19 Nikolaos Tsampras and Cheryl T. Fitzgerald The introduction of controlled ovarian stimulation (COS) for multiple follicular development significantly increased pregnancy rates (Trounson et al. 1981). Such stimulation protocols have now been developed and refined for more than 25 years in an attempt to prevent premature luteinizing hormone (LH) surge, allow oocyte maturation, and obtain an optimal number of oocytes from each treatment cycle, in order to maximize pregnancy rates per fresh embryo transfer, whilst reducing the risk of over response (Pacchiarotti et al. 2016). The good response to COS is important, as the number of oocytes obtained following stimulation correlates positively with live birth rate (Sunkara et al. 2011). The most commonly used COS protocols are summarized in Figure 19.1. Figure 19.1 Controlled ovarian stimulation (COS) protocols. The classical COS protocols use gonadotropin‐releasing hormone (GnRH) agonists or antagonists to prevent premature LH surge through pituitary suppression. Without suppression, high oestrogen levels during in vitro fertilization (IVF) cycles can result in an LH surge, with consequent lower number of oocytes retrieved and a reduced pregnancy rate (Toner 2002). Spontaneous ovulation has been reported in 16% of unsuppressed cycles (Kadoch et al. 2008). GnRH analogues are decapeptides similar to human GnRH in order to interact with GnRH receptors. These analogues have certain amino acid substitutions in the gonadotropin amino acid sequence that increases the half‐lives and the receptor‐binding affinities of analogues compared with endogenous hormones (Daya 2000; Ortmann et al. 2002). Prolonged activation of GnRH receptors by GnRH agonists leads to desensitization and consequently to suppressed gonadotropin secretion, while GnRH antagonists act as mediators of chemical hypophysectomy (van Loenen et al. 2002; Ortmann et al. 2002). Several agonistic analogues (triptorelin, leuprorelin, deslorelin, goserelin, and nafarelin) and a few antagonistic analogues (cetrorelix and ganirelix) have been introduced into clinical practice (van Loenen et al. 2002). The GnRH long agonist protocol starts with administration of the GnRH agonist in the midluteal phase of the cycle immediately preceding stimulation. This diminishes the GnRH agonist’s flare effect and suppresses endogenous follicle‐stimulating hormone (FSH) production and dominant follicle selection, aiming to promote synchronous follicular growth. After 10–14 days of agonist administration, an ultrasound scan and serum oestradiol level is used to confirm suppression, and gonadotropin stimulation begins (Pacchiarotti et al. 2016). The ovarian response is monitored with frequent transvaginal ultrasound scans and serum oestrogen levels (O’Shea et al. 1988). The adjustment of gonadotropin dose is based on follicular development. Administration of the GnRH agonist and gonadotropin continues until follicles reach 16–18 mm in size (Pacchiarotti et al. 2016). The oocytes final maturation is triggered by administration of 5000–1000 IU of human chorionic gonadotropin (hCG) (Chen et al. 2014). At 34–36 h after triggering, the oocytes are retrieved transvaginally under ultrasound guidance. A variation of the long GnRH agonist protocol is the administration of GnRH agonists for a few months prior to COS. It has been suggested that this pretreatment increases the pregnancy rate when endometriosis is present (Sallam et al. 2006). Short GnRH agonist flare protocols have been advocated for poor responders, attempting to increase oocyte yield. These protocols involve administration of the GnRH agonist in the early follicular phase, aiming to take advantage of the agonist flare effect and increase follicular recruitment (Bstandig et al. 2000). Gonadotropin stimulation begins 1–2 days later while continuing the agonist administration (Pacchiarotti et al. 2016). GnRH antagonists were developed by modifying the GnRH decapeptide at six positions. They compete with GnRH for binding of pituitary receptors. As they have no agonistic activity, they lead to almost immediate endogenous gonadotropin and hence ovarian suppression (Ortmann et al. 2002). Several GnRH antagonist protocols have been developed for assisted reproduction technology (ART). Usually the administration of the antagonist starts on day 3–7 of