GnRH antagonists in ovarian stimulation: a treatment regimen of clinicians’ second choice? Data from the German national IVF registry

Georg Griesinger1,3, Ricardo Felberbaum2 and Klaus Diedrich1

1 Department of Obstetrics and Gynecology, University Clinic of Schleswig-Holstein, Campus Luebeck, Luebeck and 2 Department of Obstetrics and Gynecology, Clinic Kempten-Oberallgaeu, Kempten, Germany

3 To whom correspondence should be addressed. Email: georg.griesinger{at}frauenklinik.uni-luebeck.de


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The place of GnRH antagonists in ovarian stimulation is controversial. Meta-analyses on studies comparing GnRH agonist and GnRH antagonist treatment regimens have suggested a comparatively lower efficacy of GnRH antagonists, which is likely to have influenced clinicians’ attitudes. This report describes GnRH antagonist utilization for ovarian stimulation in Germany from 2000–2003. Data from the national IVF registry were analysed. The majority of ovarian stimulation cycles are still performed in long GnRH agonist protocols, although a significant increase in GnRH antagonist usage has been noted (P < 0.0001). Two observations support the notion that GnRH antagonists are often utilized as a treatment option in cycles with an unfavourable a priori prognosis: (i) the proportion of GnRH antagonist cycles increases with cycle rank (P < 0.0001, {chi}2 for linear trend); and (ii) GnRH antagonist cycles are more often conducted in older patients as compared to GnRH agonist cycles (P < 0.0001). This has important implications for interpreting clinical performance of GnRH antagonists outside a research context.

Key words: assisted reproductive technology/GnRH antagonist/IVF/ovarian stimulation/registry


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Two GnRH antagonists, cetrorelix (Cetrotide®; Serono International S.A., Geneve, Switzerland) and ganirelix (Orgalutran®; Antagon®; Organon, Oss, The Netherlands), are currently commercially available for use in ovarian stimulation. In 1999, when the first GnRH antagonist attained market approval, it was expected that these new compounds would rapidly replace GnRH agonists in clinical practice due to several advantages: GnRH antagonists can be administered in the spontaneous cycle; side-effects resulting from hypoestrogenism caused by down-regulation with GnRH agonists are absent; and treatment time and gonadotrophin requirement are lower. Moreover, a reduced incidence of severe ovarian hyperstimulation syndrome after GnRH antagonist treatment can be expected.

However, uptake of GnRH antagonists has been slow and the vast majority of ovarian stimulation cycles are still performed in GnRH agonist long protocols (FIVNAT, 2002Go; Deutsches IVF Register, 2003Go). Meta-analyses on studies comparing GnRH agonist long protocols with GnRH antagonist protocols have yielded conflicting results as regards the likelihood of pregnancy achievement (Ludwig et al., 2001Go; Al-Inany and Aboulghar, 2002Go; Daya, 2005Go). Still, the picture drawn suggests that GnRH antagonists are slightly less efficacious than GnRH agonists in long protocols. This has stimulated an ongoing debate on the place of GnRH antagonists in infertility treatment which in turn is likely to have influenced clinician's attitudes towards GnRH antagonist usage. Herein we describe GnRH antagonist utilization for ovarian stimulation in Germany from 2000 to 2003 and present data from the national IVF registry supporting the notion that GnRH antagonists are clinically used as a second choice treatment option. The implications of this development are discussed.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Data were retrieved from the national German IVF registry. Since 1999 it has been compulsory to participate in the registry. A maximum of 116 centres reported data each year. All assisted reproduction treatment cycles (IVF) conducted in Germany are entered prospectively (within 8 days from beginning of the stimulation). The time-period analysed for the present study was from 2000 to 2003. A total of 272 862 initiated ovarian stimulation cycles in various protocols have been registered in these 4 years, which constitutes the denominator for the analyses presented herein. The data set includes information on the GnRH analogue employed. GnRH agonist cycles are further differentiated by long and short protocol cycles respectively. GnRH antagonist cycles comprise multiple and single dose protocol cycles. For individual cycles, the cycle rank is recorded and data were stratified according to cycle rank. Data were retrieved up to cycle number 10, above which the number of cycles becomes very small. As a demographic parameter, distribution of patients according to age categories was retrieved. Age refers to the patients’ age at initiation of the ovarian stimulation cycle.

Data are presented descriptively. {chi}2-Test for linear trend was used for statistical testing in contingency tables employing StatsDirect statistical software package, version 2.4.4.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
In 2000, the first year after GnRH antagonists became available, 37 230 GnRH agonist long protocol and 7821 GnRH antagonist ovarian stimulation cycles were initiated. By 2003, GnRH antagonist utilization had increased 2.89-fold to a total number of 22 614 initiated ovarian stimulation cycles. In contrast, 53 151 GnRH agonist cycles were recorded in 2003, which translates into a 1.42-fold increase of GnRH agonist long protocol utilization from 2000 to 2003 (Figure 1). Whereas in the year 2000 about one in seven ovarian stimulation cycles was conducted using GnRH antagonists, in 2003 about one in four cycles was conducted using GnRH antagonists (P < 0.0001).



