Abstract. Introduction. RBMOnline - Vol 7. No Reproductive BioMedicine Online; on web 16 July 2003

Transcription

1 RBMOnline - Vol 7. No Reproductive BioMedicine Online; on web 16 July 2003 Article Comparable effectiveness using flexible singledose GnRH antagonist (cetrorelix) and singledose long GnRH agonist (goserelin) protocol for IVF cycles a prospective, randomized study Dr Veljko Vlaisavljevic Veljko Vlaisavljevic attended the Medical School in Ljubljana, Slovenia. After completing his residency in Obstetrics and Gynecology, he began his career at the Department of Obstetrics and Gynecology in Maribor, Slovenia. He then studied for his PhD in Andrology at the University of Zagreb, Croatia and completed his education in reproduction at RWH in Melbourne, Australia. Now he is a Professor in the Department of Reproductive Medicine at the Maribor Teaching Hospital. At present, he is President of the Slovene Society for Ultrasound in Medicine, Vice-President of the Slovene Society of Reproductive Medicine and member of the Advisory Committee of ESHRE. His clinical practice focuses on infertility, assisted reproductive technology and research. His research interest is focused on ultrasound monitoring of follicle growth and perifollicular vascularization in natural cycles related to oocyte maturation, oocyte quality and IVF outcome. Veljko Vlaisavljevic 1, Milan Reljic, Vida Gavric Lovrec, Borut Kovacic Department of Reproductive Medicine and Gynaecologic Endocrinology, Maribor Teaching Hospital, Ljubljanska 5, SI-2000 Maribor, Slovenia 1 Correspondence: vlai@sb-mb.si Abstract This prospective randomized study compared the effectiveness of a flexible single-dose gonadotrophin-releasing hormone (GnRH) antagonist (cetrorelix) and a single-dose long GnRH agonist (goserelin) protocol for ovarian stimulation in IVF/intracytoplasmic sperm injection (ICSI) cycles. All patients from the waiting list were successively included in the study, pre-programmed with an oral contraceptive, and randomized into goserelin and cetrorelix groups. Depending on the date on which their menstrual period started, patients took oral contraceptives for one or two cycles. Ultimately, 236 patients in the first group received a single dose of depot preparation of goserelin and 224 patients received a single 3 mg dose of cetrorelix in the late follicular phase, when the mean follicle diameter exceeded 12 mm. The mean number of ampoules of FSH and the duration of stimulation was statistically significantly lower in the cetrorelix group than in the goserelin group (25.9 versus 34.5, and 9.6 versus 12.2 days, P < 0.01). The mean number of oocytes retrieved was similar (6.7 ± 4.5 versus 7.2 ± 4.6, NS). Similar results were observed in fertilization rates, blastulation rates and blastocyst transfer rates in both groups. Clinical pregnancy and delivery rates per cycle were higher in the goserelin group (34.3 and 30.1%) than in the cetrorelix group (31.9 and 28.3%), but the differences were not statistically significant. The flexible single-dose GnRH antagonist protocol is an advantageous alternative to the long GnRH agonist protocol, with similar efficacy, shorter duration, a significant reduction in the number of FSH ampoules used and without the menopause-like effects of the GnRH antagonist. Keywords: cetrorelix, flexible protocol, GnRH antagonist, IVF embryo transfer, single dose Introduction Gonadotrophin-releasing hormone (GnRH) agonists were introduced to assisted reproductive technologies (ART) about 20 years ago (Ludwig et al., 2001). Today, the so-called long GnRH protocols for ovarian stimulation (OS) have proved to be the most successful and are the most frequently used (Olivennes et al., 2000). GnRH antagonists, which have recently been introduced into protocols for OS, offer certain advantages over GnRH agonists (Devroey, 2000). GnRH antagonists such as cetrorelix (Cetrotide; Serono International S.A., Switzerland) block the pituitary GnRH receptors competitively and their administration during the follicular phase of the cycle can prevent or interrupt the LH rise (Nikolettos et al., 2001). The major cause of interest in GnRH antagonists is the absence of the flare-up effect in gonadotrophin secretion as well as the almost immediate effect of the desensitization period following antagonist administration (Olivennes et al., 2000). This mode of action, being different from that of agonists, offers some advantages. The application of an antagonist is 301

