = 0%; Analysis 1.4), with an overall effect favouring treatment with GnRH (P = 0.001). Follow‐up periods ranged from eight months to more than three years. We performed sensitivity analysis for this outcome excluding one study with high risk of bias (Waxman 1987). Only one trial remained and the result was the same (RR 2.70, 95% CI 1.52 to 4.79; 78 participants; P = 0.0007). We downgraded the evidence for this outcome to low certainty for risk of bias and imprecision.
1.4. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 4: Ovulation
Antral follicle count
Three studies reported ultrasound antral follicle count (Elgindy 2013; Gerber 2011; Giuseppe 2007). Giuseppe 2007 found there was no difference in antral follicle count between chemotherapy plus GnRH agonist and chemotherapy alone (SMD 1.11, 95% CI 0.32 to 1.90; 1 study, 29 participants; P = 0.006; Analysis 1.5). Elgindy 2013 did not provide data that could be meta‐analysed; however, this study came to the same conclusion, that there was no difference in antral follicle count between groups. Data regarding antral follicle count in Gerber 2011 were incomplete and could not be analysed.
1.5. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 5: Antral follicle count
Hormone levels
Ten included studies reported hormone levels (Badawy 2009; Elgindy 2013; Gerber 2011; Gilani 2007; Giuseppe 2007; Karimi‐Zarchi 2014; Lambertini 2015a; Leonard 2017; Munster 2012; Song 2013), however, most data were not presented in a form that could be meta‐analysed. We tried to report these data narratively wherever possible.
FSH/LH
Pooled results of Giuseppe 2007 and Karimi‐Zarchi 2014 found that there was no difference in FSH/LH levels between chemotherapy plus GnRH agonist group and chemotherapy alone group (for FSH: SMD 0.26, 95% CI −0.80 to 1.31, P = 0.63; 2 studies, 71 participants; I2 = 79%; Analysis 1.6; for LH: SMD −0.62, 95% CI −1.28 to 0.03; P = 0.06; I2 = 45%; Analysis 1.8). Three other studies found the same results (Elgindy 2013; Lambertini 2015a; Munster 2012). However, Gilani 2007 reported protective effects by GnRH agonist on ovaries in respect of FSH and LH levels (RR 1.48, 95% CI 1.02 to 2.13; 30 participants; P = 0.04; Analysis 1.7;Analysis 1.9). This conclusion was consistent with Badawy 2009 (P < 0.009 for FSH levels and P < 0.004 for LH levels), Gerber 2011 (P = 0.015 for LH levels), and Song 2013 (P < 0.05 for FSH levels).
1.6. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 6: FSH (mUI/L)
1.8. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 8: LH (mUI/L)
1.7. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 7: FSH < 20 mIU/mL
1.9. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 9: LH < 20 mIU/L
AMH
Giuseppe 2007 found there was no difference in AMH levels between groups (SMD −0.05, 95% CI −0.78 to 0.68; 29 participants; P = 0.89; Analysis 1.10). This result was consistent with Elgindy 2013. Gerber 2011 reported data regarding AMH incompletely and so we could not analyse them.
1.10. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 10: AMH (pmol/L)
Inhibin B
Giuseppe 2007 found no difference in inhibin B levels between chemotherapy plus GnRH agonist group and chemotherapy alone groups (SMD 0.06, 95% CI ‐0.67 to 0.78; 29 participants; P = 0.88; Analysis 1.11). Gerber 2011 and Munster 2012 reported data regarding inhibin B levels incompletely and we could not analyse them.
1.11. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 11: Inhibin B
Estradiol
Karimi‐Zarchi 2014 reported that GnRH agonist had a protective effect on ovaries in respect of estradiol levels (SMD 1.35, 95% CI 0.67 to 2.02; 42 participants; P < 0.0001; Analysis 1.12). Gilani 2007 found the same conclusion (RR 3.44, 95% CI 1.57 to 7.58; 30 participants; P = 0.002; Analysis 1.13). Five other studies found no difference in estradiol levels between groups (Badawy 2009; Elgindy 2013; Gerber 2011; Lambertini 2015a; Song 2013).
1.12. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 12: Estradiol
1.13. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 13: Estradiol > 20 pg/mL
Survival rate
Lambertini 2015a reported that five‐year disease‐free survival was 80.5% (95% CI 77.2% to 90.5%), in chemotherapy plus GnRH agonist group and 83.7% (95% CI 76.1% to 89.1%), in chemotherapy alone group. Moore 2015 reported that the four‐year Kaplan‐Meier estimate of the rate of disease‐free survival was 89% in the GnRH agonist group and 78% in the control group (P = 0.04), while the four‐year Kaplan‐Meier estimate of the rate of overall survival was 92% and 82% respectively (P = 0.05).
Adverse effects
Based on results from four included studies (Gerber 2011; Lambertini 2015a; Moore 2015; Song 2013), the incidence of hot flush was 172 of 369 (46.6%) in the GnRH agonist plus chemotherapy group and 141 of 362 (39.0%) in the chemotherapy alone group. The RR was 1.49 (95% CI 0.16 to 13.60; 731 participants; I2 = 99%; Analysis 1.14), with no difference between groups (P = 0.73) . We performed sensitivity analysis for this outcome as Gerber 2011 and Moore 2015 were at high risk of bias. When these two studies were excluded, the result was unchanged (RR 1.21, 95% CI 0.17 to 8.72; 2 studies, 457 participants; P = 0.85; I2 = 98%). We downgraded the evidence for this outcome to very low certainty for risk of bias, inconsistency and imprecision (Table 1).
1.14. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 14: Adverse effects: hot flush
Based on results from two included studies (Lambertini 2015a; Moore 2015), the incidence of vaginal dryness was 26 of 250 (10.4%) in the GnRH agonist plus chemotherapy group and 21 of 238 (8.8%) in the chemotherapy alone group. The RR was 1.18 (95% CI 0.68 to 2.04; 488 participants; I2 = 0%; Analysis 1.15), with no difference between groups (P = 0.55). We performed sensitivity analysis for this outcome excluding one included study with high risk of bias (Moore 2015). Only one trial remained and the result was the same (RR 1.01, 95% CI 0.48 to 2.10; 1 study, 274 participants; P = 0.98). We downgraded the evidence for this outcome to moderate certainty for risk of bias (Table 1).
1.15. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 15: Adverse effect: vaginal dryness
Based on results from two included studies (Gerber 2011; Song 2013), the incidence of urogenital symptoms was 95 of 119 (79.8%) in the GnRH agonist plus chemotherapy group and 95 of 124 (76.6%) in the chemotherapy alone group. The RR was 2.37 (95% CI 0.01 to 448.16; 2 studies, 243 participants; I2 = 96%; Analysis 1.16), with no difference between groups (P = 0.75). We performed sensitivity analysis for this outcome excluding one included study with high risk of bias (Gerber 2011). Only one trial remained and the result was the same (RR 1.00, 95% CI 0.98 to 1.02; 1 study, 184 participants; P = 1.00).
1.16. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 16: Adverse effect: urogenital symptoms
Based on results from two included studies (Lambertini 2015a; Moore 2015), the incidence of sweating was 31 of 250 (12.4%) in the GnRH agonist plus chemotherapy group and 18 of 238 (7.6%) in the chemotherapy alone group. The RR was 1.65 (95% CI 0.93 to 2.80; 2 studies, 488 participants; I2 = 0%; Analysis 1.17), with no difference between groups (P = 0.09). We performed sensitivity analysis for this outcome excluding one included study with high risk of bias (Moore 2015). Only one trial remained and the result was the same (RR 1.65, 95% CI 0.83 to 3.29; 1 study, 274 participants; P = 0.16).
1.17. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 17: Adverse effect: sweating
Based on results from two included studies (Lambertini 2015a; Moore 2015), the incidence of headache was 40 of 250 (16.0%) in the GnRH agonist plus chemotherapy group and 20 of 238 (8.4%) in the chemotherapy alone group. The RR was 2.54 (95% CI 0.54 to 11.92; 2 studies, 488 participants; I2 = 75%; Analysis 1.18), with no difference between groups (P = 0.24). We performed sensitivity analysis for this outcome excluding one study with high risk of bias (Moore 2015). Only one trial remained and the result was the same (RR 1.34, 95% CI 0.78 to 2.31; 1 study, 274 participants; P = 0.29).
1.18. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 18: Adverse effect: headache
Based on results from three included studies (Gerber 2011; Lambertini 2015a; Song 2013), the incidence of mood swings was 107 of 266 (40.2%) in the GnRH agonist plus chemotherapy group and 111 of 251 (44.2%) in the chemotherapy alone group. The RR was 1.00 (95% CI 0.98 to 1.02; 3 studies, 517 participants; I2 = 0%; Analysis 1.19), with no difference between groups (P = 0.99). We performed sensitivity analysis for this outcome excluding one study with high risk of bias (Gerber 2011). Two studies remained and the result was the same (RR 0.99, 95% CI 0.81 to 1.22; 2 studies, 214 participants; P = 0.95).
1.19. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 19: Adverse effect: mood swings
Meta‐analyses also found no difference between chemotherapy plus GnRH agonist groups and chemotherapy alone groups in rates of fatigue (RR 2.16, 95% CI 0.20 to 23.42; 1 study, 214 participants; P = 0.53; Analysis 1.20Moore 2015), joint pain (RR 0.22, 95% CI 0.01 to 4.43; 1 study, 214 participants; P = 0.32; Analysis 1.21Moore 2015), muscle pain (RR 0.54, 95% CI 0.05 to 5.85; 1 study, 214 participants; P = 0.61; Analysis 1.22Moore 2015), decrease in lipid (RR 1.62, 95% CI 0.60 to 4.38; 1 study, 214 participants; P = 0.35; Analysis 1.23; Moore 2015), thromboembolism (RR 3.23, 95% CI 0.13 to 78.43; 1 study, 214 participants; P = 0.47; Analysis 1.24; Moore 2015), agitation (RR 1.29, 95% CI 0.41 to 4.11; 1 study, 214 participants; P = 0.44; Analysis 1.25; Moore 2015), anxiety (RR 2.42, 95% CI 0.77 to 7.63; 1 study, 214 participants; P = 0.13; Analysis 1.26Moore 2015), depression (RR 3.23, 95% CI 0.90 to 11.61; 1 study, 214 participants; P = 0.07; Analysis 1.27Moore 2015), and insomnia (RR 5.00, 95 % CI 0.62 to 40.28; 1 study, 60 participants; P = 0.13; Analysis 1.28Gerber 2011).
1.20. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 20: Adverse effect: fatigue
1.21. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 21: Adverse effect: joint pain
1.22. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 22: Adverse effect: muscle pain
1.23. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 23: Adverse effect: decrease in lipid
1.24. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 24: Adverse effect: thromboembolism
1.25. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 25: Adverse effect: agitation
1.26. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 26: Adverse effect: anxiety
1.27. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 27: Adverse effect: depression
1.28. Analysis.
Comparison 1: GnRH agonist plus chemotherapy versus chemotherapy alone, Outcome 28: Adverse effect: insomnia
Comparison 2: GnRH agonist‐antagonist cotreatment plus chemotherapy versus chemotherapy alone
Only one included RCT discussed GnRH agonist‐antagonist cotreatment (Elgindy 2013).
