Abstract
The use of gonadotropin-releasing hormone (GnRH) antagonists in in vitro fertilization superovulation remains controversial. The GnRH agonist ‘long protocol’ has been seen as the gold standard for many years. Comparisons and meta-analyses of the efficacy of GnRH antagonists and agonists have been largely inconclusive, with the dataset being contaminated with outdated reports of poorer efficacy with GnRH antagonists, which have stemmed from studies of their use as a second-line drug in older women and women who were poor responders. This work cannot reflect the actual clinical effectiveness of GnRH antagonist and must be interpreted with care. The major advantages of GnRH antagonists use in superovulation include a gentler and more patient-friendly stimulation cycle with less hypoestrogenic side effects, with the potential to lower the risk of ovarian hyperstimulation and enhanced embryo growth. Our current clinical experience with GnRH antagonists in in vitro fertilization is limited, although there are a growing number of in vitro fertilization centers embracing this new technology. There is a clear need for a modern, suitably powered clinical trial to demonstrate the place of GnRH antagonist-based superovulation protocols and in subgroups of patients, such as polycystic ovary syndrome or poor responders.
The process of in vitro fertilization (IVF) involves controlled ovarian stimulation (COS) with gonadotropins to stimulate multifollicular growth. Human chorionic gonadotropin (HCG) is then used to induce final oocyte maturation before the oocytes are retrieved for IVF. It is well established that premature luteinizing hormone (LH) surges have a negative role in IVF and result in increased cancellation rates of IVF treatment. Therefore, it is desirable to suppress LH levels and premature surges in patients undergoing IVF. To achieve this, gonadotropin-releasing hormone (GnRH) agonists were introduced in the late 1980s for pituitary down-regulation. The ‘long protocol’ of using GnRH agonist for the suppression of LH has been established as the standard in ovarian stimulation ever since [1]. The inclusion of GnRH agonists in stimulation protocols brought about significant improvements by reducing cancellation rates and improving pregnancy rates [2]. In the routine use of GnRH agonists in IVF long protocol, treatment with the agonist is started at least 7 days before follicle-stimulating hormone (FSH) injection, since use of the agonist is accompanied by an initial stimulation of the gonadotropins receptors with a resulting increase in circulating LH and FSH (‘flare’) followed by subsequent depletion and desensitization. By contrast, GnRH antagonists bind competitively to GnRH receptors [3], resulting in an immediate suppression of gonadotropin release without a flare. This has allowed a significant reduction in treatment duration and avoids the estrogen-deprivation symptoms associated with GnRH agonist-induced downregulation. While the first generation of GnRH antagonists were complicated by allergic side effects secondary to histamine release, which hampered the development of these compounds, the third-generation GnRH antagonists such as ganirelix and cetrorelix do not have these side effects and exhibit a good safety profile. With the introduction of GnRH antagonists as a substitute for LH suppression in IVF stimulation, the advantages of a more patient-friendly protocol, shorter stimulation time with fewer side effects and gentler stimulation protocols showed great promise. However, the uptake of GnRH antagonists in IVF has been slower than expected. Possible reasons for this low acceptance level include the lack of experience in using antagonist in IVF, the reported lower pregnancy rates and the difficulty in patient scheduling. This article will explore the various aspects of GnRH antagonists use in IVF stimulation protocols and examine the evidence in favor of using GnRH antagonists as standard in IVF stimulation protocols.
Types of GnRH antagonists
Two third-generation GnRH antagonists have been registered for IVF: cetrotrelix (Cetrotide™, Serono international S.A, Geneva, Switzerland) and ganirelix (Orgalutran®, Antagon, Organon Inc., NJ, USA). Both have been demonstrated to be effective and safe in IVF. Thus far, no comparative trials between these GnRH antagonists have been performed.
Use of GnRH antagonists in IVF
Effective starting dose of GnRH antagonist
Early clinical trials have established that the minimal effective dosage for both GnRH antagonists is a daily dose of 0.25 mg for prevention of a premature rise of LH [4,5]. Recent reports have suggested that there is a possible linear relationship between serum antagonist levels and bodyweight. However, current trials were restricted to women of normal body mass index (BMI) (18–29 kg/m2). Consequently this dosage might not be suitable for women at the upper and lower limits of BMI.
Single versus multiple doses
Currently there are two established GnRH antagonist dose protocols developed for use in IVF. These protocols involve either multiple [6] or single [7] administration. In the multiple-dose protocol, GnRH antagonist is initiated 5–7 days after initiation of stimulation with gonadotropins and is administered daily until the start of HCG administration. In the single-dose protocol, a dose of 3 mg of GnRH antagonist is initiated on day 7 of stimulation [8], with additional single doses of GnRH antagonist added 4 days following the single dose when necessary.
