GnRH agonist for gonadal protection during chemotherapy.

Approximately 40 000 adolescent and young adult females between the ages of 15 and 39 years are diagnosed with cancer yearly. The high survival rate ofmany cancers has broadened the focus of what is considered an optimal clinical outcome to include preservation of ovarian function. Ovarian function has many implications for the cancer survivor including fertility, bonehealth andpsychosocial well-being. Preservation of ovarian function, fertility and prevention of prematuremenopause are important quality-of-life outcomes. A discussion with patients about the risks of gonadotoxicity of chemotherapy and options to prevent or decrease the devastating consequences is an important aspect of the consultation with themultidisciplinary team. Although alkylating agent-based chemotherapy and radiation therapy are often the agents most associated with ovarian failure, the complex nature of cancer treatments, and possible subsequent recurrence of disease with more treatment, makes the immediate recommendations even more challenging. The use of gonadotrophin releasing hormone agonists (GnRHa) to protect ovarian function from the effects of chemotherapy is the subject of many controversial debates and conflicting clinical reports with no clear consensus. Several systematic reviews and meta-analyses have been published with different results depending on which studies were included or excluded (Bedaiwy et al., 2011; Elgindy et al., 2015). The lack of consistency between randomized clinical trials can be attributed to many factors such as heterogeneity across the studies and multiple different end-points, including surrogate markers for ovarian reserve or even ovarian failure with few studies assessing pregnancy rates. Furthermore, it is important to understand that premature ovarian failure is not universal after exposure to alkylating agents and a significant number of patients retain ovarian function and achieve pregnancy (Schmidt et al., 2013). This clinical observation exemplifies the complex nature of chemotherapy-induced gonadal damage. Since primordial follicles are not hormone-sensitive, the biologic plausibility for the use of GnRHa has been called into question. This implies that we know the specific molecular events that lead to gonadal toxicity by a specific chemotherapeutic drug. In vitro data show a complexmechanism of damage that involves multiple components of ovarian tissue that includes the stromal cells, vascular structures, granulosa cells and the oocyte. The specific mechanism by which primordial follicle loss occurs is unclear andmay only be part of the specific target of chemotherapeutic agents (Morgan et al., 2012). In this issue ofHuman Reproduction, Bildik et al. report an in vitromodel to explore if there ismolecular evidence for-or-against the role ofGrRHa in the prevention of chemotherapy-induced ovarian damage (Bildik et al., 2015). The authors investigated whether GnRHa co-administered with the active metabolite of cyclophosphamide preserves steroidogenic activity and follicle numberandexpressionof anti-apoptotic genes.Overall, this excellent in vitro experiment shows thatGnRHadoes not affect these parameters. The laboratory model used by Bildik et al. is not an established model for how the ovary is exposed to, or responds, to parenteral gonadotoxic chemotherapy in vivo so its clinical applicability is limited. In addition, the in vitro design investigates only one potential mechanism of howGnRHa may work to protect the ovary. GnRHa probably acts at multiple levels and may be protective because of altered vascular effects; the immune system could play a role in protection against chemotherapy-induced gonadotoxicity in vivo. The in vitromodel is also limited by the temporal relationship of the administration of the agonist and chemotherapeutic agent. Typically GnRHa are administered 7–14 days before chemotherapy starts. In this in vitro study, both classes of drugs were administered concurrently or theGnRHawas administered 1–2 h before chemotherapy was given. This is quite a difference from 7 to 14 days in vivo. The study used a mixed patient population for their study materials. Ovarian cortical tissues were obtained from 15 patients (mean age+ SD: 27.8+2.7, range: 14–37) undergoing laparoscopic surgery for the removal of benign ovarian cysts between the years 2014 and 2015. Ovarian cortices embedded in the cyst wall were removed. The number of tissue samples was small (15) with a large age range and the ovarian samples were taken from cyst walls and only contained few follicles to begin with (control samples had a mean 2.5 primordial and 0.6 pre-antral/antral follicles). Most of these patients had endometriomas that can affect apoptosis. Can we extrapolate these in vitro observations to the in vivo clinical situation? The purpose of in vitro experiments is not to exclude a clinical option for patients but to ascertain if a specific