Acknowledgments
The authors wish to thank Shijun Wang for his assistance in preparing this manuscript.
While chemotherapy significantly improves the prognosis of breast cancer patients, it also damages otherwise healthy organs, such as the ovaries. Gonadotropin-releasing hormone (GnRH) agonists may have a protective effect against chemotherapy-induced ovarian toxicity in premenopausal women being treated for breast cancer; however, studies of its clinical efficacy have reported conflicting results.
This meta-analysis was designed to assess the collective data from previous studies of GnRH agonists administered concurrently with chemotherapy to prevent chemotherapy-induced ovarian toxicity in premenopausal women with breast cancer.
Electronic literature databases (Cochrane Library, Medline, and Embase) were searched for relevant randomized controlled trials (RCTs) published prior to April 2012. Only RCTs that compared GnRH agonists plus chemotherapy to chemotherapy alone for premenopausal women with breast cancer were selected. A random-effects model was used to calculate the risk ratios (RRs) for premature ovarian failure (POF) within one year after chemotherapy treatment and rates of resumed menses and spontaneous pregnancy during the follow-up period after cessation of treatment.
Five RCTs composed of 528 patients (GnRH agonist combination, n = 274; chemotherapy alone, n = 254) were included in the meta-analysis. Significantly fewer women treated with GnRH agonist experienced post-chemotherapy POF, yielding a RR of 0.40 (vs. chemotherapy alone, 95% confidence interval [CI] 0.21–0.75). In contrast, both treatment groups experienced similar rates of resumed menses (RR = 1.31, 95% CI 0.93–1.85) and spontaneous pregnancy (RR = 0.96, 95% CI 0.20–4.56).
Concurrent administration of GnRH agonists during chemotherapy treatment of breast cancer in premenopausal women appears to protect against chemotherapy-related POF in the first year after treatment, but appears to have no effect on resumed menses or spontaneous pregnancy rates.
Breast cancer remains one of the most frequently diagnosed malignancies in women. Although the incidence trends of breast cancer decreased from 1998 through 2006, however, the probability of developing breast cancer before age 40 years is 1 for every 206 women.1 Major advances in treatment of breast cancer have achieved significant benefit from adjuvant systemic chemotherapy in terms of prolonged disease-free and overall survival.2 and 3 The increased survival of patients with breast cancer has given rise to fertility problem related to cancer treatment. While the traditional method of adjuvant chemotherapy to treat early stage breast cancer has proven to be even more effective at improving prognosis in this younger population of women,2 it can damage the otherwise healthy ovaries. The manifestations of chemotherapy-induced ovarian toxicity include irregular menses or amenorrhea, which ultimately lead to premature ovarian failure (POF) and infertility. Therefore, preservation of the ovarian function following adjuvant chemotherapy has become a major concern of both patients and clinician-scientists.
Gonadotropin-releasing hormone (GnRH) is a key regulator of the hypothalamic–pituitary–gonadal reproductive axis, and studies of therapeutic administration of GnRH agonists have revealed that long-term exposure produces a reversible repression of the reproductive axis.4 Therefore, GnRH agonists were applied as adjuvant endocrine therapy to silence ovarian activity during chemotherapy treatment, thereby protecting the ovary in premenopausal patients with breast cancer so that ovarian function may be restored upon therapy cessation. Many subsequent studies have demonstrated that GnRH agonists treatment is associated with a reduced rate of chemotherapy-associated ovarian toxicity.5, 6, 7, 8, 9, 10, 11, 12, 13 and 14However, most of these studies were designed as retrospective and non-randomized trials, which precluded their ability to robustly assess ovarian and fertility preservation in the GnRH-treated patients.
Based on the previously published clinical evidence, two well-designed meta-analyses15 and 16 determined that GnRH agonists treatment during chemotherapy significantly benefited ovarian preservation in premenopausal women. Another meta-analysis17 provided evidence that the GnRH agonists treatment can preserve both ovarian function and fertility. In each of these meta-analyses, however, there was substantial heterogeneity in the types of disease, including hematologic diseases (lymphoma and leukemia), systemic lupus erythematosus, and breast cancer. Therefore, the strength of association between GnRH-related preservation of ovarian function with breast cancer remains controversial. The recent publication of two new RCTs13 and 14 that investigated the effects of GnRH agonists on ovarian preservation in breast cancer patients instigated our efforts to perform a focused meta-analysis using the larger set of accumulated clinical evidence.
