Phase II Study of Gonadotropin-Releasing Hormone Analog for Ovarian Function Preservation in Hematopoietic Stem Cell Transplantation Patients

Results of a phase II clinical trial of leuprolide administered before conditioning chemotherapy in hematopoietic stem cell transplantation patients to reduce the incidence of premature ovarian failure are presented. Leuprolide did not preserve ovarian function in patients who underwent hematopoietic stem cell transplantation using either myeloablative or nonmyeloablative regimens.

Abstract

Purpose.

Premature ovarian failure occurs in 40%–70% of patients who receive conventional chemotherapy alone. However, the incidence is higher, 70%–100%, in patients who undergo myeloablative chemotherapy with hematopoietic stem cell transplantation (HSCT). Gonadotropin-releasing hormone (GnRH) analogs, such as leuprolide, in a continuous-release formulation, may protect the ovaries from the gonadotoxic effects of chemotherapy. In non-HSCT settings, GnRH analogs have reduced the risk for premature ovarian failure to <10%. We conducted a phase II clinical trial based on the hypothesis that giving leuprolide before conditioning chemotherapy in HSCT patients reduces premature ovarian failure incidence.

Patients and Methods.

Eligible patients were women aged ≤40 years who were HSCT candidates, were premenopausal, and had both follicle-stimulating hormone (FSH) and luteinizing hormone (LH) levels ≤20 IU/L. Two 22.5-mg leuprolide doses were delivered in 3-month depot i.m. injections, the first within 2 months before HSCT. Patients were monitored for menstruation return, and ovarian function tests (FSH, LH, and estradiol) were done every 2 months starting 90 days after the last leuprolide dose.

Results.

Sixty eligible patients were enrolled, 59 underwent HSCT, and 44 were evaluable (median age, 25 years; median follow-up, 355 days). Only seven of 44 patients (16%) regained ovarian function. Of the 33 who received myeloablative regimens, six (18%) regained ovarian function. However, among the 11 who received nonmyeloablative regimens, only one (9%) regained ovarian function (p = .66).

Conclusion.

Leuprolide did not preserve ovarian function in patients who underwent HSCT using either myeloablative or nonmyeloablative regimens. Other measures that protect ovarian function need to be investigated.

Introduction

With advances in hematopoietic stem cell transplantation (HSCT) technology, more patients with different diseases are able to receive this form of treatment and, as a result, survive longer. Many long-term survivors suffer a decrease in overall quality of life—physical, psychological, social, and spiritual [1, 2]. This decrease is often a result of the side effects of the conditioning chemotherapy and/or radiation therapy given before stem cell transplantation. Traditionally, the conditioning regimen is fully myeloablative. Recently, new reduced-dose conditioning regimens have been developed that use less intense chemotherapy [3], but although these regimens are nonmyeloablative, they still cause significant side effects. The reproductive system is commonly affected by chemotherapy, but, unfortunately, is usually given less attention than other body systems.

With the increasing numbers of long-term cancer survivors who are young women, the incidence of treatment-related premature ovarian failure and subsequent infertility is increasing. Patients with premature ovarian failure will suffer the consequences of menopause, such as hot flushing, sweating, depression, emotional lability, crying crisis, asthenia, decreased libido, dyspareunia, and osteoporosis. Infertility has a significant psychological and social impact on patients. In 2006, the American Society of Clinical Oncology (ASCO) published recommendations on fertility preservation in cancer patients [4]. The guideline recognized that treatment-related infertility is an important issue for cancer patients and that oncologists have a role in addressing this issue in addition to cancer management.

In HSCT patients, the commonly used fully myeloablative regimens are high-dose chemotherapy and total-body irradiation. The chemotherapy dosages used in these combinations are very gonadotoxic. The reported incidences of premature ovarian failure in patients who receive these regimens are in the range of 70%–100% [5–8]. In contrast, the chemotherapy regimens used in the nontransplantation setting are usually milder and less gonadotoxic. The incidence of premature ovarian failure in this setting is in the range of 40%–70% [9–11]. The development of nonmyeloablative chemotherapy was based on the recognition that a graft-versus-tumor effect supplements chemotherapy's killing effects in eradicating a malignancy [12]. Using a reduced dose of chemotherapy as the conditioning regimen permits the development of mixed chimerism and the graft-versus-tumor effect. However, even these nonmyeloablative regimens cause significant side effects. The incidence of premature ovarian failure in the nonmyeloablative setting has not been reported, but it is reasonable to predict that it is similar to that of patients who receive conventional chemotherapy alone (40%–70%, as stated above).