stimulation and is continuous until the hCG triggering (de Jong et al. 2001; Devroey et al. 2009). Latterly, flexible antagonist protocols have been developed, where the initiation of the antagonist depends on follicular growth (Tarlatzis et al. 2006). The antagonist administration continues, along with gonadotropin stimulation, until follicles reach 16–18 mm. In an antagonist protocol, final oocyte maturation can be achieved either with hCG or alternatively with an agonist trigger. As the duration of the LH surge with the GnRH agonist trigger is short (approximately 34 h), compared with the one provoked by hCG (>6 days), it has been shown to be beneficial for the prevention of ovarian hyperstimulation syndrome (OHSS) (Casper 2015). The effectiveness of different GnRH agonist protocols as adjuncts to COS was extensively assessed in a recent Cochrane review. Thirty‐eight randomized controlled trials (RCT) were included and nine different comparisons between protocols were performed. Several protocols were evaluated regarding administration of the agonist during the follicular or luteal phase, for 2 versus 3 weeks before stimulation, the dose of agonist, and time of discontinuation. The authors concluded that there was no evidence that the GnRH agonist long protocol was associated with an increase in live birth and ongoing clinical pregnancy rates in comparison with the GnRH agonist short protocol, although there was moderate evidence of an increase in clinical pregnancy rates. None of the other comparisons showed any difference in live birth or pregnancy rates between the protocols compared. There was insufficient evidence to make any conclusions regarding adverse effects (Siristatidis et al. 2015). There is very limited evidence comparing the fixed and the flexible GnRH antagonist protocols. So far, antagonist administration from day 3 onward does not appear to reduce the incidence of an LH rise compared with fixed antagonist administration on day 6 of stimulation (Kolibianakis et al. 2011) and the treatment outcome is not compromised (Tannus et al. 2013). A comprehensive review by the Cochrane Collaboration compared the GnRH agonist and antagonist protocols in 2011. The study included 47 RCTs and 7511 patients undergoing ART. The authors concluded that there was no evidence of a difference in live birth rates. However, the use of antagonist compared with long GnRH agonist protocols was associated with a large reduction in OHSS (Al‐Inany et al. 2011). In addition, antagonist protocols allow the use of a GnRH agonist trigger, for oocyte final maturation, reducing further the risk of OHSS in potentially high responders (Youssef et al. 2014). Therefore, GnRH antagonist protocol is preferable for patients with polycystic ovarian syndrome (PCOS) (Pundir et al. 2012; Nardo et al. 2013; Mathur and Tan, 2014; Tannus et al. 2015). The GnRH antagonist protocol is recommended for patients diagnosed with a malignancy, undergoing COS before chemotherapy/radiotherapy for oocyte storage for future use. In such an emergency setting, delay of COS may result in a significant delay in cancer treatment that may lead to patients forgoing fertility preservation. Use of an antagonist allows random‐start ovarian stimulation, decreasing total time for the COS cycle, without compromising oocyte yield and maturity (Cakmak and Rosen 2015). Literature suggests that fertilization, implantation, and pregnancy rates with ART are lower in patients with endometriosis compared with patients with tubal factor infertility, possibly as result of poorer oocyte quality (Barnhart et al. 2002). A meta‐analysis of three small randomized studies (Sallam et al. 2006) in which prolonged, 3–6 months pretreatment with a GnRH agonist was compared with controls, showed a significant improvement of clinical pregnancy rates. Therefore, the GnRH agonist protocol, with prolonged ovarian suppression prior to COS might be preferable for patients with endometriosis (Nardo et al. 2013) The goal of gonadotropin stimulation in ART is the synchronous growth of multiple dominant follicles. FSH is the key hormone in this regard. Therefore, gonadotropin administration is an integral part of stimulation protocols (Practice Committee of the American Society for Reproductive Medicine 2008). Gonadotropin preparations can be classified, based on the