Embryo implantation and GnRH antagonists

GnRH antagonists do not activate the GnRH receptor

Bernadette Mannaerts1,3 and Keith Gordon2

1 Organon, Oss, The Netherlands and 2 Organon Inc., West Orange, USA

Abstract

Recent suggestions that gonadotrophin-releasing hormone (GnRH) antagonists activate the GnRH receptor are discussed. Most of the studies cited in support of this suggestion are in-vitro studies, testing supra-pharmacological doses of GnRH analogues in cancer cell lines, whereas GnRH antagonists, e.g. ganirelix or cetrorelix, do not affect the steroidogenesis of human granulosa cells in vitro. In patients treated with GnRH antagonists prior to IVF or intracytoplasmic sperm injection (ICSI), oocyte maturity and fertilization rates are equal to those achieved following a long protocol of GnRH agonists. Although there is a tendency towards a lower pregnancy rate (not statistically significant) in the initial trials using GnRH antagonist with either recombinant FSH or human menopausal gonadotrophin (HMG) for ovarian stimulation, this new treatment option of GnRH antagonists facilitates short and simple treatment and improves the convenience and safety for the patient. As with GnRH agonists in the past, the clinical outcome of GnRH antagonist treatment will improve with time as more clinical experience is gained (learning curve) and the treatment protocol is optimized. Moreover, a GnRH agonist instead of human chorionic gonadotrophin (HCG) may be used for triggering ovulation and will decrease the cancellation rate and minimize the risk for developing ovarian hyperstimulation syndrome (OHSS).

Key words: GnRH agonist/GnRH antagonist/GnRH receptor

In a recent article (Hernandez, 2000Go), the author creatively weaves together the scant literature in support of his hypothesis that gonadotrophin-releasing hormone (GnRH) antagonists interact with the mitotic programming of cells involved in folliculogenesis, blastomere formation and endometrium development.

We have two major problems with the rationale presented. First, throughout the manuscript there is an underlying assumption that GnRH antagonists activate the GnRH receptor (see for example the last paragraph of the section `At the level of endometrial cell'). This assumption clearly illustrates a marked discrepancy with how most researchers in this field view the mechanism of action of GnRH antagonists. To the best of our knowledge GnRH antagonists lack intrinsic effects at the GnRH receptor level and can thus only counteract the actions of native GnRH or its agonistic analogues. A second major criticism is that most of the studies from which the author is extrapolating his mechanistic arguments are in-vitro studies, utilizing cancer cell lines, which are by definition abnormal in their regulatory mechanisms. Hernandez cites the work of Emons, Ortmann and Schally in support of this contention. Yet, those same authors have themselves stated that `Findings from our laboratory suggest that the classical LHRH receptor signal transduction mechanisms, known to operate in the pituitary are not involved in the mediation of antiproliferative effects of LHRH analogues in cancer cells' (Emons et al., 1997Go2). Is Dr Hernandez really suggesting that follicles, embryos and endometrium behave like cancer cells?

As researchers that have been involved with GnRH antagonist development for many years we share the clinicians concern that in the early trials there has been a tendency for lower pregnancy rates in patients undergoing ovarian stimulation with a GnRH antagonist regimen versus rates achieved with the long GnRH agonist protocol. However, analysis of three large phase III trials conducted with 0.25 mg ganirelix reveals that even with these large numbers of patients, the difference in pregnancy rates is not statistically significant and <5%. In addition, there are some points that the readers should be aware of.

Numerous in-vitro experiments have failed to show any significant impact of GnRH antagonists such as ganirelix or cetrorelix on steroidogenesis of rat or human granulosa cells obtained prior to or after luteinization (Ortmann et al., 1999Go; Verbost et al., 1999Go).

Oocyte quality and maturity (>80% at metaphase II) seen with GnRH antagonist protocols are comparable with those achieved after a long GnRH agonist protocol. Fertilization rates are equal in all studies to date.

The number of good quality embryos obtained was in all studies similar to that obtained with a long GnRH agonist protocol, though the short antagonist regimen revealed on average one oocyte less.

Replacement of frozen embryos obtained after antagonist treatment produces pregnancy rates comparable with those obtained after agonist treatment, regardless of the antagonist dose used. The elimination half-life for ganirelix is only 13 h, thus, following the last daily 0.25 mg dose, there will have been approximately six half-lives before embryo transfer is performed. Serum concentrations of ganirelix are very low to undetectable at the time of transfer, even with day 3 transfer.

The dose-finding study with ganirelix demonstrated that implantation rates in the high dose groups were unfavourable (Ganirelix Dose Finding Study Group, 1998). Thus we question whether the outcomes of multiple low dose protocols with 0.25 mg/day GnRH antagonist can be compared with the outcomes of the single high dose (3 mg) cetrorelix protocol.

The pattern of steroid hormone concentrations during the gonadotrophin-induced ovarian stimulation with co-administration of GnRH antagonists during the latter stages is different from that seen with the GnRH agonists and, thus, it is likely that the endometrium also develops differently. Transfers were done at the same time regardless of the analogue used and this may not have been the optimal time for transfer in the GnRH antagonist groups.

Although there is a tendency towards a lower pregnancy rate in the initial trials using either recombinant FSH or human menopausal gonadotrophin (HMG) for ovarian stimulation, this new treatment option of GnRH antagonists facilitates short and simple treatment and improves the convenience and safety for the patient. Moreover, in the most recent controlled trial (n = 337) in patients undergoing IVF or ICSI, implantation rates after treatment with 0.25 mg ganirelix were high (22.9%), and equal to those after treatment with a long protocol of triptorelin (0.1 mg/day).

We anticipate that in time the clinical outcome of ganirelix treatment can be improved by gaining more clinical experience (learning curve), by optimization of the protocol and by specific small dose adjustments. Moreover, a GnRH agonist instead of human chorionic gonadotrophin (HCG) may be used for triggering ovulation and will decrease the cancellation rate and minimize the risk for developing ovarian hyperstimulation syndrome (OHSS).

Notes

3 To whom correspondence should be addressed at: NV Organon PO Box 20, 5340 BH Oss, The Netherlands. E-mail: b.mannaerts{at}organon.oss.akzonobel.nl Back

This debate was previously published on Webtrack, June 16, 2000

References

Emons, G., Ortmann, O., Schulz, K.-D. and Schally A.V. (1997) Growth inhibitory action of analogues of luteinizing hormone releasing hormone on tumor cells. Trends Endocrinol. Metab., 8, 355–362.[ISI]

Ganirelix Dose Finding Group (1998) A double-blind, randomized, dose finding study to assess the efficacy of the GnRH antagonist Ganirelix (Org 37462) to prevent premature luteinizing hormone surges in women undergoing ovarian stimulation with recombinant follicle stimulating hormone (Puregon). Hum. Reprod., 13, 3023–3031.[Abstract]

Hernandez, E.R. (2000) Embryo implantation: the Rubicon for GnRH antagonists. Hum. Reprod., 15, 1211–1216.[Abstract/Free Full Text]

Ortmann, O., Weiss, J.M., Gürke, E.M. et al. (1999) Ovarian effects of GnRH antagonists. [Abstr. no. 102.] Gynecol. Endocrinol., 13 (Suppl.), 51.

Verbost, P.M., Smitz, J., Peddemars, C. et al. (1999) Effects of GnRH-agonist and antagonists on steroid production in rat and human granulosa cells. [Abstr. no. 101.] Gynecol. Endocrinol., 13 (Suppl.), 51.