2 302 scheduled from the middle of the follicular phase of the OS cycle. The length of stimulation, the amount of gonadotrophins used for ovarian stimulation and the incidence of ovarian hyperstimulation syndrome (OHSS) are reduced (Albano et al., 2000; Ludwig et al., 2001; Nikolettos et al., 2001). Other advantages include the possibility of triggering ovulation with native GnRH or GnRH agonists and the possibility of using antagonists in minimal-stimulation or natural cycles (Bouchard and Fauser, 2000). There are also no side effects related to hormonal depletion, such as hot flushes, bleeding and vaginal dryness (Olivennes et al., 2000). On the other hand, there is an ongoing discussion as to whether there is a significant reduction in pregnancy rates when GnRH antagonists are used, since it has been reported in several studies that pregnancy rates are lower in antagonist than in long GnRH agonist protocols (Ludwig et al., 2001; Al-Inany and Aboulghar, 2002). In almost all of these studies, the singleor multiple-dose antagonists were used in a so-called fixed protocol. This means that an antagonist was injected on a certain day of the OS cycle, not taking into account individual cycle characteristics. According to some authors, the clinical outcome of GnRH antagonist cycles may be further improved by developing a more flexible regimen (Kol, 2000; Ludwig et al., 2001; Al-Inany and Aboulghar, 2002). The aim of this prospective randomized study was to compare the efficacy of the GnRH antagonist cetrorelix in a flexible single-dose regimen with that of the depot preparation GnRH agonist goserelin (Zoladex 3.6 mg; Astra-Zeneca Pharmaceuticals, Cheshire, UK) in a standard long protocol for OS in IVF/intracytoplasmic sperm injection (ICSI) cycles. A comparison of the pregnancy rate and delivery rate per cycle between the depot GnRH agonist goserelin and subcutaneous daily application of GnRH agonist did not show any difference (Vlaisavljevic et al., 2000) Material and methods This study was prospective and included 462 consecutive patients undergoing IVF or ICSI in ovarian stimulation cycles between August 1999 and December The inclusion criteria were as follows: 18- to 45-year-old patients requiring infertility treatment by OS and IVF embryo transfer with or without ICSI and with normal menstrual cycles, FSH concentrations <15 IU/l on day 2 or day 3 and no more than three cycles of ART. Patients with PCOS, severe endometriosis, known poor responders, smokers and patients over 40 were not excluded from the study. Patients were included in the study successively from the waiting list for IVF/ICSI procedure at the Department of Reproductive Medicine at Maribor Teaching Hospital. All patients from the waiting list received the same invitation letter and programme for synchronization of their menstrual period with oral contraceptives. Randomization was done 9 11 weeks in advance on the basis of the menstrual period in two groups. The patients got the list of dates in the next 9 11 weeks and instructions about when to start with oral contraceptives, depending on the date of commencement of their menstrual period. Depending on the date of inclusion in the synchronization programme, patients took oral contraceptives for one or two cycles, but all who were randomized in the same group finished oral contraceptives on the same day. The patients who got their period in the first half of the month after they had taken the last pill were included in the same group and started with the same protocol regimen. There were only two groups of patients per month (2 3 weeks apart from each other). This makes decisions during cycle monitoring more uniform for the whole group, with no influence of the previous protocol characteristics on decision-making. The groups of patients were treated with the agonist or antagonist protocol one after another. All patients gave informed consent to participate in the study, which was approved by the local Ethics Committee. Programming the cycle All patients started with oral contraceptives containing 0.05 mg ethinyloestradiol and 0.5 mg norgestrel (Stediril; Krka, Novo Mesto, Slovenia) between day 2 and day 7 of menstrual bleeding. The last day of oral contraceptive administration was fixed for the whole group of patients. The patients in the same group were treated with oral contraceptives for a minimum of 18 and a maximum of 35 days. Some of the patients had oral contraceptives for two cycles. The reason for this was the wish to programme their ART procedure weeks in advance. All patients had their first ultrasound scan on the day of the last contraceptive pill before the menstrual period started. The final oral contraceptive dose was taken on a Friday, enabling the beginning of stimulation on a Wednesday for the cetrorelix group and on a Friday for the goserelin group of patients. The schedule is shown in Figure 1. Goserelin group Seven days before the last oral contraceptive pill was taken, a single dose of depot goserelin was administered. Stimulation with gonadotrophins in this group began 7 days after the last oral contraceptive pill. Recombinant (Gonal-F; Serono, Geneva, Switzerland) or highly purified urinary FSH (Metrodin HP; Serono) was administered subcutaneously in the thigh or abdominal region. The FSH dosage was adjusted for 7 days in advance. On the first 3 days, 150 or 225 IU of FSH was administered daily. The dose was then increased by one ampoule until day 8, when the first folliculometry was performed. After that the daily FSH dose and ultrasound monitoring were adjusted to the ovarian response. Figure 1. Adjusting the menstrual cycle for treatment at a specific time.