Here, we present the first concentrated meta-analysis using the most comprehensive, up-to-date clinical data available in the public literature to quantitatively assess the efficacy of GnRH agonists to protect against chemotherapy-induced ovarian damage in premenopausal women with breast cancer.
The medical literature electronic databases of PubMed, Embase, and the Cochrane Library were searched for relevant randomized controlled trials (RCTs) published prior to April 2012. No language restrictions were applied. Potentially relevant RCTs were identified by various combinations of the following search terms: breast cancer, buserelin, triptorelin, goserelin, leuprolide, histrelin, nafarelin, gonadotropin-releasing hormone agonists, luteinizing hormone (LH)-releasing hormone agonists, chemotherapy, and ovarian failure. When multiple publications were identified for the same population, only the most recent publication was selected. In addition, the reference list of each selected relevant publication was manually searched to identify any additional eligible trials.
The following inclusion criteria were used to select RCTs for the meta-analysis: (1) patient group consisted of women who were premenopausal with histological diagnosis of breast cancer; (2) patients had been treated with GnRH agonists plus chemotherapy and compared to patients treated with chemotherapy alone; and (3) GnRH agonists interventions were administered concurrently with chemotherapy. For trials with more than two treatment arms, each valid pairwise comparison was considered separately. Trials were excluded based upon the following criteria: (1) patient group consisted of women with locally advanced or metastatic disease; (2) different chemotherapy regimens had been compared between the GnRH agonists and control groups; or (3) case–control design, reviews, letters, or case reports.
The primary outcome measures were the proportion of POF within one year after completion of the chemotherapy treatment and the incidence of women with resumption of menses during the follow-up period after cessation of treatment. Amenorrhea means the absence of menstruation for 3 or more months in women with past menses. POF is defined by the individuals' investigators. Secondary outcome measures included the proportions of women with spontaneous pregnancy or adverse effects during the follow-up period after cessation of treatment. If more than one pregnancy was reported for a single patient, only the first pregnancy was included in order to avoid a possible unit-of-analysis error.
Two reviewers (Bo Yang and Weiwei Shi) independently extracted the data from each trial using a standardized form with predefined criteria that had been developed specifically for this review and which included the following items: (1) baseline demographics: author, year of publication, and country of study; (2) participants: sample size and age; (3) GnRH agonists intervention; (4) chemotherapy intervention; (5) duration of intervention; (6) outcome measures; and (7) adverse effects. Discrepancies between the extracted datasets were resolved by discussion.
The methodological quality of each study was assessed in accordance with the guidelines in the Cochrane reviewers' handbook.18 The following trial design features were assessed: (1) measured or unmeasured baseline characteristics due to the method by which trial participants had been selected or assigned; (2) care provided apart from the intervention being evaluated; (3) method by which outcomes were ascertained, diagnosed, and verified; and (4) withdrawal or exclusion of participants throughout the course of the trial. If all quality criteria were met, the trial was considered to have low risk of bias (score: A). If one or more of the quality criteria were only partially met, the trial was considered to have moderate risk of bias (score: B), and if one or more criteria not met, the trial was considered to have high risk of bias (score: C).
Included trials were stratified according to the type of intervention and the subsequent comparisons of GnRH agonists plus chemotherapy vs. chemotherapy treatment. All analyses were performed with STATA statistical software (version 12.0; STATA Corp LP, College Station, TX, USA). Dichotomous data were used to calculate relative risk (RR) with a 95% confidence interval (CI). Homogeneity of RR across studies was assessed by using the Cochrane Q statistic (with p < 0.10 indicating statistically significant heterogeneity) and the I2 statistic (with <40% indicating “heterogeneity might not be important” and >75% indicating “considerable heterogeneity”, based on the suggestion of the Cochrane Handbook for Systematic Review of Interventions19). As there was substantial heterogeneity in the types of chemotherapy used between the different trials, a random effects model (Mantel–Haenszel heterogeneity) was used to calculate the pooled RR.