Primary outcome
Menstruation recovery or maintenance
The limited evidence from Elgindy 2013 showed the incidence of menstruation recovery or maintenance was 20 of 25 (80%) in both the GnRH agonist plus chemotherapy group and the chemotherapy alone group during the 12‐month follow‐up period. The RR was 1.00 (95% CI 0.76 to 1.32; 1 study, 50 participants; Analysis 2.1), with no difference between groups (P = 1.00). In the GnRH agonist plus chemotherapy group, 13 of 25 participants had menstruation recovery or maintenance (52.0%) in comparison to the chemotherapy alone group, in which 14 of 25 participants had menstruation recovery or maintenance (56.0%) during follow‐up period longer than 12 months. The RR was 0.93 (95% CI 0.56 to 1.55; 1 study, 50 participants; Analysis 2.1) with no difference between groups (P = 0.78). We downgraded the evidence for this outcome to very low certainty for risk of bias and serious imprecision (Table 2).
2.1. Analysis.
Comparison 2: Agonist‐antagonist cotreatment plus chemotherapy versus chemotherapy alone, Outcome 1: Menstruation recovery or maintenance
Pregnancy
The limited evidence from Elgindy 2013 showed the incidence of pregnancy rate was 1 of 25 (4.0%) in the GnRH agonist plus chemotherapy group and 0 of 25 (0%) in the chemotherapy alone group. The RR was 3.00 (95% CI 0.13 to 70.30; 1 study, 50 participants; Analysis 2.2), with no difference between groups (P = 0.49). We downgraded the evidence for this outcome to very low certainty for risk of bias and serious imprecision (Table 2).
2.2. Analysis.
Comparison 2: Agonist‐antagonist cotreatment plus chemotherapy versus chemotherapy alone, Outcome 2: Pregnancy
Discussion
Summary of main results
The review included 12 RCTs involving 1369 participants to evaluate the effect of GnRH analogues on the prevention of chemotherapy‐induced premature ovarian failure in premenopausal women. The participants were diagnosed with breast malignancy, ovarian malignancy, or Hodgkin's lymphoma. Most women of included studies received alkylating‐ or platinum complexes‐based chemotherapy.
Protection of GnRH agonist against chemotherapy‐induced premature ovarian failure
Pooled results showed that the incidence of menstruation recovery or maintenance in chemotherapy plus GnRH agonist groups was higher than that in chemotherapy alone groups during the 12‐month follow‐up period, but we observed no difference during follow‐up periods longer than 12 months (low‐certainty evidence; Table 1). Treatment‐related premature ovarian failure, which was defined as no resumption of menstrual activity with postmenopausal FSH levels at least one year after the completion of chemotherapy, occurred less often when GnRH agonist was given with chemotherapy compared to chemotherapy alone (moderate‐certainty evidence; Table 1). Evidence on the effects of GnRH agonist in the age subgroup was little and inconsistent.
GnRH agonist had a protective effect on ovulation rates (low‐certainty evidence; Table 1), however, there was no difference in pregnancy rates between chemotherapy plus GnRH agonist groups and chemotherapy groups (low‐certainty evidence; Table 1). We are cautious about this conclusion because there were insufficient data about whether the participants intended to become pregnant.
Although most included studies discussed hormone levels, most results were not presented in a form that we could meta‐analyse and the results were discrepant. Evidence of ultrasound antral follicle count and survival rate was very sparse overall, and often not presented in a form that we could meta‐analyse, making interpretation difficult.
Most common adverse effects of GnRH analogues include vasomotor symptoms, hot flushes, vaginal dryness, urogenital symptoms, headaches, sleep disturbance, sweating, depression and mood swings. The pooled analyses of safety data showed no difference in adverse effects between GnRH agonist groups and control groups. The certainty of evidence ranged from very low to moderate (Table 1).
Protection of GnRH agonist‐antagonist cotreatment against chemotherapy‐induced premature ovarian failure
A combination of GnRH antagonist and GnRH agonist appears to be useful because it combines the quick onset of action of GnRH antagonist with the long‐lasting effects of GnRH agonist. However, we only included one RCT involving GnRH agonist‐antagonist cotreatment (Elgindy 2013). The limited evidence from this study showed cotreatment had no protective effect on the ovaries in respect of menstruation recovery or maintenance, regardless of the follow‐up periods. There was also no difference in pregnancy rates between groups (Table 2).
Overall completeness and applicability of evidence
Despite most included studies being at unclear risk of bias, we found evidence to support a conclusion that the GnRH agonist was effective in protecting ovarian function during chemotherapy. Incidence of treatment‐related premature ovarian failure was lower and ovulation was more prevalent in GnRH agonist plus chemotherapy groups than that in chemotherapy‐alone groups.
Elgindy 2013 found that some women who menstruated in the 12‐month follow‐up period developed amenorrhoea later in both chemotherapy plus GnRH agonist groups and chemotherapy alone groups; while Lambertini 2015a found substantial restoration of menstruation after chemotherapy‐related amenorrhoea in long‐term follow‐up. Researchers believed that regular menses shortly after chemotherapy completion does not rule out the development of premature ovarian failure at a later date, and restoration of menstruation after chemotherapy‐related amenorrhoea is also possible (Lutchman‐Singh 2005). Thus, the protective effect of GnRH agonist on menstruation is complex and needs more research.
Most oncologists recommend waiting two to five years after concluding cancer treatments before trying to achieve pregnancy because the majority of cancers recur during this time frame and due to concerns that exposure to chemotherapy or radiotherapy (or both) may have genetically damaged maturing ovaries. Pooled results in this review demonstrated no difference in pregnancy rates between groups. However, these studies did not give details on the length of time after the end of chemotherapy before pregnancy was achieved, nor the use of contraception.
FSH and LH levels are commonly used hormone markers for premature ovarian failure. There were discrepancies in the results relating to FSH/LH levels. Five included studies reported protective effects by GnRH agonist on ovaries in respect of FSH/LH levels (Badawy 2009; Gerber 2011; Gilani 2007; Karimi‐Zarchi 2014; Song 2013), whereas the other four studies (Elgindy 2013; Giuseppe 2007; Lambertini 2015a; Munster 2012), found that there was no difference in FSH/LH levels between GnRH agonist‐chemotherapy groups and chemotherapy‐alone groups. Anti‐Mullerian hormone (AMH) is also a valid and valuable marker of ovarian follicle reserve. Two studies found that there was no difference regarding AMH between groups. Inhibin A is secreted by dominant follicles and corpus lutea, while inhibin B is secreted by developing pre‐antral and small antral follicles. Low levels of inhibin are typical in postmenopausal women (De Koning 2000; Luborsky 2000; Shelling 2000). Levels of inhibin may be an alternative indicator for premature ovarian failure than the currently used FSH/LH levels, for which there are cyclical changes and inter‐patient variability. However, only one study provided data regarding inhibin B levels and found that there was no difference between groups (Giuseppe 2007). Other commonly measured hormone markers include estradiol and progesterone. Five out of seven studies reported no differences in estradiol levels between groups (Badawy 2009; Elgindy 2013; Gerber 2011; Lambertini 2015a; Song 2013). On the whole, most data were incompletely reported and we could not meta‐analyse them. The limited opportunities for meta‐analysis might give an incomplete picture of the effects of GnRH agonist on hormone levels.
Ovaries become more sensitive to chemotherapeutic agents in older women (Hickman 2018; Leonard 2017). Chemotherapy‐induced ovarian damage may be drastically different in women under the age of 35 than in those between 35 and 40 and over 40. We found very little evidence on the effects of GnRH for age subgroup. Even where data were available, it was often incompletely reported and the results were totally different (Karimi‐Zarchi 2014; Leonard 2017; Song 2013). Leonard 2017 found a protective effect of goserelin on both the rate of amenorrhoea alone and on premature ovarian failure in women aged 40 or older; while the effect was less clear in women over 40 years. However, Song 2013 reported no difference between different age subgroups (women ≤ 35 years and younger and women older than 35 years) regarding premature ovarian failure between chemotherapy plus GnRH‐agonist groups and chemotherapy alone groups. Karimi‐Zarchi 2014 found that GnRH agonist had a protective effect in woman over 35 years instead of woman under 35 years. The limited opportunities for meta‐analysis made it difficult to know the effect of GnRH agonist on different age subgroups.
Chemotherapeutic agents are often used in combination and therefore it is difficult to evaluate the risk of ovarian damage of each individual agent. Some reports concluded that alkylating agents, cisplatin and adriamycin are among the most toxic agents (Blumenfeld 1999; Falcone 2005). Almost all recruited women in this review underwent treatment with chemotherapeutic agents that are considered to be the most toxic towards ovarian function, treatment durations were usually above six courses, and the administration of agents were usually via the intravenous route. However, it was not possible to get an understanding of the relationship between effects of GnRH analogues and treatment regimen for cumulative dose, dose‐intensity, duration and the route of administration in this review.
Four included studies reported data regarding adverse effects of GnRH agonist injection, such as hot flushes, vaginal dryness, urogenital symptoms, headaches, sleep disturbance, sweating, depression and mood swings (Gerber 2011; Lambertini 2015a; Moore 2015; Song 2013). We did not detect any difference in respect of adverse effects between groups. Since GnRH analogues could decrease utero‐ovarian perfusion and exposure of ovaries to chemotherapy, administration of GnRH analogues before or in parallel to chemotherapy may affect anti‐tumour effects of chemotherapy, especially for uterine or ovarian malignancies. However, evidence on survival rate was very sparse, and often not presented in a form that could be meta‐analysed, making interpretation difficult. Lambertini 2015a reported five‐year disease‐free survival and Moore 2015 reported the four‐year Kaplan‐Meier estimate of the rate of disease‐free survival. Those two studies included participants with breast cancer. Neither studies found any difference between groups.
Quality of the evidence
We rated the certainty of the evidence and have presented our findings in Table 1 and Table 2 for key outcomes. The review included 12 RCTs involving 1369 participants.
Most included studies were single‐centre trials with small numbers of participants. We made explicit judgements about whether these studies were at high risk of bias, according to the criteria given in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2017). We judged seven included RCTs as unclear risk of bias due to one or more criteria assessed being unclear (Badawy 2009; Gilani 2007; Giuseppe 2007; Karimi‐Zarchi 2014; Lambertini 2015a; Leonard 2017; Song 2013), while another five were at high risk of bias due to one or more criteria assessed being high (Elgindy 2013; Gerber 2011; Moore 2015; Munster 2012; Waxman 1987). Random sequence generation, allocation concealment and blinding in most studies were unclear. Although most studies did not indicate whether blinding was implemented, the impact on primary outcomes, including menstruation recovery or maintenance, premature ovarian failure and pregnancy, were limited, as the primary outcomes were objective indicators.
For comparisons between GnRH agonist plus chemotherapy versus chemotherapy alone, we judged the certainty of evidence on premature ovarian failure as moderate due to risk of bias caused by unclear study methodology or study design limitations. Evidence that we rated as low certainty included evidence on menstruation recovery or maintenance, pregnancy, and ovulation. Evidence was often undermined by risk of bias and inconsistency (I2 > 50%). Additional reasons for downgrading the certainty of evidence included imprecision (small sample size) and indirectness (the population should be restricted to women who intended to become pregnant for the outcome of pregnancy rate). Hormone levels as secondary outcomes were reported in almost all studies; however, most data were incompletely reported and could not be meta‐analysed. The limited opportunities for meta‐analysis might give an incomplete picture of the effects of GnRH agonist on hormone levels. Ultrasound antral follicle count and survival rate were seldom reported. The evidence was generally of moderate to very low certainty for adverse effects, and often undermined by unclear study methodology or study design limitations, inconsistency, and imprecision (wide confidence interval crossing the line of no effect).