source, into urine‐derived and those produced using recombinant techniques. Human menopausal gonadotropins (hMG) are derived from the pooled urine of menopausal women and contain both LH and FSH activity. The LH activity may derive from hCG, which can be preferentially concentrated during the purification process or sometimes added to the preparation (Requena et al. 2014). Initial preparations of hMG were only 5% pure. Further improvement of urine‐derived gonadotropins led to the production of purified urinary FSH (urofollitropin, FSH‐P), a product with little LH activity (<1 unit per vial as opposed to 75 IU per vial of hMG), but significant nongonadotropin protein content. Continuous attempts to purify the FSH activity further using monoclonal antibodies directed against FSH, resulted in highly purified urinary FSH (FSH‐HP). This product has less than 0.1 IU LH activity and 75 IU FSH activity per vial, and less than 5% unidentified urinary protein, making it suitable for subcutaneous use. At the same time, the specific activity of FSH in highly purified preparations was significantly higher, and batch‐to batch variability lower, than the previous generation of products (Practice Committee of the American Society for Reproductive Medicine 2008). Advances in recombinant technology techniques allowed the production of gonadotropin in cell lines that have been transfected with plasmids containing the cloned gene for producing gonadotropin subunits. A major advantage of recombinant products is their high batch‐to‐batch consistency (Levi Setti 2006). In addition, the activity of the gonadotropin can be assessed by protein content, rather than by biological assays as in the case of urine‐derived products (Bassett and Driebergen 2005). This allows superior quality control in recombinant preparations. FSH, LH, and hCG preparations have been produced with recombinant technology. There are two variants of recombinant FSH (r‐FSH) available today: (1) follitropin‐alpha (α) and (2) follitropin‐beta (β). Both variants have the same α‐ and β‐chains as native FSH, but differ in their sialic acid residues as a result of different purification processes. It is not considered that these differences have any clinical implications (Practice Committee of the American Society for Reproductive Medicine 2008). Modification of the FSH β‐subunit using recombinant techniques has led to the production of a new molecule, consisting of the α‐subunit of human FSH and a hybrid subunit composed of the C‐terminal peptide of the β‐subunit of hCG coupled with the FSH β‐subunit. This molecule is a long‐acting FSH, named corifollitropin alfa (Elonva) or FSH‐CTP (carboxy terminal peptide) (Corifollitropin Alfa Dose‐finding Study Group 2008; Fauser et al. 2009). With such an array of available preparations, the clinician should consider efficacy, safety, consistency, ease of administration, and cost, in order to choose the most suitable for each patient. Van Wely et al. published the landmark Cochrane review, including 42 trials with a total of 9606 cycles, in 2012. Comparison of r‐FSH to all other gonadotropins combined did not result in any evidence of a statistically significant difference in live birth rate (28 trials, 7339 couples). When different urinary gonadotropins were considered separately, there were significantly fewer live births after r‐FSH than hMG (11 trials, n = 3197), although differences were small and considered unlikely to be clinically significant. There was no evidence of a difference in live births when r‐FSH was compared with FSH‐P (five trials, n = 1430) or when r‐FSH was compared with FSH‐HP (13 trials, n = 2712 (van Wely et al. 2012). The efficacy of corifollitropin‐α was confirmed by a Cochrane review including 2335 patients (Pouwer et al. 2015). Ovarian stimulation carries the risk of excessive response and OHSS. The risk of OHSS relates primarily to patient factors, such as polycystic ovaries. The choice of gonadotropin does not appear to influence the risk (Mathur and Tan 2014
Ovarian Stimulation Protocols
Introduction
Prevention of Premature LH Surge
GnRH Agonist Protocols
GnRH Antagonist Protocols
Comparison of Stimulation Protocols
Follicular Growth in Controlled Ovarian Stimulation: Gonadotropins
Choice of Gonadotropins
Efficacy
Safety
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