Potential publication bias was also assessed by both Begg's rank correlation test20 and Egger's linear regression test,21 with p <0.10 indicating statistical significance. Finally, sensitivity analysis was used to investigate the influence of a single study on the overall risk estimate, and was carried out by sequentially omitting one study at each turn with the metaninf algorithm in STATA. A p-value of <0.05 was considered statistically significant.
A total of 276 potentially relevant publications were identified by the initial electronic search. After reviewing the full-texts, only five studies6, 10, 11, 12, 13 and 14 met the inclusion criteria (Fig. 1).
Fig. 1. Flow chart of trial selection process for meta-analysis.
Characteristics of the five RCTs are listed in Table 1. The five trials were composed of 528 premenopausal breast cancer patients, including 274 who had received GnRH with chemotherapy and 254 who had received chemotherapy alone. GnRH treatment consisted of goserelin10, 11 and 12 or triptorelin.13 and 14 Both of the GnRH agonists were administered once every four weeks throughout the chemotherapy treatment period. In most of the trials,12, 13 and 14 the GnRH agonist treatment was initiated at least one week prior to the chemotherapy treatment.
Table 1. Summary of clinical studies included in the meta-analysis.
Abbreviations: ER, estrogen receptor; PR, progesterone receptor; CMF, cyclophospamide, methotrexate, and fluorouracil; NP, not report; FSH, follicle-stimulating hormone; LH, luteinizing hormone; POF, premature ovarian failure.
All of the included trials were classified as having low or moderate risk of bias according to the methodological quality assessment.
Four of the RCTs10, 12, 13 and 14 evaluated POF rate within one year after the chemotherapy treatment had been completed. One trial10 assessed the rate within eight months after treatment, while the other three assessed the rate for 12 months. As shown in Fig. 2, although the heterogeneity among the four trials was insignificant (I2 = 48.7%, p = 0.119), we also used the random effect model method because of the obvious clinical heterogeneity. Administration of GnRH agonists was associated with reduced POF rates (vs. chemotherapy alone: RR = 0.40, 95% CI 0.21–0.75).
Fig. 2. Forest plots showing RR of POF rate of eligible studies comparing GnRH agonists plus chemotherapy with chemotherapy alone in a random effect model.
All five of the RCTs evaluated the resumption of menses in relation to GnRH agonist treatment during the longest follow-up. As shown in Fig. 3, the rates of menses resumption were not significantly different between the GnRH agonist-treated patients and the patients treated with chemotherapy alone. The RR of 1.31 (95% CI, 0.93–1.85) indicated no significant association with menses resumption. The heterogeneity among the five trials was significant (I2 = 92.2%, p = 0.000). However, there was some change of RR in a fixed-effect model (RR = 1.32; 95% CI, 1.16–1.50).
Fig. 3. Forest plots showing RR of resumption of menses rate of eligible studies comparing GnRH agonists plus chemotherapy with chemotherapy alone in a random effect model.
Three of the RCTs12, 13 and 14 evaluated the spontaneous pregnancy rate in relation to GnRH agonist treatment during the longest follow-up. As shown in Fig. 4, although the heterogeneity among the three trials was not significant (I2 = 6.4%, p = 0.344), we also used the random effect model method because of the obvious clinical heterogeneity. The rates of spontaneous pregnancy were not significantly different between the GnRH agonist-treated patients and the patients treated with chemotherapy alone. The RR of 0.96 (95% CI 0.20–4.56) indicated no significant association with spontaneous pregnancy.
Fig. 4. Forest plots showing RR of spontaneous pregnancy rate of eligible studies comparing GnRH agonists plus chemotherapy with chemotherapy alone in a random effect model.
Two of the five RCTs12 and 13 reported adverse events experienced by patients in the treatment groups. Hot flashes, mood swings, urogenital symptoms, insomnia, headache, and sweating were reported. As shown inFig. 5, the random effect model method was used because of the obvious clinical heterogeneity. The RR for adverse effects was 1.24 (95% CI 0.91–1.68) and the difference between the two treatment groups was not statistically significant (p = 0.17).