Only one included RCT, with a small sample size, compared GnRH agonist‐antagonist cotreatment plus chemotherapy versus chemotherapy alone. Thus, the evidence was very low certainty due to concern about risk of bias and serious imprecision (single RCT with small sample size).
Potential biases in the review process
We prepared this review using Cochrane methodology, and were guided by both the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011b), and the standard methods of Cochrane Gynaecological, Neuro‐Oncology and Orphan Cancers. We strictly followed the review protocol in study selection, data extraction and data analysis. Two review authors independently performed study selection, data extraction and assessment of risk of bias. We used standardized data extraction forms.
However, this review had limitations. Some important information from studies was unclear and this could result in bias. The included studies differed in some important aspects of design, namely the definition of menstruation recovery or maintenance and premature ovarian failure, and time to follow‐up, which could lead to biases relating to information and misclassification. We tried to contact the authors of included studies for additional or missing data and individual patient data, but no one replied. Thus, we extracted all data from published articles. We used the random‐effects model for analyses. The limitations of this approach were down‐weighting large studies when statistical heterogeneity was present, and giving more equal weighting to the studies combined. Women recruited in this review were between the ages of 12 and 51.1 years; however, most studies had no age‐determined subgroup analysis. This could also introduce bias due to the fact that chemotherapy‐induced ovarian damage may be drastically different in women of different ages. Moreover, the default zero‐cell correction could potentially bias the result of the meta‐analysis towards no difference between the two groups, as the RR was calculated when there were no events (pregnancy) in GnRH group.
There are five studies awaiting classification because we were unable to obtain the full texts. This could generate information bias and publication bias within the methodology and processes of the systematic review.
Agreements and disagreements with other studies or reviews
Several recently published systematic reviews and meta‐analyses have discussed the protective effects of GnRH agonist on ovaries. Hickman 2018 and Lambertini 2015b found that GnRH agonists had a protective effect against chemotherapy‐related premature ovarian failure. Although this conclusion was consistent with ours, one major problem with these two reviews was that the review authors simply classified un‐recovered menstruation during the follow‐up period as premature ovarian failure and analysed real premature ovarian failure and amenorrhoea together. Munhoz 2016 found that GnRH agonist given with chemotherapy was associated with increased rates of menstruation recovery after a minimum of six months after the last cycle of chemotherapy. This result was partially consistent with ours, which demonstrated that menstruation recovery or maintenance after chemotherapy plus GnRH agonist was higher than for chemotherapy alone during a 12‐month follow‐up period, but no difference was observed during follow‐up longer than 12 months. Lambertini 2015b and Munhoz 2016 found that GnRH agonist seemed to increase the pregnancy rate, whereas our results demonstrated no difference in pregnancy rates between groups. Discrepancies might be due to differences in the studies and the included population (it included only women with breast cancer) and absence of subgroup analyses at different follow‐up periods.
Authors' conclusions
Implications for practice.
Gonadotropin‐releasing hormone (GnRH) agonist appears to be effective in protecting the ovaries during chemotherapy, in terms of menstruation recovery or maintenance, treatment‐related premature ovarian failure and ovulation. Evidence for protection of fertility was insufficient and needs further investigation. Evidence was also insufficient to assess the effect of GnRH agonist and GnRH antagonist cotreatment on ovarian protection against chemotherapy.
Implications for research.
Large and well‐designed randomised controlled trials should be conducted to clarify the effects of GnRH analogues in preventing chemotherapy‐induced ovarian failure, especially for the use of GnRH antagonists. We suggest that further studies on the protective effect of GnRH analogues supplementation on different age groups or different chemotherapy regimens should be conducted. Furthermore, studies should have longer follow‐up duration and address the effects on pregnancy rates, anti‐tumour therapy (including five‐year‐survival) and direct adverse effects (symptoms of hypoestrogenaemia such as hot flashes, headaches and osteoporosis).
What's new
| Date | Event | Description |
|---|---|---|
| 12 August 2021 | Review declared as stable | Conclusions unlikely to change with the addition of new studies. No update planned. |
History
Protocol first published: Issue 4, 2009 Review first published: Issue 11, 2011
| Date | Event | Description |
|---|---|---|
| 15 April 2019 | Amended | Amendment made to PLS for clarity. |
| 28 February 2019 | New search has been performed | New searches run in November 2018. Text of the review updated as required. Author team revised. |
| 28 February 2019 | New citation required but conclusions have not changed | Eight additional RCTs involving 1212 participants were included. Conclusions remain unchanged. |
| 1 April 2015 | Amended | Contact details updated. |
| 27 March 2014 | Amended | Contact details updated. |
Acknowledgements
The authors wish to thank the Cochrane Gynaecological, Neuro‐oncology and Orphan Cancers editorial team: Jo Morrison (Co‐ordinating Editor) for clinical and editorial advice, Clare Jess (Managing Editor) and Tracey Harrison (Assistant Managing Editor) for their editorial support, and Jo Platt (Information Specialist) for her assistance with the search.
This project was supported by the National Institute for Health Research, via Cochrane Infrastructure funding to the Cochrane Gynaecological, Neuro‐oncology and Orphan Cancer Group. The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the Systematic Reviews Programme, NIHR, NHS or the Department of Health.
Appendices
Appendix 1. CENTRAL search strategy
#1 MeSH descriptor Ovarian Failure, Premature explode all trees #2 ovar* and (failure or function or damage) #3 MeSH descriptor Fertility explode all trees #4 fertility #5 (#1 OR #2 OR #3 OR #4) #6 chemotherap* #7 gonadal toxicity or gonadotoxic* #8 MeSH descriptor Antineoplastic Agents explode all trees #9 MeSH descriptor Antineoplastic Combined Chemotherapy Protocols explode all trees #10 (#6 OR #7 OR #8 OR #9) #11 MeSH descriptor Gonadotropin‐Releasing Hormone explode all trees #12 GnRH #13 gonadotropin‐releasing hormone #14 LHRH #15 luteinizing‐hormone releasing hormone #16 (#11 OR #12 OR #13 OR #14 OR #15) #17 (#5 AND #10 AND #16)
Appendix 2. MEDLINE OVID search strategy
1 exp Ovarian Failure, Premature/ 2 (ovar* and (failure or function or damage)).mp. 3 exp Fertility/ 4 fertility.mp. 5 1 or 2 or 3 or 4 6 chemotherap*.mp. 7 (gonadal toxicity or gonadotoxic*).mp. 8 exp Antineoplastic Agents/ 9 exp Antineoplastic Combined Chemotherapy Protocols/ 10 6 or 7 or 8 or 9 11 exp Gonadotropin‐Releasing Hormone/ 12 GnRH.mp. 13 gonadotropin‐releasing hormone.mp. 14 LHRH.mp. 15 luteinizing‐hormone releasing hormone.mp. 16 11 or 12 or 13 or 14 or 15 17 5 and 10 and 16 18 "randomized controlled trial".pt. 19 "controlled clinical trial".pt. 20 randomized.ab. 21 placebo.ab. 22 drug therapy.fs. 23 randomly.ab. 24 trial.ab. 25 groups.ab. 26 18 or 19 or 20 or 21 or 22 or 23 or 24 or 25 27 17 and 26 28 Animals/ 29 Humans/ 30 28 not (28 and 29) 31 27 not 30
Key: mp=title, original title, abstract, name of substance word, subject heading word ab=abstract fs=floating subheading
Appendix 3. Embase OVID search strategy
1 exp premature ovarian failure/ 2 (ovar* and (failure or function or damage)).mp. 3 exp fertility/ 4 fertility.mp. 5 1 or 2 or 3 or 4 6 exp chemotherapy/ or chemotherap*.mp. 7 (gonadal toxicity or gonadotoxic*).mp. 8 exp antineoplastic agent/ 9 6 or 7 or 8 10 exp gonadorelin/ 11 GnRH.mp. 12 gonadotropin‐releasing hormone.mp. 13 LHRH.mp. 14 luteinizing‐hormone releasing hormone.mp. 15 10 or 11 or 12 or 13 or 14 16 5 and 9 and 15 17 exp controlled clinical trial/ 18 randomized.ab. 19 placebo.ab. 20 dt.fs. 21 randomly.ab. 22 trial.ab. 23 groups.ab. 24 17 or 18 or 19 or 20 or 21 or 22 or 23 25 16 and 24
key: mp=title, abstract, subject headings, heading word, drug trade name, original title, device manufacturer, drug manufacturer name ab=abstract fs=floating subheading
Appendix 4. CBM search strategy
#1. 卵巢早衰 ti. #2. 卵巢早衰 tx. #3. 卵巢早衰 ab. #4. 化疗 ti. #5. 化疗 tx. #6. 化疗 ab. #7. 促性腺激素释放激素 ti. #8. 促性腺激素释放激素 tx. #9. 促性腺激素释放激素 ab. #10. OR 1‐3 #11 OR 4‐5 #12 OR 6‐9 #13 10 AND 11 AND 12
Appendix 5. Data extraction form items
Review ID
Study ID
Reference ID
Person extracting data
Date of date extraction
Year of study publication
Title
Author
Type of study design
Unit of randomisation
Describe setting
Funding source
Inclusion criteria
Exclusion criteria
Experimental intervention
Control/comparison intervention
Outcomes
Adequate sequence generation: was the allocation sequence adequately generated?
Allocation concealment: was allocation concealment adequate?
Blinding: was knowledge of the allocated intervention adequately prevented during the study?
Incomplete outcome data addressed: were complete outcome data adequately addressed?
Free of selective reporting bias: are reports of study free of suggestions of selective reporting bias?
Free of other bias: was the study apparently free of other problems that could put it at high risk of bias?
Outcome measures
Very brief summary of study authors' main findings/conclusions
Reasons for exclusion
Appendix 6. Risk of bias criteria
Generation of allocation sequence
Yes ‐ adequate methods applied e.g. random number tables, computer‐generated random numbers, coin tossing, or card shuffling
Unclear ‐ no information was provided
No ‐ other methods of allocation that are potentially biased
Allocation concealment
Yes ‐ adequate measures such as central randomisation; serially numbered, opaque, sealed envelopes; or another description that contained convincing elements of concealment
Unclear ‐ information unavailable
No ‐ inadequate measures that do not fall into one of the categories under the 'adequate methods' category
Blinding
Blinding of women (yes, no, or unclear)
Blinding of caregivers (yes, no, or unclear)
Blinding of outcome assessment (yes, no, or unclear)
Incomplete outcome data
Yes ‐ clearly described loss of participants to follow‐up at each data collection point and exclusion of participants after randomisation
Unclear ‐ no information was provided
No ‐ unreported/missing numbers of losses at follow‐up, or no reasons for missing data provided
Free of selective reporting bias
Yes ‐ all of the study's pre‐specified outcomes and all expected outcomes of interest to the review have been reported
Unclear ‐ information unavailable
No ‐ incomplete reporting of study's pre‐specified outcomes; one or more reported primary outcomes were not pre‐specified; outcomes of interest were reported incompletely and thus could not be used; study failed to include results of a key outcome that would have been expected to have been reported
Free of other bias
We described for each included study any important concerns we had about other possible sources of bias, for example, the baseline balance.
Yes ‐ no evidence of bias from other sources
Unclear ‐ insufficient information to permit judgement of adequacy or otherwise of other forms of bias
No ‐ potential bias present from other sources (e.g. early stopping of trial, fraudulent activity, extreme baseline imbalance, or bias related to specific study design)
Data and analyses
Comparison 1. GnRH agonist plus chemotherapy versus chemotherapy alone.