In the late 1980s, Ataya and colleagues reported promising results using a gonadotropin-releasing hormone (GnRH) analog to protect the ovaries from chemotherapy effects in animal studies [13, 14]. The mechanism of action of GnRH analogs was believed to involve direct suppression of gonadotropin receptors in the ovary, leading to a reduction in oogenesis and thereby rendering the ovarian follicles less susceptible to the gonadotoxic effects of chemotherapy. We hypothesized that a long-acting GnRH analog, such as leuprolide, given before the start of the conditioning regimen for stem cell transplantation could reduce the incidence of premature ovarian failure. We performed a phase II clinical study with the goals of preserving the fertility of patients who undergo HSCT and preventing them from suffering the immediate and long-term side effects of premature menopause.

Patients and Methods

Trial Design

All patients provided written informed consent prior to participating in the study, and the study was reviewed and approved by the institutional review board at The University of Texas MD Anderson Cancer Center. Eligible patients were women aged <40 years who were HSCT candidates, were postmenarchal and premenopausal before the start of transplantation, and either had a follicle-stimulating hormone (FSH) level ≤20 IU/L and a luteinizing hormone (LH) level ≤20 IU/L within 1 week of initiation of the study or had their ovarian function status determined by the gynecologic oncology service. FSH and LH levels were determined based on the second International Reference Preparation World Health Organization Reference Standard 78/549. Patients who had breast cancer or ovarian cancer, who were pregnant at the time of the study, or who were known to have a hypersensitivity reaction to any GnRH analog were excluded.

Treatment

The first 22.5-mg dose of leuprolide, in a 3-month depot i.m. injection, was given within 2 months before stem cell transplantation. The second dose of leuprolide was given 3 months after the first injection. If the platelet count at the time of either injection was <50,000/μL, leuprolide was given with platelet transfusion support. Patients were monitored for return of menstruation, and ovarian function tests (FSH, LH, and estradiol) were done every 2 months starting 90 days after the second dose of leuprolide. All toxic effects encountered during the study were evaluated according to the National Cancer Institute Common Toxicity Criteria, version 3.0, and recorded accordingly.

Study Evaluation Criteria

Normal ovarian function after transplantation was defined as spontaneous menstruation for ≥3 months with the following hormonal parameters: (a) serum FSH level ≤20 IU/L, (b) serum LH level ≤20 IU/L, and (c) serum estradiol level ≥30 pmol/L. Ovarian failure after transplantation was defined as continuation of cessation of menstruation for ≥3 months in a patient who was premenopausal at the beginning of the study with the following hormonal parameters: (a) serum FSH level >20 IU/L, (b) serum LH level >20 IU/L, and (c) serum estradiol level <30 pmol/L.

The following conditioning regimens were considered myeloablative: busulfan plus cyclophosphamide; busulfan plus fludarabine; cyclophosphamide plus total-body irradiation (>10 Gy); carmustine, etoposide, cytarabine, and melphalan; and a single melphalan dose >140 mg/m². The following conditioning regimens were considered nonmyeloablative: fludarabine plus cyclophosphamide; fludarabine plus melphalan; fludarabine, idarubicin, and cytarabine; a total-body irradiation dose of 2 Gy, and a single melphalan dose ≤140 mg/m².

Statistical Considerations

Patient data were divided into myeloablative and nonmyeloablative groups according to the conditioning regimen given before HSCT. In a separate analysis, the data were also divided by type of transplantation (allogeneic or autologous). Standard descriptive statistics, including the median, range, and ratio, were used in the analysis of the data for the whole group and for the subgroups. Fischer's exact test was used to determine the difference between subgroups.

Results

Patient and Tumor Characteristics

Sixty eligible patients were enrolled in the study, but one of these patients did not undergo HSCT. Of the 59 patients who underwent HSCT, nine patients refused the second dose of leuprolide and were subsequently taken off study. Six patients died of either disease progression or transplantation-related complications before their ovarian function status was evaluated. For the 44 evaluable patients, the median age was 25 years (range, 15–39 years) and the median follow-up was 355 days (range, 102–1,676 days) (Table 1). Thirty-three patients received myeloablative conditioning regimens and 11 patients received nonmyeloablative conditioning regimens. Seven patients received total-body irradiation as a conditioning regimen but two were not evaluable. Twenty-nine patients underwent allogeneic HSCT and 15 patients underwent autologous HSCT. All patients but two received at least one prior chemotherapy regimen. The median number of prior chemotherapy regimens before HSCT was two (range, 0–8), and 12 patients also received prior local radiation. In 10 patients, the local radiation was given above the diaphragm. Only two patients received local radiation in the left hemipelvis and left inguinal area. The most common disease among the 44 patients was Hodgkin's lymphoma, followed by acute myelogenous leukemia (Table 2). The chemotherapy conditioning regimens used are listed in Table 3.

Table 1.

Patient characteristics

Abbreviation: HSCT, hematopoietic stem cell transplantation.