Fig. 5. Forest plots showing RR of adverse effects of eligible studies comparing GnRH agonists plus chemotherapy with chemotherapy alone in a random effect model.
As shown in Fig. 6, the quantitative summary measure of RR (95% CI) for POF changed very little by sequential omission of individual trials. Moreover, the two trials with the most influence on the results,10 and 14 did not change the direction of the effect.
Fig. 6. RR and 95% CI by omitting each study from the included studies of POF rate.
Neither the Begg's rank correlation test (p = 1.00) nor the Egger's linear regression test (p = 0.925) showed any evidence of publication bias for the trials reporting adjusted RR of POF rate.
The major findings of the current meta-analysis provide prospective evidence that administration of GnRH agonists concurrently with chemotherapy can significantly decrease the POF rate in premenopausal women within one year after completion of the breast cancer treatment. Specifically, the POF rate of women who received GnRH agonists concurrently with chemotherapy is 60% lower than in their counterparts who received chemotherapy alone.
Premature (or primary) ovarian failure (also known as primary ovarian insufficiency) is a form of infertility that affects women under the age of 40. Clinically defined as primary or secondary amenorrhea lasting more than four consecutive months, POF can be diagnosed by increased serum level of follicle-stimulating hormone (FSH; to >40 IU/mL, confirmed by a second measurement one month later) coupled with decreased estrogen levels.22, 23, 24 and 25 However, the five RCTs included in the current meta-analysis used various definitions of POF to identify their study populations. Most of trials11, 13 and 14 defined POF as no resumption of the menstrual cycle after cessation of chemotherapy, while one of the trials10, 12 and 14 included measurements of FSH, LH, estradiol, or inhibin A and B in their diagnoses. It is important to note that this difference between the trials may have impacted our meta-analysis results.
The incidence rates reported for permanent POF following systemic therapy for breast cancer have ranged from 33% to 76%.26 and 27 For example, Badawy et al.10 reported that POF in the control group (75%) may be related to the acute temporary effect of the chemotherapy. It is well-known that menstruation alone does not accurately reflect ovarian function,28 and biochemical measurement of a panel of established serum markers of ovarian function or reserve (including anti-Mullerian hormone (AMH), FSH, LH, estradiol, and inhibin-B) is the preferred clinical indicator. Monitoring of these hormone markers will help to detect chemotherapy-related ovarian toxicity prior to the physical sign of menstrual cessation. Measurement of AMH coupled with ultrasound-assisted antral follicle counting has been suggested as an effective means of assessing the ovarian reserve following breast cancer chemotherapy.29 Unfortunately, we were unable to include an analysis of the hormone markers to more accurately assess POF or ovarian reserve in the pooled patients from the RCTs in the current meta-analysis due to the lack of data.
Clinical studies have shown that age represents a significant risk factor for chemotherapy-induced POF.30Normal ovarian function itself is age-dependent, with ovarian function declining with age. Not surprisingly, older women experience a higher incidence of complete POF and permanent infertility following chemotherapy, as compared to their younger counterparts.31 and 32 Moreover, the risk of chemotherapy-induced amenorrhea is also directly related to the particular chemotherapy regimen.26 Again, we were unable to perform subgroup analysis of POF based on age and chemotherapy regimen due to the limited data. The age range of the patients from the five RCTs was inconsistent, as were the chemotherapy regimens. It should be noted that in older patients, the therapeutic effect of GnRH agonist may be decrease with decline in ovarian function.
Infertility is another potential complication of adjuvant chemotherapy that is experienced by some premenopausal women. These women may continue to menstruate or resume their menses after chemotherapy, despite having abnormal fertility. In the current meta-analysis, three of the RCTs12, 13 and 14evaluated spontaneous pregnancy rate but no statistically significant increase in spontaneous pregnancies was found for GnRH agonist treatment. However, the follow-up duration was too short to evaluate the real influence on fertility.