View larger version (18K):
[in this window]
[in a new window]
 
Figure 1. Initiated ovarian stimulation cycles per year (numbers on columns represent proportions).

 
Stratified by cycle rank, the proportion of GnRH antagonist cycles increases from 23.3%in first treatment cycles, to 35.5% in fifth treatment cycles, and to 47.8% in tenth treatment cycles (P < 0.0001; Figure 2). The same trend was observed for stimulation protocols other than GnRH agonist long or GnRH antagonist (e.g. GnRH agonist short, or stimulation regimen without GnRH analogue; data not shown). From a total of 272 862 cycles, 117 854 (40.2%) were conducted in a first treatment cycle. With higher cycle ranks, the number of conducted cycles steadily decreases (data not shown).



View larger version (19K):
[in this window]
[in a new window]
 
Figure 2. Proportion of long GnRH agonist and GnRH antagonist cycles stratified by cycle rank (numbers on columns represent proportions of GnRH antagonist cycles).

 
Table I shows distribution of patients according to age categories in GnRH agonist long and GnRH antagonist cycles respectively. The proportions of patients in age categories in different columns vary from row to row, which was statistically significant by total {chi}2-test (P < 0.0001). Furthermore, a significant linear trend was found (P < 0.0001), suggesting that the number of patients decreases with age differently in GnRH agonist long and GnRH antagonist protocols. This was found for pooled data (2000–2003), as well as when individual years were analysed (data not shown).


View this table:
[in this window]
[in a new window]
 
Table I. Proportion of patients in different age categories being treated in either GnRH agonist long or GnRH antagonist protocols from 2000 to 2003 ({chi}2-test for linear trend)

 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
GnRH antagonists have increasingly gained ground in clinical use in recent years. Data from the German registry are similar to data from the French registry, which show an increase in the proportion of GnRH antagonist cycles from 0.6% in 1999 to 19.6% and 24.2% in 2002, for IVF and ICSI respectively (FIVNAT, 2002Go). Although GnRH antagonists have obviously found acceptance by clinicians, the place of GnRH antagonists remains controversial.

In this analysis we demonstrate patient selection in favour of GnRH agonist protocols for both age and cycle rank. It is well established that younger patients as well as patients in lower treatment ranks have a comparatively higher chance of achieving pregnancy than older patients or patients in higher treatment cycle ranks. The data from the German registry thus suggest that GnRH antagonists are comparatively more often used in cycles which have an unfavourable a priori prognosis. It appears likely that the majority of clinicians favour the well-established GnRH agonist long protocol as a first choice for ovarian stimulation, whereas for patients who failed to become pregnant, or patients of older age, GnRH antagonist protocols are more often employed. Inherent to this approach is a danger of a ‘self-fulfilling prophecy’ as regards pregnancy rates in the perception of the clinician. Presumably, clinicians use other variables, such as body mass index, presence of polycystic ovaries, risk of ovarian hyperstimulation, presence of endometriosis, cycle irregularities etc., to primarily select for either GnRH agonist or GnRH antagonist protocols. Here we report only two variables, age and cycle rank, that could be unambiguously and completely retrieved from the registry. Since older patients are more prone to a decreased ovarian response to exogenous gonadotrophins, it may appear rational to clinicans to use GnRH antagonist protocols to avoid ovarian suppression at the stage of follicular recruitment. However, little evidence exists to support the notion that GnRH antagonists are superior in older patients or patients with a decreased ovarian response (Tarlatzis et al., 2003Go) as compared to GnRH agonist-based treatment regimen.

In the annual report from the Deutsches IVF Register (2003)Go, clinical pregnancy achievement per embryo transfer is recorded as 25.75% for GnRH antagonist protocol cycles as compared to 30.52% for GnRH agonist long protocol cycles. The size of effect of an unfavourable patient selection on the pregnancy rate in the GnRH antagonist treatment cycles as suggested by the present analysis is difficult to estimate, but it has to be assumed that such an effect exists. In support of this assumption is a subanalysis in the registry report on women with tubal infertility, first treatment cycles, and age <35 years, where the pregnancy rate was 37.83 and 36.07% for the GnRH agonist long protocol and GnRH antagonist protocol respectively. However, when pooled data from an annual statistic are interupted, it must be taken into consideration that a comparatively small number of patients are treated in higher cycle ranks. Moreover, not all patients that are recorded as being treated in a higher cycle rank in the German registry necessarily have a poor prognosis, because patients have their treatment cycles chronologically counted, irrespective of pregnancy achievement or live birth in a previous cycle. Obviously, only comparative controlled trials allow a conclusion on the efficacy of the two protocols. Thus, we herein refrained from reporting pregnancy.