3 Cetrorelix group Stimulation with gonadotrophins began 5 days after the last oral contraceptive pill (Figure 2). Recombinant (Gonal-F) or highly purified urinary FSH (Metrodin HP) was administered subcutaneously. On the first 3 days, 150 or 225 IU of FSH was administered daily. The dose was then increased by one ampoule until day 5 of stimulation when the first folliculometry was performed. After that, the daily FSH dose was adjusted according to the response of the ovaries to stimulation. Ultrasound monitoring of follicular development started from day 5 of stimulation and continued regularly every day or every second day, depending on the mean follicle diameter and follicle growth rate. From the day when the mean follicle diameter passed 10 mm, folliculometry was advised more frequently. If the increase in mean follicular diameter between 2 consecutive days was <2 mm/day, then ultrasound was scheduled every second day, if the increase was 2 mm/day or more, then folliculometry was done on the next day. The criteria for subcutaneous administration of cetrorelix were fulfilled when the mean follicle diameter passed mm and follicular growth was confirmed between two measurements. An additional vial of gonadotrophin was added to the stimulation regimen on the day of cetrorelix administration. In fast responders, cetrorelix was administered on the day when the mean follicle diameter was 12mm or larger. In slow responders, cetrorelix was administered when the mean follicular diameter passed 14 mm. If the criteria for human chorionic gonadotrophin (HCG) administration were not met within 4 days of administration of the single dose of 3 mg of cetrorelix, daily administration of 0.25 mg cetrorelix was carried out up to and including the day of HCG administration. On the day on which folliculometry ascertained that the mean diameter of the dominant follicle had reached >17 mm (two largest diameters perpendicular to each other), the time for HCG (Profasi; Serono, Aubonne, Switzerland) injection was set. All patients received HCG in doses of 10,000 IU subcutaneously by self-administration. On the morning of the day of HCG administration, the LH surge was tested. Follicle aspiration was performed h after HCG administration. After retrieval, the oocytes for IVF were inseminated and cultured in IVF media (MediCult, Jyllinge, Denmark). The oocytes for the ICSI procedure were exposed to 80 IU/ml of Figure 2. Cetrorelix flexible single dose protocol used in the reported study (begun in Maribor in 1999). hyaluronidase for s (MediCult). Those oocytes that were in metaphase II (MII) were injected immediately after the procedure. Sperm injection was performed after sperm immobilization using polyvinylpirrolidone (PVP) (MediCult). ICSI was performed on an inverted microscope (Olympus, Tokyo, Japan) and Narishige micromanipulators with micropipettes (Hunter Scientific, Essex, UK). All embryos were cultured for 5 days in sequential BlastAssist System (MediCult) before blastocyst stage embryo transfer. After fertilization was assessed, the embryo was placed in a new droplet of B1 medium covered with liquid paraffin (MediCult). On day 2, embryo quality was evaluated as described previously (Kovacic et al., 2002). On day 3 of culture, the embryo was transferred to a droplet of B2 medium for a further 48 h of development. Embryos of the same quality were co-cultured in the same droplet of the medium. A maximum of two embryos, available among those that attained the blastocyst or morule stage on day 5, were transferred. The embryo transfer was performed directly from the culture medium covered with liquid paraffin (MediCult) during the embryo transfer procedure (Kovacic et al., 2002). In all transfers, a minimum of one and a maximum of two expanded blastocysts were transferred. In a small number of blastocyst transfers, one compact morula was added to the single expanded blastocyst. The luteal phase was supported in both groups of patients with 600 mg of vaginal progesterone (Utrogestan; Laboratoires Besins Iscovesco, Paris, France) until the confirmation of pregnancy. Implantation was confirmed by determination of serum β-hcg 16 days after oocyte retrieval. Pregnancy was confirmed as clinical only in case of ultrasonic evidence of a