Despite the protective effect of GnRH agonist treatment on preserving ovarian function in premenopausal women treated with chemotherapy, a major concern of both patients and treating physicians is the potential side effects of the GnRH agonists themselves. GnRH agonists suppression of the reproductive axis causes typical symptoms of menopause, such as hot flashes, headaches, mood changes, sweating, and dry skin, as well as decreased bone density and possible predisposition to osteoporosis or bone fracture. However, the current meta-analysis found no statistically significant differences on adverse effects between the women treated with GnRH agonists plus chemotherapy and those treated with chemotherapy alone (p = 0.17). Another concern is that GnRH agonists may reduce the effectiveness of the chemotherapy.33 For this issue, a previous meta-analysis34 assessed pooled data of 11,906 premenopausal women with early breast cancer from 16 RCTs and concluded that the administration of LH-RH agonists as adjuvant treatment did not reduce the effectiveness of the chemotherapy. Interestingly, most of the studies that did not meet the inclusion criteria for the current meta-analysis also found a beneficial effect of GnRH for reducing chemotherapy-associated POF in premenopausal patients with breast cancer. 5, 7, 8 and 9 However, these studies lacked a control group, so that it was impossible to rule out age, rather than GnRH agonists treatment, as the main determinant of ovarian function preservation.
Although only RCTs were included in the current meta-analysis, several potential limitations exist that may have impacted the results. First, despite that fact that no language restrictions were applied to our literature search, only English languages RCTs were identified. It is possible that some relevant clinical data published in other languages may have been overlooked. Second, although there was no significant overall heterogeneity (I2 = 19.7% in POF) among the five included studies, the obvious lack of uniform chemotherapy regimens, follow-up duration, and POF definition may have impacted the analysis. In addition, there is a major heterogeneity of the primary trials with regards to inclusion or exclusion of estrogen receptor positive patients and treatment with tamoxifen. Treatment with endocrine agents is a significant potential source of bias of individual studies. Third, women recruited to these five RCTs were between the ages of 13 and 47, and only one of the RCTs10 specifically concentrated on women under 40 years old. The inconsistent age of patients included in these RCTs may have lead to bias in the current meta-analysis. Fourth, environmental factors, such as occupational exposure to chemicals, has been implicated in POF25; it is possible that an unidentified subgroup of patients represents a particular risk factor and biased the results of the current meta-analysis. Finally, the limited length of follow-up in the RCTs precludes our ability to determine the long-term impact of GnRH agonists on preservation of ovarian function and fertility.
Concurrent administration of GnRH agonists during chemotherapy appears to be an effective method to reduce the POF rate in premenopausal women with breast cancer. However, additional well-designed studies with larger and more diverse populations are needed to confirm the protective effects of GnRH agonists against chemotherapy-induced ovarian damage and determine the related factors.
There are no conflicts of interest for any of the authors.
The authors wish to thank Shijun Wang for his assistance in preparing this manuscript.
Cancer statistics, 2010
CA Cancer J Clin, 60 (2010), pp. 277–300
Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomised trials
Lancet, 365 (2005), pp. 1687–1717
Adjuvant chemotherapy in oestrogen-receptor-poor breast cancer: patient-level meta-analysis of randomised trials
Lancet, 371 (2008), pp. 29–40
Gonadotropin-inhibitory hormone (GnIH) and its control of central and peripheral reproductive function
Front Neuroendocrinol, 31 (2010), pp. 284–295
Goserelin as ovarian protection in the adjuvant treatment of premenopausal breast cancer: a phase II pilot study
Anticancer Drugs, 13 (2002), pp. 417–424