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
|---|---|---|---|---|
| 1.1 Menstruation recovery or maintenance | 11 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 1.1.1 Follow‐up time ≤ 12 months | 5 | 460 | Risk Ratio (M‐H, Random, 95% CI) | 1.60 [1.14, 2.24] |
| 1.1.2 Follow‐up time > 12 months | 8 | 869 | Risk Ratio (M‐H, Random, 95% CI) | 1.08 [0.95, 1.22] |
| 1.2 Premature ovarian failure (POF) | 4 | 780 | Risk Ratio (M‐H, Random, 95% CI) | 0.44 [0.31, 0.61] |
| 1.3 Pregnancy | 7 | 703 | Risk Ratio (M‐H, Random, 95% CI) | 1.59 [0.93, 2.70] |
| 1.4 Ovulation | 2 | 95 | Risk Ratio (M‐H, Random, 95% CI) | 2.47 [1.43, 4.26] |
| 1.5 Antral follicle count | 1 | Std. Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.6 FSH (mUI/L) | 2 | 71 | Std. Mean Difference (IV, Random, 95% CI) | 0.26 [‐0.80, 1.31] |
| 1.7 FSH < 20 mIU/mL | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 1.8 LH (mUI/L) | 2 | 71 | Std. Mean Difference (IV, Random, 95% CI) | ‐0.62 [‐1.28, 0.03] |
| 1.9 LH < 20 mIU/L | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 1.10 AMH (pmol/L) | 1 | Std. Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.11 Inhibin B | 1 | Std. Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.12 Estradiol | 1 | Std. Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.13 Estradiol > 20 pg/mL | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 1.14 Adverse effects: hot flush | 4 | 731 | Risk Ratio (M‐H, Random, 95% CI) | 1.49 [0.16, 13.60] |
| 1.15 Adverse effect: vaginal dryness | 2 | 488 | Risk Ratio (M‐H, Random, 95% CI) | 1.18 [0.68, 2.04] |
| 1.16 Adverse effect: urogenital symptoms | 2 | 243 | Risk Ratio (M‐H, Random, 95% CI) | 2.37 [0.01, 448.16] |
| 1.17 Adverse effect: sweating | 2 | 488 | Risk Ratio (M‐H, Random, 95% CI) | 1.61 [0.93, 2.80] |
| 1.18 Adverse effect: headache | 2 | 488 | Risk Ratio (M‐H, Random, 95% CI) | 2.54 [0.54, 11.92] |
| 1.19 Adverse effect: mood swings | 3 | 517 | Risk Ratio (M‐H, Random, 95% CI) | 1.00 [0.98, 1.02] |
| 1.20 Adverse effect: fatigue | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 1.21 Adverse effect: joint pain | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 1.22 Adverse effect: muscle pain | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 1.23 Adverse effect: decrease in lipid | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 1.24 Adverse effect: thromboembolism | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 1.25 Adverse effect: agitation | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 1.26 Adverse effect: anxiety | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 1.27 Adverse effect: depression | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected | |
| 1.28 Adverse effect: insomnia | 1 | Risk Ratio (M‐H, Random, 95% CI) | Totals not selected |
Comparison 2. Agonist‐antagonist cotreatment plus chemotherapy versus chemotherapy alone.
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
|---|---|---|---|---|
| 2.1 Menstruation recovery or maintenance | 1 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 2.1.1 Follow‐up time ≤ 12 months | 1 | 50 | Risk Ratio (M‐H, Random, 95% CI) | 1.00 [0.76, 1.32] |
| 2.1.2 Follow‐up time > 12 months | 1 | 50 | Risk Ratio (M‐H, Random, 95% CI) | 0.93 [0.56, 1.55] |
| 2.2 Pregnancy | 1 | 50 | Risk Ratio (M‐H, Random, 95% CI) | 3.00 [0.13, 70.30] |
Characteristics of studies
Characteristics of included studies [ordered by study ID]
Badawy 2009.
| Study characteristics | ||
| Methods |
Study design: single‐centre RCT Study duration: not mentioned |
|
| Participants |
Inclusion criteria: women diagnosed with unilateral adenocarcinoma of the breast, with positive or negative lymph node status, and with no metastasis who had undergone modified radical mastectomy or breast‐conserving surgery plus full axillary lymph node dissection. All women were menstruating normally and aged 18 to 40 years (basal FSH levels < 10 mU/mL)
Exclusion criteria: women with macroscopic metastatic spread of the disease |
|
| Interventions |
Treatment group
Control group
|
|
| Outcomes | Women were followed up monthly for up to 8 months after completion of treatment:
|
|
| Notes |
|
|
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | Not mentioned |
| Allocation concealment (selection bias) | Low risk | Quote: "sealed, dark envelopes" were used |
| Blinding (performance bias and detection bias) All outcomes | Unclear risk | Not mentioned whether participant/personnel/outcome assessment was blinded or not |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | Only 1 woman in each group dropped out. Reasons not mentioned |
| Selective reporting (reporting bias) | Low risk | Results of all outcomes mentioned in Methods section were reported in Results section |
| Other bias | Unclear risk | Protocol was not mentioned. There was no difference between groups regarding age, body weight, height, number of dissected lymph nodes, serum LH, and chemotherapy dose |
Elgindy 2013.
| Study characteristics | ||
| Methods |
Study design: 2‐centre, 4‐armed, open‐label RCT Study duration: December 2009 to August 2011 |
|
| Participants |
Inclusion criteria: women (18 to 40 years old) with primary hormone insensitive breast cancer (stage I–IIIa) scheduled for cyclophosphamide‐based chemotherapy; and with a history of regular menstrual periods, transvaginal ultrasound‐confirmed presence of both ovaries, and absence of ovarian tumours or cysts > 40 mm
Exclusion criteria: women with advanced breast cancer (stage IIIb–IV), primary ovarian cancer or pelvic metastases, history of chemotherapy or abdominal or pelvic radiation; women receiving or planning to receive hormone therapy; and pregnant and nursing women |
|
| Interventions |
Treatment group
Control group
|
|
| Outcomes | Women were followed up on a 6‐month basis for 18 months
|
|
| Notes |
|
|
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Quote: "computer‐generated block randomisation scheme with variable block sizes" was used |
| Allocation concealment (selection bias) | Low risk | Quote: "The randomisation list was produced by a statistician not involved with patient recruitment. Allocation concealment was achieved with sequentially numbered, dark, opaque, sealed envelopes" |
| Blinding (performance bias and detection bias) All outcomes | High risk | Quote: It's an "open‐label trial" |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | Five women died during the follow‐up period: one woman in each group except the GnRH agonist plus delayed chemotherapy group, in which two women died |
| Selective reporting (reporting bias) | High risk | Adverse events as outcomes were not reported |
| Other bias | Unclear risk | Protocol was not mentioned. There was no significant difference in baseline characteristics between groups |
Gerber 2011.
| Study characteristics | ||
| Methods |
Study design: randomised, open‐label, controlled, multi‐centre phase II study Study duration: March 2005 to December 2007 |
|
| Participants |
Inclusion criteria: premenopausal women (aged 18 to 45 years) with primary hormone‐insensitive breast cancer. They had to have had regular and spontaneous menstrual periods before study entry, with FSH < 15 mlU/mL in the follicular phase of the menstrual cycle. Women had to use adequate non‐hormonal contraceptive measures during study treatment. Treatment with sex hormones was not allowed
Exclusion criteria: women were excluded for known hypersensitivity reaction to the investigational compounds, prior cytotoxic treatment for any reason, and distant metastases and if (primary or secondary) ovarian insufficiency was suspected |
|
| Interventions |
Treatment group
Control group
|
|
| Outcomes | Women were followed up on a 6‐month basis for at least 24 months
|
|
| Notes |
|
|
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | Not mentioned |
| Allocation concealment (selection bias) | Unclear risk | Not mentioned |
| Blinding (performance bias and detection bias) All outcomes | High risk | Quote: it's an "open‐label trial" |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | The risk of bias in the primary outcomes is low as only two women in each group withdrew due to adverse event, progression of disease and unwillingness to participate. Antral follicle count was only measured in 10 women, thus the risk of bias for this secondary outcome was high |
| Selective reporting (reporting bias) | Low risk | Results of all outcomes mentioned in protocol were reported |
| Other bias | Unclear risk | There was no significant difference in baseline characteristics between groups |
Gilani 2007.
| Study characteristics | ||
| Methods |
Study design: RCT Study duration: not mentioned |
|
| Participants |
Inclusion criteria: post‐menarchal women aged 12 to 40 years with ovarian malignancy; history of normal menstrual rhythm; women who had 1 or 2 ovaries and normal laboratory tests after surgery; alkylating or alkylating‐like metronomic chemotherapy; different histological ovarian tumours and any FIGO stage
Exclusion criteria: < 12 or > 40 years; had received radiotherapy; had received > 6 months' chemotherapy, single agent chemotherapy or non‐alkylating or alkylating‐like metronomic chemotherapy |
|
| Interventions |
Treatment group
Control group
|
|
| Outcomes | Women were followed up for 6 months after chemotherapy
|
|
| Notes |
|
|
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | Not mentioned |
| Allocation concealment (selection bias) | Unclear risk | Not mentioned |
| Blinding (performance bias and detection bias) All outcomes | Unclear risk | Not mentioned whether participant/personnel/outcome assessment was blinded or not |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | All women were followed up to the end of the trial |
| Selective reporting (reporting bias) | Low risk | Results of all outcomes mentioned in Methods section were reported in Results section |
| Other bias | Unclear risk | Protocol was not mentioned. There was no significant difference in baseline characteristics between groups |
Giuseppe 2007.
| Study characteristics | ||
| Methods |
Study design: single‐centre, RCT Study duration: 1996 to 2002 |
|
| Participants |
Inclusion criteria: female cancer survivors previously affected by Hodgkin's lymphoma, aged 20 to 38 years, treated with the same standardized chemotherapeutic schedule
Exclusion criteria: not mentioned |
|
| Interventions |
Treatment group
Control group
|
|
| Outcomes | Women were followed up every 6 months after cessation of chemotherapy (mean follow‐up time was 4.2 ± 2.8 years)
|
|
| Notes |
|
|
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | Not mentioned |
| Allocation concealment (selection bias) | Unclear risk | Not mentioned |
| Blinding (performance bias and detection bias) All outcomes | Unclear risk | Not mentioned whether participant/personnel/outcome assessment was blinded or not |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | All women were followed up and included in the analyses |
| Selective reporting (reporting bias) | Low risk | Results of all outcomes mentioned in Methods section were reported in Results section |
| Other bias | Unclear risk | Protocol was not mentioned. There was no significant difference in baseline characteristics between groups |
Karimi‐Zarchi 2014.
| Study characteristics | ||
| Methods |
Study design: RCT Study duration: 2010 to 2011 |
|
| Participants |
Inclusion criteria: women with breast cancer aged 25 to 45 years; oestrogen receptor‐negative and progesterone receptor‐negative breast cancer; primary Stages to Stage II (T2N/M0); candidate for chemotherapy
Exclusion criteria: advanced stages of breast cancer; low complaint cases in receiving diphereline monthly; oestrogen receptor‐positive and progesterone receptor‐positive breast cancer; concurrent malignancies; FSH ≥ 30 IU/L |
|
| Interventions |
Treatment group
Control group
|
|
| Outcomes | Women were followed up at the end of months 3 and 6 after cessation of chemotherapy
|
|
| Notes |
|
|
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | Not mentioned |
| Allocation concealment (selection bias) | Unclear risk | Not mentioned |
| Blinding (performance bias and detection bias) All outcomes | Unclear risk | Not mentioned |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | All women were followed up |
| Selective reporting (reporting bias) | Low risk | Results of all outcomes mentioned in Methods section were reported in Results section |
| Other bias | Unclear risk | Protocol was not mentioned. There was no mention of baseline characteristics between groups |
Lambertini 2015a.