The authors are aware of three independent meta-analyses that compare GnRH antagonist protocols to the established GnRH agonist treatment regimen. Al-Inany and Aboulghar (2002)Go pooled data from five prospective, randomized controlled trials and found an OR of 0.79 (95% CI 0.63–0.99) for clinical pregnancy rate per woman in favour of the GnRH agonist long protocol. This meta-analysis has recently been updated and confirmed the finding of a lower pregnancy rate in GnRH antagonist cycles (Aboulghar, 2004Go). The meta-analysis by Ludwig et al. (2001)Go reported a trend towards a lower clinical pregnancy rate per woman randomized (OR 0.85; 95% CI 0.70–1.03) with a confidence interval that crossed the unity. Another recent meta-analysis by Daya (2005)Go including 5000 treatment cycles on an intention-to-treat basis found an OR of 0.87 (95% CI 0.77–0.99) for clinical pregnancy in GnRH antagonist cycles; however, this marginal effect statistically vanished when only ongoing pregnancies were considered.

Due to the economic pressure on the side of the patient and the competition between centres, reproductive medicine has become dominated by the outcome parameter pregnancy rate. Pregnancy achievement without consideration of side-effects of the treatment can no longer be considered the objective of assisted reproduction. For the reduction of multiple gestations, a clinical trend towards elective single embryo transfer exists, although this will be associated for some patients with a decreased likelihood of pregnancy achievement per cycle (or embryo transfer) and the necessity to undergo comparatively more treatment (or embryo transfer) cycles. For ovarian stimulation, no consensus exists as yet on the endpoint measure that defines success.

For ovarian stimulation we advocate the balance of benefits and drawbacks of individual stimulation regimens. GnRH antagonists have helped to overcome some major disadvantages of GnRH agonist-based stimulation protocols, especially of the long protocol, resulting in safe and patient-friendly ovarian stimulation that does not deserve the place of a second choice treatment option. With GnRH antagonists, LH surges can reliably be prevented; there is absence of the flare-up effect, thus treatment time is considerably shorter and ovarian cysts are rare; no symptoms due to hypoestrogenism occur; there is no inadvertent administration of the GnRH analogue in early pregnancy; gonadotrophin consumption is lower and ovarian stimulation is softer; a trend in favour of GnRH antagonists is present for the incidence of severe OHSS (Al-Inany and Aboulghar, 2002Go). Furthermore, in our clinical experience, chemical castration by means of a GnRH agonist long protocol is emotionally disturbing for women who try to conceive. To fit the organizational regimen of IVF units, GnRH antagonist cycles can also reliably be programmed with an oral contraceptive pill (Fischl et al., 2001Go).

In conclusion, GnRH antagonists offer a number of clinical advantages, but are apparently often not employed as a first line treatment. The pregnancy rates achieved are reduced as compared to GnRH agonist long protocol cycles in both controlled studies and observational registry data. However, in registry reports the effect of an unfavourable patient selection for GnRH antagonists has to be taken into consideration. We advocate that GnRH antagonist protocols deserve optimization rather than second place.


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Aboulghar MA (2004) Meta-analysis of RCTs comparing GnRH-agonist and GnRH-antagonist. Abstract of the ESHRE Campus symposium 2004 on GnRH-antagonist in IVF, p 13.

Al-Inany H and Aboulghar M (2002) GnRH antagonist in assisted reproduction: a Cochrane review. Hum Reprod 17, 874–885.[Abstract/Free Full Text]

Daya S (2005) GnRH-agonists versus antagonists in assisted reproduction—a systematic review. Abstract Book of the 8th International Symposium on GnRH-analogues in Cancer and Human Reproduction, A100.

Deutsches IVF Register (2003) Annual report from the German IVF registry 2003 (http://www.deutsches-ivf-register.de).

Fischl F, Huber JC and Obruca A (2001) Zeitliche Optimierung der kontrollierten Hyperstimulation (KOH) in Kombination mit GnrH-Antagonisten und Ovulationshemmer in einem IVF-Programm. J Fertilität Reprod 11, 50–51.

FIVNAT (2002) Annual report from the French IVF registry. Bilan provisoire 2002 (http://www.perso.wanadoo.fr/fivnat.fr).

Ludwig M, Katalinic A and Diedrich K (2001) Use of GnRH antagonists in ovarian stimulation for assisted reproductive technologies compared to the long protocol Meta-analysis. Arch Gynecol Obstet 265, 175–182.[CrossRef][Medline]

Tarlatzis BC, Zepiridis L, Grimbizis G and Bontis J (2003) Clinical management of low ovarian response to stimulation for IVF: a systematic review. Hum Reprod Update 9, 61–76.[Abstract/Free Full Text]

Submitted on March 3, 2005; resubmitted on April 10, 2005; accepted on April 18, 2005.