Ovarian preservation with gonadotropin-releasing hormone analog during chemotherapy
Asia-Pac J Clin Oncol, 3 (2007), pp. 79–83
Gonadotropin-releasing hormone analogues added to adjuvant chemotherapy protect ovarian function and improve clinical outcomes in young women with early breast carcinoma
Cancer, 106 (2006), pp. 514–523
Prevention of chemotherapy-induced menopause by temporary ovarian suppression with goserelin in young, early breast cancer patients
Ann Oncol, 17 (2006), pp. 74–78
Ovarian protection with goserelin during adjuvant chemotherapy for pre-menopausal women with early breast cancer (EBC)
Breast Cancer Res Treat, 110 (2008), pp. 411–416
Gonadotropin-releasing hormone agonists for prevention of chemotherapy-induced ovarian damage: prospective randomized study
Fertil Steril, 91 (2009), pp. 694–697
Adjuvant goserelin and ovarian preservation in chemotherapy treated patients with early breast cancer: results from a randomized trial
Breast Cancer Res Treat, 117 (2009), pp. 561–567
Effect of luteinizing hormone-releasing hormone agonist on ovarian function after modern adjuvant breast cancer chemotherapy: the GBG 37 ZORO study
J Clin Oncol, 29 (2011), pp. 2334–2341
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
J Am Med Assoc, 306 (2011), pp. 269–276
Randomized trial using gonadotropin-releasing hormone agonist triptorelin for the preservation of ovarian function during (neo)adjuvant chemotherapy for breast cancer
J Clin Oncol, 30 (2012), pp. 533–538
Ovarian preservation by GnRH agonists during chemotherapy: a meta-analysis
J Womens Health (Larchmt), 18 (2009), pp. 311–319
Adjuvant gonadotropin-releasing hormone analogues for the prevention of chemotherapy induced premature ovarian failure in premenopausal women
Cochrane Database Syst Rev (2011) CD008018
Gonadotropin-releasing hormone analog cotreatment for preservation of ovarian function during gonadotoxic chemotherapy: a systematic review and meta-analysis
Fertil Steril, 95 (2011), pp. 906.e1-4–914.e1-4
Cochrane handbook for systematic reviews of interventions version 5.0.2
(2009)
Measuring inconsistency in meta-analyses
BMJ, 327 (2003), pp. 557–560
Operating characteristics of a rank correlation test for publication bias
Biometrics, 50 (1994), pp. 1088–1101
Bias in meta-analysis detected by a simple, graphical test
BMJ, 315 (1997), pp. 629–634
Premature ovarian failure
Br Med Bull, 56 (2000), pp. 643–649
Premature ovarian failure
Obstet Gynecol, 113 (2009), pp. 1355–1363
General mechanisms of premature ovarian failure and clinical check-up
Gynecol Obstet Fertil, 36 (2008), pp. 857–861
Occupational exposures to chemicals as a possible etiology in premature ovarian failure: a critical analysis of the literature
Reprod Toxicol, 33 (2012), pp. 269–279
Ovarian function in premenopausal women treated with adjuvant chemotherapy for breast cancer
J Clin Oncol, 14 (1996), pp. 1718–1729
The rates of chemotherapy-induced amenorrhea in patients treated with adjuvant doxorubicin and cyclophosphamide followed by a taxane
Am J Clin Oncol, 30 (2007), pp. 126–132
Ovarian reserve testing and the use of prognostic models in patients with subfertility
Hum Reprod Update, 7 (2001), pp. 581–590
The effects of chemotherapy and long-term gonadotrophin suppression on the ovarian reserve in premenopausal women with breast cancer
Hum Reprod, 21 (2006), pp. 2583–2592
American Society of Clinical Oncology recommendations on fertility preservation in cancer patients
J Clin Oncol, 24 (2006), pp. 2917–2931
Fertility and the impact of systemic therapy on hormonal status following treatment for breast cancer
Curr Oncol Rep, 2 (2000), pp. 587–593
Chemotherapy-induced amenorrhea and fertility in women undergoing adjuvant treatment for breast cancer
Cancer Control, 9 (2002), pp. 466–472
Increased thromboembolic complications with concurrent tamoxifen and chemotherapy in a randomized trial of adjuvant therapy for women with breast cancer. National Cancer Institute of Canada Clinical Trials Group Breast Cancer Site Group
J Clin Oncol, 14 (1996), pp. 2731–2737
Use of luteinising-hormone-releasing hormone agonists as adjuvant treatment in premenopausal patients with hormone-receptor-positive breast cancer: a meta-analysis of individual patient data from randomised adjuvant trials
Lancet, 369 (2007), pp. 1711–1723