The advantage of a single dose is obviously the need for fewer injections, but in 10% of cycles, additional single daily doses are required [9]. The sole possible disadvantage of the single-dose protocol is the possibility of excessive suppression of endogenous LH during the first few days following injection, which may have an adverse effect on pregnancy rates. Currently, no significant difference in pregnancy rates have been shown in a randomized controlled trial (RCT) comparing these two protocols [10,11].
Alternate-day injection of GnRH antagonist might be another novel protocol of GnRH administration. Only one RCT explored this method and reported similar LH suppression efficacy compared with daily injection of GnRH antagonists and a significantly lower GnRH antagonist usage. This study was underpowered, but suggested that alternate-day dosing was as effective as single-day dosing in preventing premature luteinization [12].
Fixed versus flexible administration
GnRH antagonists have the advantage of flexibility in that they may be administered at any time during the follicular phase of a treatment cycle. Several studies have been performed to ascertain the optimum treatment schedule. Two general approaches have emerged for the commencement of antagonist treatment in IVF cycles; the flexible start or the fixed start.
For the flexible start protocol, a GnRH antagonist is initiated when the lead follicle is 14–15 mm in diameter [13], while in the fixed protocol, GnRH antagonist is initiated on day 5–6 of the stimulation cycle. The principle underlying both approaches is that antagonist administration should commence when there is adequate follicular development but before there is a risk of premature LH elevation due to positive feedback from rising estradiol concentrations in serum. It is also possibly worth minimizing exposure to GnRH antagonist due to its possible negative effect on endometrial receptivity [14].
Four RCTs have compared fixed (day 6) with flexible (follicle diameter 14–15 mm) protocols of GnRH-antagonist administration. Current evidence does suggest a lower pregnancy rate in the flexible start protocol but the difference is not significant (odds ratio [OR]: 0.70; 95% confidence interval [CI]: 0.47–1.05). Both regimens were equally effective in preventing a premature LH surge. There was a statistically significant reduction, both in number of antagonist ampoules (OR: −1.2; 95% CI: −1.26 to −1.15) and amount of gonadotropins (OR: 95.5 IU; 95% CI: 74.8–116.1) used in the flexible protocol [13]. Larger RCTs are required to further explore this question.
Despite anxieties regarding endometrial receptivity after prolonged exposure to GnRH antagonists, early initiation of GnRH antagonists in the early follicular phase did not show any detrimental effects on follicular development or IVF outcomes [14,15]. Centers with experience in the use of GnRH antagonists have reported promising results with a fixed start and an earlier initiation of GnRH antagonist on day 5 of stimulation. This approach has the advantage of being a simple and patient-friendly protocol.
GnRH agonist versus human chorionic gonadotropin for triggering
Standard triggering (final oocyte maturation) in most GnRH agonist and antagonist assisted reproduction technology (ART) cycles is by a single HCG injection. However, HCG has a long half-life and has the propensity to stimulate ongoing growth of less mature follicles, increasing the risk of developing ovarian hyperstimulation syndrome (OHSS) [16]. Antagonist cycles provide the option of using a single GnRH agonist injection for triggering final oocyte maturation. GnRH agonists have short half-lives and could therefore decrease the risk of OHSS [17]. However, the use of GnRH agonists for triggering a GnRH antagonist cycle induces an LH surge of profile and duration that is unlike that seen in the natural cycle [18,19]. Most studies to date have shown a lower pregnancy rate with GnRH agonist triggering compared with routine HCG injection for final oocyte maturation, but lack evidence of effectiveness in OHSS prevention [20]. More studies are required to establish the optimum dose of GnRH agonist and possibly the frequency of multiple agonist injections for triggering. Other alternative management options for OHSS can be considered such as coasting or embryo cryopreservation with transfer of thawed embryos in another cycle [16,21], and again more evidence is needed to compare the efficacy of these methods.
Scheduling of GnRH antagonist treatment with oral contraceptive pill
With oral contraceptive (OC) (pre)treatment, the precycle estrogen–progestin cycle may augment the patient's subsequent response to gonadotropins, preventing premature LH surges and improving synchronization of the recruitable cohort of ovarian follicle [22]. The most important practical advantage of an antagonist protocol with OC is the fact that the treatment does not need to be synchronized with the patient's cycle. This allows the time of oocyte recovery to be coordinated and planned to occur on a weekday [23]. However, Rombauts and colleagues have highlighted the limitations of this scheduling protocol with longer duration of treatment and increased gonadotropin requirements [24]. No significant differences were observed in oocyte number, embryo quality or pregnancy outcome in three RCTs [24–26]. Although scheduling with OC in antagonist cycles can be helpful in reducing clinic opening times, further work is needed before this approach can be recommended [27].