gestation sac. Luteal support was the same for all pregnant patients in both groups. All pregnant women continued with luteal supplementation with a daily dose of 30 mg didrogesterone (Dabroston; Solvay Pharmaceuticals, Weesp, The Netherlands) until week 8 of pregnancy. The data were processed using the statistical program by Statistica (StatSoft, USA). Analysis of discrete variables was done by chi-square analysis. Student s t-test was used for parametric analysis. Results The prospective randomized study included 462 stimulated IVF/ICSI cycles. A single dose of cetrorelix was received by 226 patients (cetrorelix group) and a single dose of goserelin was received by 236 patients (goserelin group). The mean day of the first cetrorelix administration was 7.7 ± 1.7 (day 5 to day 13). One of the women in the cetrorelix group was given an additional dose of 0.25 mg of cetrorelix. None of the 236 patients from the goserelin group experienced an LH surge before HCG administration. In the cetrorelix group, two patients had positive LH on the day of HCG administration. Oocyte recovery was carried out in both patients and one became pregnant. 303

4 The patients and cycles characteristics are presented in Table 1. Both groups were balanced for patient age and rate of IVF and ICSI cycles. Two patients in the cetrorelix group and 11 in the goserelin group did not undergo oocyte retrieval because an impaired ovarian response was observed. Oocyte retrieval was successfully performed in 223 cycles in both groups. The mean length of stimulation (9.6 ± 1.8 versus 12.2 ± 4.4, P < 0.01) and the mean number of ampoules (25.9 ± 9.1 versus 34.5 ± 8.7, P < 0.01) administered was significantly lower in the cetrorelix group. The mean number of oocytes was lower in the cetrorelix than in the goserelin group, but this difference was not statistically significant (6.7 ± 4.5 versus 7.2 ± 4.6, not significant). The fertilization rate, abnormal fertilization rate and number of cycles without fertilization were similar in the two groups. The number of good and poor quality embryos was also not different in the two groups. No differences were observed when the numbers of blastocysts and morules on the day of embryo transfer, the total number of blastocysts transferred, the number of cycles with embryo transfer and the mean number of embryos per transfer were compared (Table 2). The clinical outcomes are presented in Table 3. The pregnancy, delivery and miscarriage rates were statistically comparable in the cetrorelix and goserelin groups. There was a trend toward a lower number of babies born per replaced embryos in the cetrorelix group (26.8 versus 30.3%, not significant), although the differences did not reach statistical significance. The clinical results related to the period of time were analysed for each consecutive group of 50 patients in the cetrorelix group. The aim of the analysis was to confirm whether a learning curve exists in such a study design (Table 4). Discussion All patients included in this study were pretreated with oral contraceptives for a minimum of 18 days and a maximum of 35 days in one cycle. Usually such an approach was popularized as a method of avoiding oocyte retrievals during weekends. The incorporation of OC pretreatment into the OS protocol involving cetrorelix and rfsh offers the prospect of increased control over the workload for IVF clinics, together with increased convenience for the patients (Howles, 2002). The objective of this study was to randomize patients in two separate groups per month, which offers a more uniform monitoring procedure for each type of stimulation protocol. In the literature, two basic protocols for the use of the GnRH antagonist cetrorelix can be found. In the multiple-dose protocol, small doses of the GnRH antagonist are injected from the middle of the follicular phase until HCG administration (Diedrich et al., 1994; Albano et al., 1997). Consecutive low daily doses of 0.25 mg of the antagonist cetrorelix are administered from day 5 or day 6 of stimulation. The single-dose protocol envisages the administration of a higher dose injected later in the follicular phase, when the LH surge is most likely to occur unexpectedly. Effective suppression depends on the applied dose of the