| Study characteristics | ||
| Methods |
Study design: multi‐centre, randomised phase 3 trial Study duration: October 2003 to January 2008 |
|
| Participants |
Inclusion criteria: histologically proven stage I, II, or III breast cancer; aged 18 to 45 years, premenopausal
Exclusion criteria: previous chemotherapy, radiotherapy, for both cancer or nonneoplastic diseases; evidence of distant metastases; other malignancies in the previous 5 years, except basal or squamous cell carcinoma of the skin or adequately treated in situ carcinoma of the cervix; and pregnancy or lactation |
|
| Interventions |
Treatment group
Control group
|
|
| Outcomes | Women were followed up every 12 months (the median long‐term follow‐up was 7.3 years)
|
|
| Notes |
|
|
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Quote: "Randomisation lists were prepared with the use of permuted blocks of different sizes and a 1:1 ratio" |
| Allocation concealment (selection bias) | Low risk | Quote: "Randomisation was performed centrally by faxing the Clinical Trials Unit of the National Cancer Research Institute in Genoa (Italy)" |
| Blinding (performance bias and detection bias) All outcomes | Unclear risk | Not mentioned whether participant/personnel/outcome assessment was blinded or not |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | In 12‐month follow‐up, 6 women in chemotherapy‐alone group did not receive chemotherapy and 6 were lost to follow‐up, while 2 women in chemotherapy plus triptorelin group did not receive chemotherapy and 7 were lost to follow‐up. At the study cut‐off date of long follow‐up period (median follow‐up was 7.3 years), 18 and 17 women were lost to follow‐up in chemotherapy‐alone group and chemotherapy plus triptorelin group respectively |
| Selective reporting (reporting bias) | Low risk | All outcomes mentioned in the Protocol were reported |
| Other bias | Low risk | No information specified. There was no significant difference in baseline characteristics between groups |
Leonard 2017.
| Study characteristics | ||
| Methods |
Study design: multi‐centre RCT Study duration: August 2004 to December 2009 |
|
| Participants |
Inclusion criteria: premenopausal women (defined as regular menses in the 12 months prior to chemotherapy) with histologically confirmed breast cancer who were to receive adjuvant or neo‐adjuvant chemotherapy; women with estradiol receptor‐negative tumours; women with estradiol receptor‐positive tumours for whom the investigator did not deem ovarian suppression necessary as part of the treatment; the breast cancers could be up to stage IIIB (T1‐T4 with N0‐2) and complete excision of the tumour before adjuvant chemotherapy or planned after neoadjuvant therapy
Exclusion criteria: women who had prior chemotherapy or endocrine therapy |
|
| Interventions |
Treatment group
Control group
|
|
| Outcomes | Women were followed up 6‐monthly for 24 months
|
|
| Notes |
|
|
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Quote: "Randomisation was centrally performed by telephone to the trial centre, eligibility was confirmed verbally, and treatment was allocated by computer‐generated lists" |
| Allocation concealment (selection bias) | Low risk | Quote: "Randomisation was centrally performed by telephone to the trial centre, eligibility was confirmed verbally, and treatment was allocated by computer‐generated lists" |
| Blinding (performance bias and detection bias) All outcomes | Unclear risk | Not mentioned whether participant/personnel/outcome assessment was blinded or not |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | Three in each arm had died within 24 months of randomisation. For a further 19 women (11 in the control arm and 8 in the intervention arm), menstrual status during the interval between the 12‐month follow‐up visit and the 24‐month follow‐up visit could not be determined from the data available |
| Selective reporting (reporting bias) | Low risk | All outcomes mentioned in Protocol were reported |
| Other bias | Low risk | The original protocol restricted the entry of women to those with estradiol receptor‐negative tumours only, but women with estradiol receptor‐positive tumours for whom the investigator did not deem ovarian suppression necessary as part of the treatment were subsequently allowed entry to the trial after a protocol amendment. There was no significant difference in baseline characteristics between groups |
Moore 2015.
| Study characteristics | ||
| Methods |
Study design: open‐label, international multi‐centre, randomised phase 3 trial Study duration: February 2004 to May 2011 |
|
| Participants |
Inclusion criteria: premenopausal women, 18 to 49 years of age were eligible if they had operable stage I‐IIIA estradiol receptor‐negative and progesterone receptor‐negative breast cancer. Eligible participants had taken no oestrogens, antioestrogens, selective estradiol receptor modulators, aromatase inhibitors, or hormonal contraceptives within the month before inclusion
Exclusion criteria: Exceptions were made for the use of hormonal contraception in women < 35 years of age that was discontinued before randomisation and for hormonal treatment for up to 2 months for the purposes of in vitro fertilization and cryopreservation of embryos or oocytes before randomisation |
|
| Interventions |
Treatment group
Control group
|
|
| Outcomes | Women were followed up annually (the median follow‐up time at the time of the end‐point analysis was 4.1 years)
|
|
| Notes |
|
|
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | Not mentioned |
| Allocation concealment (selection bias) | Unclear risk | Not mentioned |
| Blinding (performance bias and detection bias) All outcomes | High risk | Quote: it's an "open‐label trial" |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | Data on pregnancy and survival were collected in 218 women, and data on adverse events were collected in 214 women. Attrition mainly due to withdrawal, loss to follow‐up and non‐response. The attrition bias for these outcomes were at low risk. Data on POF were available for only 88 and 69 women in treatment group and control group respectively. Attrition was due to incomplete data on menstruation or FSH levels. The attrition bias for POF was at high risk |
| Selective reporting (reporting bias) | Low risk | All outcomes mentioned in Protocol were reported |
| Other bias | High risk | The study closed early owing to loss of funding for study‐drug distribution. ClinicalTrials.gov Identifier: NCT00196846 |
Munster 2012.
| Study characteristics | ||
| Methods |
Study design: RCT Study duration: July 2003 to January 2007 |
|
| Participants |
Inclusion criteria: premenopausal women (< 45 years of age with a FSH level < 40 mIU/mL and ≥ 2 menstrual periods in the preceding 6 months) with newly diagnosed early‐stage breast cancer (stages I‐III) and planned adjuvant or neoadjuvant chemotherapy
Exclusion criteria: pregnancy or lactation, prior chemotherapy, or bilateral oophorectomy or ovarian irradiation before enrolment; history of other cancers; personal or familial (first‐degree relative) premature ovarian failure; plan to undergo an oophorectomy/hysterectomy within 2 years; and oral contraceptive administration |
|
| Interventions |
Treatment group
Control group:
|
|
| Outcomes | Women were followed up for at least 2 years after the last cycle of chemotherapy (median follow‐up time was 18 months from completion of last chemotherapy)
|
|
| Notes |
|
|
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Quote: "computerized randomisation" was used |
| Allocation concealment (selection bias) | Unclear risk | Not mentioned |
| Blinding (performance bias and detection bias) All outcomes | Unclear risk | Not mentioned whether participant/personnel/outcome assessment was blinded or not |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | 1 participant in the triptorelin arm withdrew consent before first treatment. Another, in the control arm, was found to have residual disease after neoadjuvant chemotherapy and was treated with additional chemotherapy not permitted by protocol |
| Selective reporting (reporting bias) | High risk | Protocol was mentioned in the published article. Levels of serum inhibin A were not reported |
| Other bias | High risk | The study initially planned to enrol 124 women, but the trial was stopped for futility after 49 women were enrolled. The planned follow‐up time was 5 years but was stopped early because no difference between the 2 groups was seen |
Song 2013.
| Study characteristics | ||
| Methods |
Study design: single‐centre, prospective randomised phase II study Study duration: February 2010 to March 2012 |
|
| Participants |
Inclusion criteria: premenopausal women, 18 to 45 years; proven histological stage I–III breast cancer; received primary surgical therapy before adjuvant chemotherapy; no history of prior chemotherapy or hormone therapy; no distant metastases or localized cancer relapse; no prior or concomitant malignancy; no history of primary or secondary ovarian insufficiency; and no pregnancy or nursing
Exclusion criteria: not mentioned |
|
| Interventions |
Treatment group
Control group
|
|
| Outcomes | Women were followed up for 12 months after the end of chemotherapy
|
|
| Notes |
|
|
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | Not mentioned |
| Allocation concealment (selection bias) | Unclear risk | Not mentioned |
| Blinding (performance bias and detection bias) All outcomes | Unclear risk | Not mentioned whether participant/personnel/outcome assessment was blinded or not |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | Sixteen in chemotherapy‐alone group (2 women failed to complete at least one cycle of chemotherapy and 14 women were lost to follow‐up) and 21 in chemotherapy plus GnRH agonist group (4 women failed to complete at least one cycle of chemotherapy and 17 women were lost to follow‐up) were not included in the evaluation |
| Selective reporting (reporting bias) | Low risk | Results of all outcomes mentioned in Methods section were reported in Results section |
| Other bias | Unclear risk | There was no significant difference in baseline characteristics between groups |
Waxman 1987.
| Study characteristics | ||
| Methods |
Study design: RCT Study duration: not mentioned |
|
| Participants |
Inclusion criteria: women aged 17 to 46 years old with Hodgkin's disease
Exclusion criteria: not mentioned |
|
| Interventions |
Treatment group
Control group:
|
|
| Outcomes | Women were followed up for 3 years after completing treatment
|
|
| Notes |
|
|
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | Not mentioned |
| Allocation concealment (selection bias) | Unclear risk | Not mentioned |
| Blinding (performance bias and detection bias) All outcomes | Unclear risk | Not mentioned whether participant/personnel/outcome assessment was blinded or not |
| Incomplete outcome data (attrition bias) All outcomes | Low risk | One woman died during treatment |
| Selective reporting (reporting bias) | Low risk | Results of all outcomes mentioned in Methods section were reported in Results section |
| Other bias | High risk | Control groups included women who were > 45 years old while GnRH‐agonist group included women less than 34 years old. Besides, low dose of buserelin used in the study could have led to incomplete pituitary downregulation, which might have affected the results |
ABVD: Doxorubicin, Bleomycin, Vinblastine, Dacarbazine; AMH: anti‐Mullerian hormone; BEP: Bleomycin, Etoposide, Cisplatin; CMOPP: Cyclophosphamide, Mitoxantrone, Vincristine, Prednisone, Procarbazine; DHAP: Rituximab, Dexamethasone, Ara‐C‐cytarabine, Platinol; FAC: 5‐Fluorouracil, Doxorubicin, Cyclophosphamide; FIGO: International Federation of Gynaecology and Obstetrics; FSH: follicle‐stimulating hormone; GnRH: gonadotropin‐releasing hormone; LH: luteinizing hormone; MVPP: Mustine, Vinblastine, Procarbazine, Prednisone; POF: premature ovarian failure; RCT: randomised controlled trial; TC: Paclitaxel, Carboplatin; TP: Paclitaxel, Cisplatin; VAC:Vincristine, Actinomycin, Cyclophosphamide; VAC: Vincristine, Actinomycin, Cyclophosphamide
Characteristics of excluded studies [ordered by study ID]
| Study | Reason for exclusion |
|---|---|
| Bernhard 2007 | Not intervention of interest |
| Demeestere 2016 | Wrong population |
| Mina 2013 | Not outcome of interest |
| Regan 2017 | Not intervention of interest |
| Sverrisdottir 2009 | Wrong population |
| Zhang 2018 | Wrong population |
Characteristics of studies awaiting classification [ordered by study ID]
Anderson 2016.