Exogenous gonadotropin use in GnRH antagonist treatment
In current antagonist protocols, gonadotropin (FSH) treatment is started on day 2 of menses and continued through the follicular phase of the natural cycle. Endogenous gonadotropin secretion will not be suppressed at this time, as seen in agonist cycles [28]. Endogenous gonadotropin secretion may still be suppressed by administration of exogenous gonadotropin but overall it is likely that there will be a reduction in dose of exogenous gonadotropin required to achieve adequate superovulation when compared with agonist cycles. This is shown by analysis of comparative Phase III trials [29], and it appears that dose schedules may be comparable with the long protocol. There is also RCT evidence that increasing dose of exogenous gonadotropins in antagonist cycles does not result in higher pregnancy rates, although slightly more oocytes may be collected [30–32].
The serum concentration of gonadotropins does not change on initiation of GnRH antagonists [33], therefore increasing the dose of gonadotropins at GnRH antagonist initiation to compensate does not appear to be necessary and does not result in a higher probability of pregnancy [34].
An area that might merit more research is the preferred starting day of exogenous gonadotropins. The intercycle rise of endogenous FSH may allow for a later start of exogenous gonadotropins as compared with a fixed start at cycle day 2, and use of a milder ovarian stimulation protocol may result in better-quality embryos [35].
Luteinizing hormone supplementation in GnRH antagonist treatment
Extremely low circulating concentrations of LH may result in reduced live birth rates [5] and increased rate of pregnancy loss [36]. Concern over excessive suppression of endogenous LH secretion in the mid-follicular phase in GnRH antagonist cycles has prompted some to suggest LH supplementation using recombinant LH. However, RCTs to test this hypothesis have not shown benefit [23,37]. On the basis of current evidence, LH supplementation in GnRH antagonist cycles is unhelpful.
Criteria for human chorionic gonadotropin administration
Different protocols for the use of GnRH antagonists have used markedly different criteria for triggering final oocytes maturation [38]. Recent data have highlighted the need for careful optimization of timing HCG administration. Prolongation of the follicular phase may result in a higher incidence of endometrial advancement, which may, in turn, lead to lower pregnancy rates [39]. Further studies with stricter criteria for HCG triggering are required to explore the optimal timing of HCG administration.
Luteal-phase supplementation
Luteal-phase support is necessary in GnRH antagonist cycles [40–42]. Current observational studies have reported extremely low pregnancy rates in unsupported luteal phase after GnRH antagonist administration [19]. Use of GnRH antagonists allows the clinician to consider the use of GnRH agonist as an alternative to luteal support with progesterone or HCG. This approach has been reported to be as effective as conventional micronized progesterone [43] and has the potential added benefit of an enhancing effect on embryo implantation and development [44]. This is promising and requires further study. Table 1 summarizes the studies concerning GnRH antagonist protocols.
Summary of studies concerning GnRH antagonist protocols.
FSH: Follicle-stimulating hormone; GnRH: Gonadotropin-releasing hormone; HCG: Human chorionic gonadotropin; LH: Luteinizing hormone; OCP: Oral contraceptive pill; RCT: Randomized controlled trial. Modified from [54].
Extrapituitary effects of GnRH antagonist
The current knowledge of the mechanism of action by which the GnRH antagonist controls the growth and differentiation of tissues and organs is incomplete. There are concerns that the extrapituitary effect of GnRH antagonist may compromise the process of embryo development and implantation [45]. In a dose-finding study it was clearly demonstrated that high doses of GnRH antagonist will impair implantation [5]. However, recent meta-analyses did not show evidence of any effect of GnRH antagonist on the extrapituitary GnRH receptors in the ovary and endometrium at doses currently used for IVF superovulation [46,47]. Conversely, embryos cryo-preserved during cycles with high doses of GnRH antagonist have resulted in acceptable pregnancy rates after thawing [48]. This demonstrates that GnRH antagonists may not have a detrimental effect on early embryos; however, this merits further studies.
Efficacy of GnRH antagonist in IVF
Meta-analyses of studies comparing GnRH agonist long protocols with GnRH antagonist protocols have yielded conflicting results. In the Cochrane meta-analysis, the clinical efficacy of GnRH antagonists was found to be significantly lower that in GnRH agonist cycles. The conclusion drawn from analysis of RCTs reported poorer outcomes for antagonist in pregnancy rates (OR: 0.84; 95% CI: 0.72–0.97) and live birth rates (p = 0.03; OR: 0.82; 95% CI: 0.69–0.98). The study also reported a significantly shorter duration of stimulation and lower gonadotropin requirement in GnRH antagonist cycles. The incidence of OHSS was lower after the use of GnRH antagonist (p = 0.01; relative risk [RR]: 0.61; 95% CI: 0.42–0.89) [11]. However, other recent RCTs have led to different conclusions regarding the efficacy of GnRH antagonists, with no differences in pregnancy rates [35,55].