drug. The administration of 3 mg of cetrorelix was scheduled mostly on day 7 of stimulation and selected as a safe choice with a protection period of at least 4 days (Olivennes et al., 1994, 1998, 2000; Leroy et al., 1995) The patients must receive an additional 0.25 mg of cetrorelix daily if HCG is not administered in 4 days (96 h) after the 3 mg injection of cetrorelix. Both protocols were effective in preventing premature LH surges, but there may be differences in ovarian responses Table 1. Patients and cycles characteristics in the two groups of patients. Cetrorelix group Goserelin group P-value Started cycles Age a (min max) 32.9 ± 4.9 (22 44) 32.9 ± 4.7 (23 43) NS >40 (%) 14 (6.2) 14 (5.9) NS IVF/ICSI rate NS rfsh/ufsh-hp rate NS No. of patients undergoing 224 (99.1) 225 (95.3) NS oocyte retrieval (%) No. of cycles with oocytes 223 (98.7) 223 (94.5) NS (%) No. of FSH ampoules a 25.9 ± ± 8.7 <0.01 Stimulation length (days) a 9.6 ± ± 4.4 <0.01 Administration of GnRH 7.7 ± 1.7, day 13.8 ± 0.2, antagonist/gnrh agonist b 5 day 13 day 14 (min max) No. of oocytes (mean ± SD) 1493 (6.7 ± 4.5) 1629 (7.2 ± 4.6) NS No. of severe OHSS (%) 4 (1.7) 7 (2.9) 304 a Values are means ± SD. b Related to the first day of stimulation. NS = not statistically significant.

5 Table 2. Fertilization and cleavage outcome in the two groups of patients. Cetrorelix group Goserelin group P-value No. of normal (2PN) fertilized 964 (64.6) 1004 (61.6) NS oocytes (% of all oocytes) No. of abnormal fertilized 71 (4.7) 69 (4.2) NS oocytes (% of all oocytes) No. of cycles without fertilization 14 (6.3) 9 (4.0) NS (% of cycles with oocytes) No. of good quality embryos 777 (80.6) 845 (84.2) NS (G1 + G2) on day 2 (%) No. of poor quality embryos 187 (19.4) 143 (14.5) NS (G2 + G3) on day 2 (%) No. of blastocysts and morules 577 (58.8) 480 (47.8) NS on day 5 (% 2PN embryos) No. of cycles without blastocysts 44 (19.7) 38 (17.0) NS (% of cycles with oocytes) No. of cycles with embryo transfer 165 (74.0) 176 (78.9) NS (% of cycles with oocytes) Blastocysts frozen No. of blastocysts/morulae NS transferred Embryo transfer One compact morula 3 8 One expanded blastocyst Two compact morulae 1 3 One compact morula and one 11 5 expanded blastocyst Two expanded blastocysts NS Mean number of embryos 1.8 ± ± 0.5 NS per transfer a a Values are means ± SD. NS = not statistically significant. Table 3. Clinical outcome (pregnancies) in the two groups of patients. Cetrorelix group Goserelin group P-value Implantation confirmed with 80 (35.4) 84 (35.6) NS HCG (% per started cycles) No. of clinical pregnancies 72 (31.9) 81 (34.3) NS (% per started cycles) No. of ectopic pregnancies 2 1 NS No. of clinical abortions (%) 6 (8.3) 9 (10.8) NS No. of implanted embryos (%) 95 (32.6) 107 (35.6) NS No. of multiple implantations (%) 22 (28.2) 23 (27.7) NS No. of deliveries NS Singleton NS Twins (%) 12 (18.7) 18 (25.3) NS Triplets (%) a 1 (1.5) 1 (1.4) NS No. of babies born 78 (26.8) 91 (30.3) NS (% of embryos replaced) Delivery rate per transfer NS Delivery rate per cycle NS NS = not statistically significant. a Singleton and one set of monozygotic twins. 305

6 Table 4. The influence of the experience with Cetrorelix protocol on clinical outcome in 211 consecutive cycles. Consecutive oocyte retrieval cycles P-value Oocyte retrieval cycles (n) Mean no. of oocytes a 5.9 ± ± ± ± 5.1 Fertilization rate (%) NS Blastocyst transfers (n) Blastocysts transferred (n) Blastocysts implanted (n) Pregnancies (n) Deliveries (n) Pregnancy per transfer (%) NS Take home baby rate (%) NS a Values are means ± SD. NS = not statistically significant. 306 between the two protocols (Hung Yu and Chung Ho, 2001). The multiple-dose regimen permits deeper suppression of LH that can eventually produce lower plasma oestradiol concentrations and, as a consequence, reduced endometrial quality when gonadotrophin preparations devoid of LH are used (Bouchard and Fauser, 2000). It has also been demonstrated that with the single 3 mg cetrorelix protocol, serum LH concentrations drop in around 40% of patients (Ludwig et al., 2001). If the antagonist is administered on the fixed-day regimen at a time when follicle development is still at an early stage of growth and when using recombinant FSH alone, this drop