| Methods | Prospective, multi‐centre, open‐label RCT |
| Participants | 227 premenopausal women with early or locally advanced breast cancer where adjuvant or neoadjuvant chemotherapy was indicated were included |
| Interventions |
|
| Outcomes |
|
| Notes | It's a conference abstract. It did not report the number of women in both groups |
Behringer 2007.
| Methods | RCT |
| Participants | Women (18 to 40 years) with advanced‐stage HL, including stage II with B symptoms and 1 of the following risk factors: extranodal disease and/or large mediastinal mass, stage III, and IV |
| Interventions |
|
| Outcomes |
|
| Notes | The recruitment of the trial was stopped early and only 23 women included in the final analysis. Unclear whether GnRH was administered before or in parallel to chemotherapy and the data were not presented in the abstract |
Ismail‐Khan 2008.
| Methods | RCT |
| Participants | Premenopausal (age < 44 years, FSH < 40 mIU/mL) women receiving (neo)adjuvant chemotherapy for breast cancer |
| Interventions |
|
| Outcomes |
|
| Notes | It's a conference abstract. Unclear whether GnRH was administrated before or in parallel to chemotherapy. It did not report the number of women in both groups |
Karimi‐Zarchi 2011.
| Methods | Double‐blinded RCT |
| Participants | 44 young women with Hodgkin lymphoma during cyclophosphamide chemotherapy were included |
| Interventions |
|
| Outcomes |
|
| Notes | It's a conference abstract. Unclear whether GnRH was administrated before or in parallel to chemotherapy. It did not report the number of women in both groups |
Rossi 2015.
| Methods | Phase III randomised study |
| Participants | 88 women with breast cancer undergoing (neo)adjuvant chemotherapy were included |
| Interventions |
|
| Outcomes |
|
| Notes | It's a conference abstract. Unclear whether GnRH was administrated before or in parallel to chemotherapy |
AMH: anti‐Mullerian hormone; FSH: serum levels of follicle‐stimulating hormone; GnRH: gonadotropin‐releasing hormone; LH: luteinizing hormone; RCT: randomised controlled trial
Characteristics of ongoing studies [ordered by study ID]
Manger 2006.
| Study name | The PREGO‐Study |
| Methods | Prospective randomised study |
| Participants | Young women with Systemic Lupus Erythematosus undergoing cyclophosphamide treatment |
| Interventions | Monthly injection of depot GnRH agonist was administered before and continued throughout the chemotherapy |
| Outcomes | Not known |
| Starting date | Not known |
| Contact information | Karin.manger@rzmail.uni‐erlangen.de |
| Notes | None |
NCT00090844.
| Study name | Triptorelin for preserving ovarian function in premenopausal women receiving chemotherapy for early‐stage breast cancer |
| Methods | randomised phase II trial |
| Participants | Women ≤ 44 years of age with breast cancer |
| Interventions |
|
| Outcomes |
|
| Starting date | Completed |
| Contact information | Not known |
| Notes | Early closure due to low accrual. Other study ID numbers: CDR0000374991; P30CA076292 (U.S. NIH Grant/Contract ); MCC‐0203 (other identifier: Moffitt CCOP Research Base); NCI‐7031 (other identifier: National Cancer Institute Division of Cancer Prevention) |
NCT00380406.
| Study name | NCT00380406 |
| Methods | Double‐blind RCT |
| Participants | Women aged 18 to 38 years |
| Interventions |
|
| Outcomes |
|
| Starting date | January 2007 |
| Contact information | Not known |
| Notes | Other study ID numbers: 2006603‐01H |
NCT02483767.
| Study name | Gonadotropin‐releasing hormone agonist for the preservation of ovarian function during chemotherapy in premenopausal breast cancer (POF) |
| Methods | Prospective randomised control clinical trial |
| Participants | Women aged 18 to 45 years with breast cancer |
| Interventions |
|
| Outcomes |
|
| Starting date | Completed on 17 October 2017 |
| Contact information | Not known |
| Notes | Other study ID numbers: PUMCH‐BC‐ovarian suppression |
NCT02856048.
| Study name | Co‐treatment with GnRH analogues on the ovarian reserve in young women treated with alkylating agents for cancer (PRESOV) |
| Methods | Multicentre, open‐liable RCT |
| Participants | Female participants aged 12 to 25 years with cancer (sarcoma, Ewing, osteosarcoma, lymphoma) |
| Interventions |
|
| Outcomes |
|
| Starting date | Active, not recruiting (27 February 2018) |
| Contact information | Not known |
| Notes | Other study ID numbers: P140932; 2015‐001121‐17 (EudraCT number) |
NCT03444025.
| Study name | Prospective phase II randomised trial of evaluating the effect of adding LHRH analogue to the neoadjuvant chemotherapy treatment of triple negative breast cancer on pathologic complete response (pCR) rates |
| Methods | Open‐label RCT |
| Participants | Premenopausal women aged 18 to 60 years with invasive carcinoma of breast |
| Interventions |
|
| Outcomes |
|
| Starting date | Not yet recruiting (23 February 2018) |
| Contact information | Not known |
| Notes | Other study ID numbers: B2017‐11 |
AMH: anti‐Mullerian hormone; FSH: follicle‐stimulating hormone; GnRH: gonadotropin‐releasing hormone; LH: luteinizing hormone; LHRH: luteinizing Hormone Releasing Hormone
Differences between protocol and review
Premature ovarian failure and ovulation were not prescribed as outcomes in the protocol (Chen 2009). Treatment‐related premature ovarian failure is defined as no resumption of menstrual activity with postmenopausal FSH levels at least one year after the end of chemotherapy. As menstruation recovery or maintenance and FSH levels were two outcomes prespecified in the protocol and premature ovarian failure reflects severe damage to ovaries by chemotherapy, we included premature ovarian failure as a primary outcome in this updated review. Ovulation is generally confirmed by serial ovarian ultrasound monitoring or serum levels of progesterone, which were also two prespecified outcomes in the protocol, and can reflect ovarian function to some extent, and we had included it as a secondary outcome in the original review (Chen 2011). Thus, we continued to analyse ovulation as an outcome in this updated review. Other clinical symptoms of premature ovarian failure, such as vasomotor symptoms, hot flushes, and vaginal dryness, were listed as secondary outcomes in the protocol. As these symptoms are mainly considered as GnRH analogue‐related adverse events, we combined them in the adverse events analyses. We discussed amending the outcome measurements with Clare Jess, Managing Editor of Cochrane Gynaecological, Neuro‐oncology and Orphan Cancers.
In the original review, we discussed the subgroup analysis based on administration routes of GnRH analogues (intramuscular, subcutaneous, and intranasal administration). In fact, intranasal GnRH analogue is rarely used nowadays and only one included study (Waxman 1987), used this method. Thus, we deleted this subgroup analysis in this update. As the effects of GnRH analogues on ovarian protection are closely related to follow‐up periods, we added subgroup analyses based on different follow‐up periods in the update.
Contributions of authors
HC was responsible for developing the protocol, searching for trials, assessing the quality of trials, extracting data, analysing data, developing the review and amending it. JL was responsible for searching for trials, assessing the quality of trials and extracting data. LC and LX were responsible for analysing data and developing the review. WH was responsible for language revision and amending the review.
Sources of support
Internal sources
West China Secondary Hospital, Sichuan University, China
External sources
Cochrane Gynaecological, Neuro‐oncology and Orphan Cancers, UK
Declarations of interest
JL: none known HC: none known LC: none known WH: none known LX: none known
Stable (no update expected for reasons given in 'What's new')
References
References to studies included in this review
Badawy 2009 {published data only}
- Badawy A, Elnashar A, El-Ashry M, Shahat M. Gonadotropin-releasing hormone agonists for prevention of chemotherapy-induced ovarian damage: prospective randomized study. Fertility and Sterility 2009;91:694-7. [DOI] [PubMed] [Google Scholar]
Elgindy 2013 {published data only}
- Elgindy EA, El-Haieg DO, Khorshid OM, Ismail EI, Abdelgawad M, Sallam HN, et al. Gonadatrophin suppression to prevent chemotherapy-induced ovarian damage: a randomized controlled trial. Obsterics and Gynecology 2013;121(1):78-86. [DOI] [PubMed] [Google Scholar]
Gerber 2011 {published data only}
- Gerber B, Minckwitz G, Stehle H, Reimer T, Felberbaum R, Maass N, et al. Effect of luteinizing hormone-releasing hormone agonist on ovarian function after modern adjuvant breast cancer chemotherapy: the GBG 37 ZORO study. Journal of Clinical Oncology 2011;29(17):2334-41. [DOI] [PubMed] [Google Scholar]
Gilani 2007 {published data only}
- Gilani MM, Hasanzadeh M, Ghaemmaghami F, Ramazanzadeh F. Ovarian preservation with gonadotropin-releasing hormone analog during chemotherapy. Asia-Pacific Journal of Clinical Oncology 2007;3:79-83. [Google Scholar]
Giuseppe 2007 {published data only}
- Giuseppe L, Attilio G, Edoardo DN, Loredana G, Cristina L, Vincenzo L. Ovarian function after cancer treatment in young women affected by Hodgkin disease (HD). Hematology 2007;12:141-7. [DOI] [PubMed] [Google Scholar]
Karimi‐Zarchi 2014 {published data only}
- Karimi-Zarchi M, Forat-Yazdi M, Nakhaie-Moghadam M, Miratashi-Yazdi A, Teimoori S, Dehghani-Tafti A. Evaluation of fertility preservation with GnRH agonist in breast cancer cases treated with cyclophosphamide as a chemotherapy drug. In: International Journal of Reproductive Biomedicine. Vol. 4. 2017:Supplement 1.