Table 2.

Disease distribution for evaluable patients

Table 3.

Conditioning regimens used

Abbreviation: TBI, total-body irradiation.

Ovarian Function After Stem Cell Transplantation

Of the 44 evaluable patients, only seven (16%) regained their ovarian function. In our analysis of the correlation between regained ovarian function and type of conditioning regimen before transplantation, of the 33 patients who received myeloablative regimens, six patients regained ovarian function (18%). However, among the 11 patients who received nonmyeloablative regimens, only one patient (9%) regained ovarian function. In this subgroup, the p-value was .66 by Fischer's exact test. We also looked at the correlation between regained ovarian function and type of transplantation. Of the 15 patients who underwent autologous transplantation, five patients regained ovarian function (33%). However, among the 29 patients who underwent allogeneic transplantation, only two patients (7%) regained ovarian function. In this subgroup, the p-value was .04 by Fischer's exact test. Of the seven patients who regained ovarian function, none reported a pregnancy. Five of these seven patients had Hodgkin's lymphoma, one had acute lymphocytic leukemia, and one had aplastic anemia. There was no correlation between regained ovarian function and disease diagnosis.

The seven patients who regained ovarian function had a younger median age (22 years) than the 37 patients with no ovarian function (26 years), and most [5] of the seven patients underwent autologous transplantation, but they had a median of three chemotherapy regimens prior to HSCT (Table 1).

Tolerability of Leuprolide

Nine of 59 patients (15%) refused the second dose of leuprolide because of intolerable menopausal side effects, including hot flushing, vaginal dryness, and mood swings. They were subsequently taken off study. No permanent or grade 4 side effects were observed.

Discussion

Our study showed that the use of long-acting leuprolide right before transplantation did not preserve ovarian function in patients who underwent HSCT using either myeloablative or nonmyeloablative regimens.

Chemotherapy-induced premature ovarian failure was first reported in the late 1950s, when three different groups noted that early menopause occurred in chronic myelogenous leukemia patients who received busulfan treatment [15–17]. Subsequently, other drugs have been found to cause premature ovarian failure. Most of these drugs are alkylating agents, such as cyclophosphamide [18–21] and melphalan [22]. The ovarian damage effect is age dependent and dose dependent [23]. The older the patient at the time of transplantation, the higher the chance of chemotherapy-induced premature ovarian failure. Also, the higher the dose of the chemotherapy used, the higher the chance of premature ovarian failure. Patients who develop premature ovarian failure not only manifest symptoms of menopause but also demonstrate abnormal hormonal levels in the blood, such as elevated FSH and LH levels and a lower estradiol level. The pathological feature of chemotherapy-induced ovarian damage is follicular destruction with replacement by fibrosis [21, 24–26].

Most of the ovarian follicles during the reproductive period in a female are inactive primordial follicles and do not undergo cell division. It is believed that alkylating agents cause damage to the undeveloped but still growing oocytes and possibly also to the pregranulosa cells of the primordial follicles [27]. Through the gonadotropin-stimulating effects of the normal hypothalamic-pituitary-gonadal axis, more primordial follicles are recruited to the growing pool and become susceptible to the damaging effects of alkylating agents. Researchers observed that ovarian function was preserved in most long-term survivors of lymphoma who were treated prepubertally, but in fewer patients who were treated as adults [28, 29]. Therefore, it is clinically reasonable to predict that removing the stimulating effects of gonadotropin on the ovary and inducing a temporary prepubertal, dormant stage in the ovary will protect it from the destructive effects of chemotherapy or even radiotherapy.

Leuprolide is a synthetic peptide analog to GnRH. Given at therapeutic doses, leuprolide suppresses gonadal steroidogenesis after a predictable initial stimulatory effect. In premenopausal women, this stimulatory effect results in estrogen-withdrawal bleeding that begins 14–28 days after the administration of leuprolide. The subsequent suppressive effect is totally reversible when the drug is discontinued. Blumenfeld et al. [30], in 1996, first addressed the issue of GnRH analogs' protective effect on ovarian function. They reported that ovarian function was regained in 93.7% of patients who received GnRH analog cotreatment with chemotherapy, compared with only 39% of patients in the control group. In a small study reported in 2001 by Pereyra Pacheco et al. [31], 12 patients received a GnRH analog together with polychemotherapy, and all had return of menstruation after chemotherapy. In contrast, four patients did not receive a GnRH analog but only polychemotherapy and all developed hypergonadotrophic hypoestrogenic amenorrhea. Badawy et al. [32] reported, in 2009, the result of their randomized study of a GnRH analog plus chemotherapy versus chemotherapy alone in patients with breast cancer. In patients who received a GnRH analog plus chemotherapy, 69.2% regained ovarian function, compared with only 25.6% of patients who received chemotherapy alone. Another randomized study of a GnRH analog plus chemotherapy versus chemotherapy alone in patients with breast cancer was reported by Del Mastro et al. [33] in 2011. Of the patients who received a GnRH analog plus chemotherapy, 8.9% had early menopause, compared with 25.9% of patients who received chemotherapy alone. In 2009, Clowse et al. [34] presented their meta-analysis of nine studies comparing a GnRH analog plus chemotherapy with chemotherapy alone. With a total of 366 patients, 178 patients received a GnRH analog during chemotherapy and 93% of them maintained ovarian function. For the other 188 patients who received chemotherapy alone, only 48% maintained ovarian function. However, several small studies suggested no benefit of using a GnRH analog during chemotherapy to preserve ovarian function [35, 36]. The explanation those investigators proposed was based on the finding that primordial follicles do not express FSH receptors, and there is no evidence that they contain GnRH receptors [37].