in LH levels may lead to a disturbance in follicle maturation. Therefore a more flexible regimen of cetrorelix and FSH might be a better approach than the use of a fixed-day protocol. This experiment used a modification of the approach of Oliviennes et al. (1994), so that in the current protocol, the GnRH antagonist was not administered on the fixed day 6 or day 7 of stimulation. Cetrorelix treatment was started when the leading follicle exceeded 12 mm in diameter and was identified as growing. This procedure ensured that the cetrorelix was not administered at an early stage of growth and that the criteria for HCG administration were fulfilled in the next 4 days, so an additional administration of cetrorelix was generally unnecessary. This was the case in only four patients (1.7%), although the interval when cetrorelix was administered was wide, between day 5 in fast responders and day 13 of stimulation in slow responders. Scheduling the day of cetrorelix administration by the size of the leading follicle improves the outcome of cycles (Ludwig et al., 2002a,b). The cost of ovarian stimulation can be further reduced if the factor of confirming growing is taken into account. Administration of cetrorelix 3 mg is more appropriate because there is no need for an additional ampoule of cetrorelix 0.25 mg before the follicles reach the appropriate size for HCG administration. On the other hand, the disadvantage of such an approach, in contrast to the agonist protocol, is the difficult management of the non-uniform cohort of follicles, where the large leading follicle is the one that determines the moment of agonist administration, despite the fact that others in the cohort are too small to reach maturity within the next 4 days. In this prospective randomized study, the numbers of patients reaching the day of HCG, fertilization rate, number of good quality embryos, blastulation rate and number of cycles with embryo transfer were similar in the cetrorelix and the goserelin groups. Therefore, these results, as well as the results of other authors (Ganirelix Dose Finding Study Group, 1998), do not support the hypothesis that GnRH antagonists interact with the mitotic programming of cells involved in folliculogenesis and blastomere formation (Hernandez, 2000), and the hypothesis that the GnRH antagonist interferes with the mechanisms involved in germinal vesicle breakdown as well as the cell signalling pathway driving the oocyte into metaphase II (De la Fuente, 1999). Regarding the results of blastocyst transfer in cycles with only one oocyte retrieved (Vlaisavljevic et al., 2001) and in poor responders (Kovacic et al., 2002), it was decided to use blastocyst transfer as routine for all patients. As reported by many other authors (Albano et al., 2000; Olivennes et al., 2000), there was an advantage in reducing the number of FSH ampoules and the duration of stimulation in the cetrorelix group. In this group, the mean number of oocytes was also lower than in the goserelin group, which is known to be related to a lower risk for OHSS (Reljic et al., 1999). The number of severe cases of OHSS was lower in the cetrorelix group (1.7 versus 2.9%), but the difference in the present study was not statistically significant. Results from multicentric studies stipulate that in the IVF centres with no experience with the antagonist, pregnancy rates were significantly lower than when the agonist was used. In most GnRH antagonist studies, there is a trend towards somewhat (but not significantly) reduced pregnancy rates. This has been attributed to the learning curve, which is the consequence of the differences in monitoring and laboratory work (Barri et al., 2002). In the centres that were experienced in antagonist usage, results were in favour of antagonist versus agonist (Fluker et al., 2002). Sequential analysis of the results of this study did not confirm the hypothesis of the learning curve for antagonists. In contrast to the agonist protocol, antagonist protocols probably need a more individual approach, taking the follicle growth rate into account as an individual characteristic of patient cycle before the administration of GnRH antagonists and more frequent ultrasonographic monitoring of follicle growth to obtain similar results.