- Karimi-Zarchi M, Forat-Yazdi M, Vafaeenasab MR, Nakhaie-Moghadam M, Miratashi-Yazdi A, Teimoori S, et al. Evaluation of the effect of GnRH agonist on menstrual reverse in breast cancer cases treated with cyclophosphamide. European Journal of Gynaecological Oncology 2014;35(1):59-61. [PubMed] [Google Scholar]
Lambertini 2015a {published data only}
- Del Mastro L, Boni L, Michelotti A, Gamucci T, Olmeo N, Gori S, et al. Effect of the gonadotropin-releasing hormone analogue triptorelin on the occurrence of chemotherapy-induced early menopause in premenopausal women with breast cancer: a randomized trial. Journal of the American Medical Association 2011;306(3):269-76. [DOI] [PubMed] [Google Scholar]
- Lambertini M, Boni L, Michelotti A, Gamucci T, Scotto T, Gori S, et al. Ovarian suppression with triptorelin during adjuvant breast cancer chemotherapy and long-term ovarian function, pregnancies, and disease-free survival: a randomized clinical trial. Journal of the American Medical Association 2015;314(24):2632-40. [DOI] [PubMed] [Google Scholar]
Leonard 2017 {published data only}
- Leonard RC, Adamson DJ, Bertelli G, Mansi J, Yellowlees A, Dunlop J, et al. GnRH agonist for protection against ovarian toxicity during chemotherapy for early breast cancer: the Anglo Celtic Group OPTION trial. Annals of Oncology 2017;28:1811-6. [DOI] [PubMed] [Google Scholar]
Moore 2015 {published data only}
- Moore HC, Unger JM, Phillips KA, Boyle F, Hitre E, Porter D, et al. Goserelin for ovarian protection during breast-cancer adjuvant chemotherapy. New England Journal of Medicine 2015;372(10):923-32. [DOI] [PMC free article] [PubMed] [Google Scholar]
Munster 2012 {published data only}
- Munster PN, Moore AP, Ismailkhan R, Cox CE, Lacevic M, Gross-King M, et al. Randomized trial using gonadotropin-releasing hormone agonist triptorelin for the preservation of ovarian function during (neo)adjuvant chemotherapy for breast cancer. Journal of Clinical Oncology 2012;30(5):533-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
Song 2013 {published data only}
- Song G, Gao H, Yuan Z. Effect of leuprolide acetate on ovarian function after cyclophosphamide-doxorubicin-based chemotherapy in premenopausal patients with breast cancer: results from a phase II randomized trial. Medical Oncology (Northwood, London, England) 2013;30(3):667. [DOI] [PubMed] [Google Scholar]
Waxman 1987 {published data only}
- Waxman JH, Ahmed R, Smith D, Wrigley PF, Gregory W, Shalet S, et al. Failure to preserve fertility in patients with Hodgkin's disease. Cancer Chemotherapy and Pharmacology 1987;19:159-62. [DOI] [PubMed] [Google Scholar]
References to studies excluded from this review
Bernhard 2007 {published data only}
- Bernhard J, Zahrieh D, Castiglione-Gertsch M, Hurny C, Gelber RD, Forbes JF, et al. Adjuvant chemotherapy followed by goserelin compared with either modality alone: the impact on amenorrhea, hot flashes, and quality of life in premenopausal patients- the International Breast Cancer Study Group trial VIII. Journal of Clinical Oncology 2007;25:263-70. [DOI] [PubMed] [Google Scholar]
Demeestere 2016 {published data only}
- Demeestere I, Brice P, Peccatori FA, Kentos A, Dupuis J, Zachee P, et al. No evidence for the benefit of gonadotropin-releasing hormone agonist in preserving ovarian function and fertility in lymphoma survivors treated with chemotherapy: final long-term report of a prospective randomized trial. Journal of Clinical Oncology 2016;34(22):2568-74. [DOI] [PubMed] [Google Scholar]
Mina 2013 {published data only}
- Mina R, Reiff A, Silva CA, Fernandez PV, Higgins GC, Imundo LF, et al. Phase 2 trial on triptorelin for ovary protection in childhood-onset systemic lupus erythematosus. In: ACR/ARHP Annual Meeting. 2013.
Regan 2017 {published data only}
- Regan MM, Walley BA, Francis PA, Fleming GF, Láng I, Gómez HL, et al. Concurrent and sequential initiation of ovarian function suppression with chemotherapy in premenopausal women with endocrine-responsive early breast cancer: an exploratory analysis of TEXT and SOFT. Annals of Oncology 2017;28:2225-32. [DOI] [PMC free article] [PubMed] [Google Scholar]
Sverrisdottir 2009 {published data only}
- Sverrisdottir A, Nystedt M, Johansson H, Fornander T. Adjuvant goserelin and ovarian preservation in chemotherapy treated patients with early breast cancer: results from a randomized trial. Breast Cancer Research and Treatment 2009;117(3):561-7. [DOI] [PubMed] [Google Scholar]
Zhang 2018 {published data only}
- Zhang Y, Ji Y, Li J, Lei L, Wu S, Zuo W, et al. Sequential versus simultaneous use of chemotherapy and gonadotropin-releasing hormone agonist (GnRHa) among estrogen receptor (ER)-positive premenopausal breast cancer patients: effects on ovarian function, disease-free survival, and overall survival. Breast Cancer Research and Treatment 2018;168(3):1-8. [DOI] [PubMed] [Google Scholar]
References to studies awaiting assessment
Anderson 2016 {published data only}
- Anderson R, Yellowlees A, Dunlop J, Thomas G, Leonard R. Administration of a GnRH agonist during chemotherapy for breast cancer reduces ovarian toxicity in women aged under 40 years. In: Human Reproduction (Oxford, England). Vol. 31. 2016:suppl 1.1.
Behringer 2007 {published data only}
- Behringer K, Wildt L, Haverkamp H, Diehl V, Pfistner B, Engert A. Results of the randomised phase II for the reduction of ovarian failure with use GnRH-analogues and oral contraceptives in young women (18-40 years) treated with intensive chemotherapy for advanced-stage Hodgkin lymphoma (HL). In: Haematologia. Vol. 92 Suppl 5. 2007:63-4.
Ismail‐Khan 2008 {published data only}
- Ismail-Khan R, Minton S, Cox C, Sims I, Lacevic M, Gross-King M, et al. Preservation of ovarian function in young women treated with neoadjuvant chemotherapy for breast cancer: a randomized trial using the GnRH agonist (triptorelin) during chemotherapy. In: Journal of Clinical Oncology: ASCO annual meeting proceedings. Vol. 26. 2008:12.
Karimi‐Zarchi 2011 {published data only}
- Karimi-Zarchi M, Forat-Yazdi M, Nakhaie-Moghadam M. Ovarian function preservation by GnRH agonists during chemotherapy in Hodgkin lymphoma patients- a double blind randomized control trial clinical. In: International Journal of Gynecological Cancer. Vol. 21. 2011:Suppl 3.
Rossi 2015 {published data only}
- Rossi G, Iadarola R, Lambertini M, Poggio F, Levaggi A, Bighin C, et al. The value of hormone serum concentration to predict the gonadotoxic effect of chemotherapy and the efficacy of LHRH analogs as a strategy to reduce treatment-related premature ovarian failure in breast cancer. In: Annals of Oncology. Vol. 26 Supp 6. 2015:vi3–vi25.
References to ongoing studies
Manger 2006 {published data only}
- Manger K, Wildt L, Kalden JR, Manger B. Prevention of gonadal toxicity and preservation of gonadal function and fertility in young women with systemic lupus erythematosus treated by cyclophosphamide: the PREGO-Study. Autoimmunity Reviews 2006;5:269-72. [DOI] [PubMed] [Google Scholar]
NCT00090844 {published data only}
- NCT00090844. Triptorelin for preserving ovarian function in premenopausal women receiving chemotherapy for early-stage breast cancer. www.clinicaltrials.gov/ct2/show/NCT00090844?term=chemotherapy&recrs=abdefghm&type=Intr&intr=GnRH&gndr=Female&rank=6.
NCT00380406 {published data only}
- NCT00380406. Protecting ovaries and fertility during chemotherapy - the proof trial: a randomized controlled trial of gonadotropin releasing hormone agonist (GnRHa) for fertility preservation in oncology patients. www.clinicaltrials.gov/ct2/show/NCT00380406?term=chemotherapy&recrs=abdefghm&type=Intr&intr=GnRH&gndr=Female&rank=12.
NCT02483767 {published data only}
- NCT02483767. Gonadotropin-releasing hormone agonist for the preservation of ovarian function during chemotherapy in premenopausal breast cancer (POF). www.clinicaltrials.gov/ct2/show/NCT02483767?term=chemotherapy&recrs=abdefghm&type=Intr&intr=GnRH&gndr=Female&rank=2.
NCT02856048 {published data only}
- NCT02856048. Co-treatment with GnRH analogs on the ovarian reserve in young women treated with alkylating agents for cancer (PRESOV) [Assessment of the effect of a co-treatment with GnRH analogs on the ovarian reserve in adolescents and young women treated with alkylating agents for cancer]. www.clinicaltrials.gov/ct2/show/NCT02856048?term=chemotherapy&recrs=abdefghm&type=Intr&intr=GnRH&gndr=Female&draw=2&rank=1.
NCT03444025 {published data only}
- NCT03444025. Prospective phase II randomized trial of evaluating the effect of adding LhRh analogue to the neoadjuvant chemotherapy treatment of triple negative breast cancer on pathologic complete response (pCR) rates [Neoadjuvant goserelin for triple negative breast cancer (NeoGONT)]. www.clinicaltrials.gov/ct2/show/NCT03444025?term=chemotherapy&recrs=abdefghm&type=Intr&intr=GnRH&gndr=Female&rank=13.
Additional references
AGO 2017
- AGO clinical practice guidelines in oncology. www.ago-online.de/en/guidelines-mamma/april-2017 Data accessed in October 2018.
Badawy 2009 [pers comm]
- Ahmed Badawy. Additional data [personal communication]. Email to: Ahmed Badawy 16 September 2018.
Bedoschi 2016
- Bedoschi G, Navarro PA, Oktay K. Chemotherapy-induced damage to ovary: mechanisms and clinical impact. Future Oncology 2016;12(20):2333-44. [DOI] [PMC free article] [PubMed] [Google Scholar]
Blumenfeld 1999
- Blumenfeld Z, Avivi I, Ritter M, Rowe JM. Preservation of fertility and ovarian function and minimizing chemotherapy-induced gonadotoxicity in young women. Journal of the Society for Gynecologic Investigation 1999;6:229-39. [DOI] [PubMed] [Google Scholar]
Blumenfeld 2007
- Blumenfeld Z. How to preserve fertility in young women exposed to chemotherapy? The role of GnRH agonist co-treatment in addition to cryopreservation of embrya, oocytes, or ovaries. The Oncologist 2007;12:1044-54. [DOI] [PubMed] [Google Scholar]
Blumenfeld 2008b
- Blumenfeld Z, Von Wolff M. GnRH-analogues and oral contraceptives for fertility preservation in women during chemotherapy. Human Reproduction Update 2008;14:543-52. [DOI] [PubMed] [Google Scholar]
D'avila 2015
- D'avila AM, Biolchi V, Capp E, Corleta H. Age, anti-mullerian hormone, antral follicles count to predict amenorrhea or oligomenorrhea after chemotherapy with cyclophosphamide. Journal of Ovarian Research 2015;8(1):82. [DOI] [PMC free article] [PubMed] [Google Scholar]
Danforth 2005
- Danforth DR, Arbogast LK, Friedman CI. Acute depletion of murine primordial follicle reserve by gonadotropin-releasing hormone antagonists. Fertility and Sterility 2005;83:1333-8. [DOI] [PubMed] [Google Scholar]
Dayangan Sayan 2018
- Dayangan Sayan C, Tulmac OB, Karaca G, Ozkan ZS, Yalcin S, Devrim T, et al. Could erythropoietin reduce the ovarian damage of cisplatin in female rats. Gynecological Endocrinology 2018;34(4):309-13. [DOI] [PubMed] [Google Scholar]
Deeks 2017
- Deeks JJ, Higgins JP, Altman DG (editors) on behalf of the Cochrane Statistical Methods Group. Chapter 9: Analysing data and undertaking meta-analyses. In: Higgins JPT, Churchill R, Chandler J, Cumpston MS (editors), Cochrane Handbook for Systematic Reviews of Interventions version 5.2.0 (updated June 2017), Cochrane, 2017. Available from www.training.cochrane.org/handbook Data accessed in October 2018.
De Koning 2000
- De Koning CH, Popp-Snijders C, Schoemaker J, Lambalk CB. Elevated FSH concentrations in imminent ovarian failure are associated with higher FSH and LH pulse amplitude and response to GnRH. Human Reproduction 2000;15(7):1452-6. [DOI] [PubMed] [Google Scholar]
Elgindy 2013 [pers comm]
- Eman A Elgindy. Additional or missing data [personal communication]. Email to: Eman A. Elgindy 15 September 2018.