Our study was the first prospective phase II study to evaluate the efficacy of a GnRH analog in reducing the incidence of premature ovarian failure in the setting of HSCT. However, our result showed that only seven of 44 patients (16%) regained ovarian function. One possible reason for such a poor ovarian outcome may be that the study patients were heavily pretreated prior to HSCT, with a median of two prior chemotherapy regimens (range, 0–8). We also looked at the correlation between ovarian outcome and the type of conditioning regimen. It was surprising that patients who received nonmyeloablative regimens had a lower rate of post-treatment ovarian function than those who received myeloablative regimens, although there was no statistical difference (p-value of .66). When we looked at the correlation between ovarian outcome and transplantation type, patients who underwent allogeneic transplantation had a statistical significant lower rate of post-treatment ovarian function than those who underwent autologous transplantation (p-value of .04). Among the 11 patients who received nonmyeloablative regimens, 10 also underwent allogeneic transplantation. It is possible that the immunogenic outcome of an allogeneic transplant, such as graft-versus-host disease, may also have a detrimental or prolonged effect on ovarian function. Another factor that may affect ovarian outcome is age at HSCT. Among the seven patients who regained ovarian function, the median age was 22 years, which was younger than that of the other 37 patients (median age, 26 years). This finding confirms that the risk for premature ovarian failure is age dependent.

In the past, ovarian preservation was not given much attention during cancer management; most of the focus was on treating the active cancer. Furthermore, systemic treatment options for advanced cancer were limited, and the outcome usually was grave. In such a setting, ovarian preservation became even less important. Nowadays, with the advance of systemic treatment and a multidisciplinary approach to cancer management, many patients are able to become long-term survivors. Quality of life after cancer has thus become a major issue. Nakayama et al. [38] conducted a survey study and reported that most patients believe that a discussion of fertility-related or menopausal-related issues was as important as a discussion of their cancer issues. The ASCO guideline recommends the discussion of ovarian preservation as early as possible in treatment planning [4]. There are many different means of ovarian preservation for patients who undergo cancer treatment that could cause treatment-related infertility. For female patients, the ASCO guideline lists several options, including embryo cryopreservation, oocyte cryopreservation, ovarian cryopreservation and transplantation, oophoropexy, and GnRH analogs. Most of these ovarian preservation options are invasive in nature, and the procedures require a significant length of time. GnRH analog treatment has the advantage of being a simple injection and ready to use. Because of its ease of administration, easy tolerability, and lack of interaction with chemotherapy, a GnRH analog is commonly used in the oncology community for ovarian preservation in patients who receive chemotherapy. Two randomized studies have already shown the benefit of a GnRH analog in ovarian preservation in patients with breast cancer [32, 33]. Although our phase II study did not demonstrate the benefit of a GnRH analog in ovarian preservation in patients who underwent HSCT, the role of GnRH analogs in the transplant setting remains unsolved. To determine the true benefit of a GnRH analog in reducing premature ovarian failure in patients undergoing an HSCT procedure, a large-scale clinical trial is still needed. Other measures protective of ovarian function also need to be investigated for patients who undergo HSCT.

See the accompanying commentary on pages 162–163 of this issue.

Author Contributions

Conception/Design: Yee Chung Cheng, Richard E. Champlin

Provision of study material or patients: Naoto T. Ueno, Yee Chung Cheng

Collection and/or assembly of data: Naoto T. Ueno, Yee Chung Cheng, Mariko Takagi

Data analysis and interpretation: Naoto T. Ueno, Yee Chung Cheng, Andrea Milbourne

Manuscript writing: Naoto T. Ueno, Yee Chung Cheng

Final approval of manuscript: Naoto T. Ueno, Yee Chung Cheng, Mariko Takagi, Andrea Milbourne, Richard E. Champlin