Premature meta-analysis of early trials has stimulated debate concerning the place of GnRH antagonists in IVF treatment. The meta-analysis can be criticized for comparing the gold standard long protocol, arrived at after many years of research and refinement, with early studies of antagonists carried out mainly in support of regulatory approval. The earlier studies assessed were open-label trials and could therefore be affected by bias in management and reporting. Many of the centers conducting the trials were using GnRH antagonists for the first time, before criteria for optimal use of GnRH antagonists had been established. It is clear that protocols for use of GnRH antagonists should not merely mimic those in standard use with the agonists. Current research should still focus on the more basic questions of physiology needed to define the target endocrine milieu during antagonist use. Table 2 summarizes the current major differences between GnRH agonist and GnRH antagonist treatments.
Major differences between GnRH agonist and GnRH antagonist treatment during IVF superovulation.
GnRH: Gonadotropin-releasing hormone; HCG: Human chorionic gonadotropin; IVF: In vitro fertilization; OCP: Oral contraceptive pill; OHSS: Ovarian hyperstimulation syndrome.
Specific patient subgroups suitable for the use of GnRH antagonists Polycystic ovary syndrome Most studies have excluded women with polycystic ovary syndrome (PCOS) as they have a higher propensity for OHSS during ovarian stimulation. The use of a gentler GnRH antagonist protocol may have its merits in this subgroup of patients.
Small studies on this subgroup of patients did not allow a reliable conclusion regarding IVF outcome following the use of a GnRH antagonist versus agonist to be drawn [56]. Nevertheless, results with the antagonist protocol gave results that were as good as those with the agonist. Elevated concentrations of LH are a common feature in PCOS. These may be detrimental to follicular development and ovulation. GnRH antagonists are capable of suppressing and normalizing elevated LH levels in PCOS patients but do not induce ovulation [57]. More prospective trials using GnRH antagonists with PCOS patients are required.
Poor responders
The use of prolonged downregulation with GnRH agonist in women who are poor ovarian responders may have a direct, deleterious effect on the ovary, resulting in decreased ovarian response and increased cancellation rate [58,59]. GnRH antagonists are given in the late follicular phase and will therefore not affect early follicle growth, which may be beneficial to poor responders. A small prospective pilot study with this group has suggested that GnRH antagonist use is associated with a significantly lower cancellation rate, but did not show an improvement in pregnancy rate. Other small studies and analyses have failed to confirm such a difference [60,61]. More studies are needed to evaluate this hypothesis.
Conclusion
Current clinical experience with GnRH antagonist in IVF is limited, although there are a growing number of IVF centers that have adopted the use of antagonists wholeheartedly, with promising results. We cannot hope to replace 20 years of experience with GnRH agonists overnight, and it is crucially important to establish the optimum standard for the use of antagonists in superovulation. More resources should be directed into properly powered clinical trials to explore the issues of safety and optimal pregnancy outcome in GnRH antagonist cycles.
Future perspective
With increasing experience in the use of GnRH antagonists there will be greater acceptance of GnRH antagonists in more centers. However, for antagonists to be fully integrated into the existing IVF protocols, answers to many questions need to be found. Research targets should include studies:
To optimize and refine current treatment regimens with GnRH antagonists. More studies on the direct effects of GnRH antagonists on the developing embryo and establishment of the ideal antagonist endocrine environment for successful pregnancy outcome;
To explore the possible use of GnRH antagonists in natural-cycle IVF;
To individualize treatment scheduling and protocols for patients;
To establish better scheduling and synchronization of treatment cycles without compromising overall success rates.
Executive summary
No difference between single- or multiple-dose gonadotropin-releasing hormone (GnRH) antagonist administration.
Fixed GnRH antagonist administration on the fifth or sixth day of stimulation appears to be superior to flexible initiation by a follicle diameter of 14–16mm.
Increasing follicle-stimulating hormone doses during stimulation or initiation of GnRH antagonist does not appear to be necessary.
Oral contraceptive pretreatment for cycle scheduling is feasible with no detrimental effect on pregnancy outcome, but adds complexity.
Addition of luteinizing hormone during stimulation does not appear to be necessary.
Prolonged follicular phase is associated with a lower probability of pregnancy.
Triggering with human chorionic gonadotropin is superior to GnRH agonist, but optimal timing for triggering needs to be further explored.
Luteal-phase support is essential in GnRH antagonist cycles.
Use of GnRH antagonist in mild-stimulant protocols merit further studies.