Falcone 2005
- Falcone T, Bedaiwy MA. Fertility preservation and pregnancy outcome after malignancy. Current Opinion in Obstetrics and Gynecology 2005;17:21-6. [DOI] [PubMed] [Google Scholar]
Gerber 2011 [pers comm]
- Sibylle Loibl. Additional data [personal communication]. Email to: Sibylle Loibl 15 September 2018.
Giuseppe 2007 [pers comm]
- Loverro Giuseppe. Additional or missing data [personal communication]. Email to: Loverro Giuseppe 16 April 2018.
Guenther 2017
- Guenther V, Alkatout I, Junkers W, Bauerschlag D, Maass N, Otte S. Fertility preservation in female patients with breast cancer - a current overview. Geburtshilfe Frauenheilkd 2017;77(10):1088-94. [DOI] [PMC free article] [PubMed] [Google Scholar]
Hasky 2015
- Hasky N, Uri-Belapolsky S, Goldberg K, Miller I, Grossman H, Stemmer SM, et al. Gonadotrophin-releasing hormone agonists for fertility preservation: unraveling the enigma? Human Reproduction 2015;30(5):1089-101. [DOI] [PubMed] [Google Scholar]
Hickman 2018
- Hickman LC, Llarena NC, Valentine LN, Liu X, Falcone T. Preservation of gonadal function in women undergoing chemotherapy a systematic review and meta-analysis of the potential role for gonadotropin-releasing hormone agonists. Journal of Assisted Reproduction and Genetics 2018;35(4):571-81. [DOI] [PMC free article] [PubMed] [Google Scholar]
Higgins 2003
- Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003;327:557-60. [DOI] [PMC free article] [PubMed] [Google Scholar]
Higgins 2011a
- Higgins JP, Altman DG, Gøtzsche PC, Jüni P, Moher D, Oxman AD, et al. The Cochrane Collaboration's tool for assessing risk of bias in randomised trials. BMJ 2011;343:d5928. [DOI] [PMC free article] [PubMed] [Google Scholar]
Higgins 2011b
- Higgins JP, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. Available from www.handbook.cochrane.org Data accessed in September 2018.
Higgins 2017
- Higgins JP, Altman DG, Sterne JA (editors). Chapter 8: Assessing risk of bias in included studies. In: Higgins JPT, Churchill R, Chandler J, Cumpston MS (editors), Cochrane Handbook for Systematic Reviews of Interventions version 5.2.0 (updated June 2017), Cochrane, 2017. Available from www.training.cochrane.org/handbook Data accessed in September 2018.
Horicks 2015
- Horicks F, Van Den Steen G, Houben S, Englert Y, Demeestere I. Folliculogenesis is not fully inhibited during GnRH analogues treatment in mice challenging their efficiency to preserve the ovarian reserve during chemotherapy in this model. PLoS One 2015;10(9):e0137164. [DOI] [PMC free article] [PubMed] [Google Scholar]
Horicks 2018
- Horicks F, Van Den Steen G, Gervy C, Clarke HJ, Demeestere I. Both in vivo FSH depletion and follicular exposure to gonadotrophin-releasing hormone analogues in vitro are not effective to prevent follicular depletion during chemotherapy in mice. Molecular Human Reproduction 2018;24(4):221-32. [DOI] [PubMed] [Google Scholar]
Huang 2009
- Huang YH, Zhao XJ, Zhang QH, Xin XY. The GnRH antagonist reduces chemotherapy-induced ovarian damage in rats by suppressing the apoptosis. Gynecologic Oncology 2009;112(2):409-14. [DOI] [PubMed] [Google Scholar]
Imai 2007
- Imai A, Sugiyama M, Furui T, Tamaya T, Ohno T. Direct protection by a gonadotropin-releasing hormone analog from doxorubicin-induced granulosa cell damage. Gynecologic and Obstetric Investigation 2007;63:102-6. [DOI] [PubMed] [Google Scholar]
Kalich‐Philosoph 2013
- Kalich-Philosoph L, Roness H, Carmely A, Fishel-Bartal M, Ligumsky H, Paglin S, et al. Cyclophosphamide triggers follicle activation and "burnout"; AS101 prevents follicle loss and preserves fertility. Science Translational Medicine 2013;5(185):185ra62. [DOI] [PubMed] [Google Scholar]
Kano 2017
- Kano M, Sosulski AE, Zhang L, Saatcioglu HD, Wang D, Nagykery N, et al. AMH/MIS as a contraceptive that protects the ovarian reserve during chemotherapy. Proceedings of the National Academy of Sciences of the United States of America 2017;114(9):E1688-97. [DOI] [PMC free article] [PubMed] [Google Scholar]
Karimi‐Zarchi 2014 [pers comm]
- M Karimi-Zarchi. Additional data [personal communication]. Email to: M. Karimi-Zarchi 8 December 2018.
Lambertini 2015 [pers comm]
- Lucia Del Mastro. Additional data [personal communication]. Email to: Lucia Del Mastro 15 September 2018.
Lambertini 2015b
- Lambertini M, Ceppi M, Poggio F, Peccatori FA, Azim HA, Ugolini D, et al. Ovarian suppression using luteinizing hormone-releasing hormone agonists during chemotherapy to preserve ovarian function and fertility of breast cancer patients: a meta-analysis of randomized studies. Annals of Oncology 2015;51(12):S318-S319. [DOI] [PubMed] [Google Scholar]
Lambertini 2016
- Lambertini M, Del Mastro L, Pescio MC, Andersen CY, Azim HA Jr, Peccatori FA, et al. Cancer and fertility preservation: international recommendations from an expert meeting. BMC Medicine 2016;14:1. [DOI] [PMC free article] [PubMed] [Google Scholar]
Leonard 2017 [pers comm]
- Robert C F Leonard. Additional or missing data [personal communication]. Email to: Robert C. F. Leonard 26 November 2018.
Levine 2015
- Levine JM, Kelvin JF, Quinn GP, Gracia CR. Infertility in reproductive-age female cancer survivors. Cancer 2015;121(10):1532-9. [DOI] [PubMed] [Google Scholar]
Li 2013
- Li X, Kang X, Deng Q, Cai J, Wang Z. Combination of a GnRH agonist with an antagonist prevents flare-up effects and protects primordial ovarian follicles in the rat ovary from cisplatin-induced toxicity: a controlled experimental animal study. Reproductive Biology and Endocrinology 2013;11:16. [DOI: 10.1186/1477-7827-11-16] [DOI] [PMC free article] [PubMed] [Google Scholar]
Luborsky 2000
- Luborsky J, Llanes B, Roussev R, Coulam C. Ovarian antibodies, FSH and inhibin B: Independent markers associated with unexplained infertility. Human Reproduction 2000;15(5):1046-51. [DOI] [PubMed] [Google Scholar]
Lutchman‐Singh 2005
- Lutchman Singh K, Davies M, Chatterjee R. Fertility in female cancer survivors: pathophysiology, preservation and the role of ovarian reserve testing. Human Reproduction Update 2005;11:69–89. [DOI] [PubMed] [Google Scholar]
Maggi 2016
- Maggi R, Cariboni AM, Marelli MM, Moretti RM, Andrè V, Marzagalli M, et al. GnRH and GnRH receptors in the pathophysiology of the human female reproductive system. Human Reproduction Update 2016;22(3):358-81. [DOI] [PubMed] [Google Scholar]
Meirow 2004
- Meirow D, Assad G, Dor J, Rabinovici J. The GnRH antagonist cetrorelix reduces cyclophosphamide-induced ovarian follicular destruction in mice. Human Reproduction 2004;19(6):1294-9. [DOI] [PubMed] [Google Scholar]
Moore 2015 [pers comm]
- Halle CF Moore. Additional or missing data [personal communication]. Email to: Halle C.F. Moore 21 May 2018.
Morgan 2012
- Morgan S, Anderson RA, Gourley C, Wallace WH, Spears N. How do chemotherapeutic agents damage the ovary. Human Reproduction Update 2012;18(5):525-35. [DOI] [PubMed] [Google Scholar]
Munhoz 2016
- Munhoz RR, Pereira AA, Sasse AD, Hoff PM, Traina TA, Hudis CA, et al. Gonadotropin-releasing hormone agonists for ovarian function preservation in premenopausal women undergoing chemotherapy for early-stage breast cancer: a systematic review and meta-analysis. JAMA Oncology 2016;2(1):65-73. [DOI] [PMC free article] [PubMed] [Google Scholar]
Munster 2012 [pers comm]
- Pamela N Munster. Additional or missing data [personal communication]. Email to: Pamela N. Munster 15 September 2018.
Peng 2007
- Peng P, Yang DZ, Zheng CY, Mo YQ, He YM, Zhang QX. Effects of gonadotropin releasing hormone analogues on chemotherapy-induced ovarian function damage in rats. Zhonghua Fu Chan Ke Za Zhi 2007;42:546-50. [PubMed] [Google Scholar]
Review Manager 2014 [Computer program]
- Nordic Cochrane Centre, The Cochrane Collaboration Review Manager 5 (RevMan 5). Version 5.3. Copenhagen: Nordic Cochrane Centre, The Cochrane Collaboration, 2014.
Shelling 2000
- Shelling AN, Burton KA, Chand AL, Van Ee CC, France JT, Farquhar CM, et al. Inhibin: a candidate gene for premature ovarian failure. Human Reproduction 2000;15(12):2644-9. [DOI] [PubMed] [Google Scholar]
Soleimani 2011
- Soleimani R, Heytens E, Darzynkiewicz Z, Oktay K. Mechanisms of chemotherapy-induced human ovarian aging: double strand DNA breaks and microvascular compromise. Aging (Albany NY) 2011;3(8):782-93. [DOI] [PMC free article] [PubMed] [Google Scholar]
Song 2013 [pers comm]
- Guiping Song. Additional data [personal communication]. Email to: Guiping Song 15 September 2018.
Tan 2014
- Tan SJ, Lee LJ, Tzeng CR, Wang CW, Hsu MI, Chen CH. Targeted anti-apoptosis activity for ovarian protection against chemotherapy-induced ovarian gonadotoxicity. Reproductive Biomedicine Online 2014;29(5):612-20. [DOI] [PubMed] [Google Scholar]
Trivers 2014
- Trivers KF, Fink AK, Partridge AH, Oktay K, Ginsburg ES, Li C, et al. Estimates of young breast cancer survivors at risk for infertility in the U.S. Oncologist 2014;19(8):814-22. [DOI] [PMC free article] [PubMed] [Google Scholar]
References to other published versions of this review
Chen 2009
- Chen H, Lina H. Adjuvant gonadotropin-releasing hormone analogues for chemotherapy induced premature ovarian failure in premenopausal women. Cochrane Database of Systematic Reviews 2009, Issue 4. Art. No: CD008018. [DOI: 10.1002/14651858.CD008018] [DOI] [PubMed] [Google Scholar]
Chen 2011
- Chen H, Li J, Cui T, Hu L. Adjuvant gonadotropin-releasing hormone analogues for the prevention of chemotherapy induced premature ovarian failure in premenopausal women. Cochrane Database of Systematic Reviews 2011, Issue 11. Art. No: CD008018. [DOI: 10.1002/14651858.CD008018.pub2] [DOI] [PubMed] [Google Scholar]