7 In the present study, the pregnancy rates and delivery rates were high and very similar in both groups. Comparison of clinical pregnancy rates per attempt between different studies where antagonists (cetrorelix, ganirelix) were compared with agonists (goserelin, triptorelin buserelin, leuprolide and triptorelin) showed that the differences between different studies are bigger than those between different analogues used (Felderbaum and Diedrich, 2002). The same characteristics were confirmed when implantation rates per embryo replaced were analysed. A meta-analysis of five randomized trials shows an overall significantly lower pregnancy rate of 5% in the GnRH antagonist regimen than in the GnRH agonist regimen (Huirne and Lambalk, 2001). Unfortunately, this meta-analysis included studies that compared a single-dose analogue regimen with a multiple-dose cetrorelix and ganirelix regimen. It is therefore difficult to compare the results of this analysis with those of the present study, particularly because Ludwig et al. demonstrate with their meta-analysis that cetrorelix and ganirelix have to be analysed separately (Ludwig et al., 2001). The reason is that cetrorelix and not ganirelix will result in the same pregnancy rate as the long protocol. The studies comparing the cetrorelix and the long agonist protocol did not demonstrate statistically different results for the pregnancy rate per cycle (OR 0.91; 95% CI ). In this metaanalysis, the studies using cetrorelix showed lower (OR = 0.85) but not significantly different clinical pregnancy and ongoing pregnancy rates compared with the long agonist protocol. Two studies mentioned in this analysis report the use of the single-dose cetrorelix and long agonist. Equivalent results (OR = 1) in the antagonist and agonist groups were demonstrated only in the study using a more flexible agonist protocol. In the study of Olivennes et al., which was not included in this meta-analysis, a single antagonist protocol with administration of cetrorelix on day 7 was compared with a long agonist protocol (Olivennes et al., 2000). The trend towards a slightly lower pregnancy rate was observed in the cetrorelix group. In the present study, very similar pregnancy rates and only slightly lower delivery rates were found. However, it is impossible to conclude from these results that the flexible single-dose antagonist protocol is better than the fixed single-dose antagonist protocol. Only a prospective randomized study comparing these two antagonist protocols can answer this question. It can be concluded that the flexible single-dose antagonist protocol gave similar clinical results to the long agonist protocol, but the antagonist protocol is more patient friendly. Such an approach can additionally improve ovarian stimulation for assisted reproduction. It offers similar efficacy to the long agonist regimen, but at the same time the cetrorelix regimen has numerous advantages: it reduces the duration of the treatment cycle and reduces the number of ampoules of gonadotrophin in each treatment cycle. Using the antagonists in ovarian stimulation, oestrogen deprivation symptoms associated with agonist-induced down-regulation can be avoided, but at the same time the cetrorelix protocol, in combination with oral contraceptive synchronization, provides the same flexibility in programming the cycle. In both groups of patients included in this study, luteal supplementation with vaginal progesterone was advised, although the effect of GnRH antagonists on human steroidogenesis still remains unclear. The majority of studies on agonist influence support the inhibitory effects on basal or HCG induced progesterone secretion in vitro (Tarlatzis and Kolibianakis, 2002). The presence of GnRH receptors has been studied in the preovulatory follicle, corpus luteum and granulosa luteal cells. Several studies examining the influence of GnRH antagonist on progesterone and oestrogen biosynthesis in the ovary gave conflicting results. It appears that the effect of GnRH antagonist and agonist on human ovarian steroidogenesis in vitro is similar (Terlatzis and Kolibianakis, 2002). It can be concluded that the flexible single-dose antagonist protocol gave similar clinical results to the long agonist protocol. However, because of numerous advantages, it offers a valid alternative treatment regimen for ART. So far, single and multiple dose protocols have not been compared prospectively. The single dose is simple but requires strict monitoring of the cycle. Multiple doses almost eliminate the need for hormone assessment (Olivennes, 2002). Antagonist protocols for ovarian stimulation are more patient-friendly than long agonist protocols. Additional simplification can be expected from the new long-acting formulation of recombinant human FSH, which is designed to offer patients the alternative of a single administration rather than multiple daily injections of rfsh (Ludwig et al., 2002a,b). Acknowledgement The authors wish to thank Marijana Gajsek-Marchetti, translator from the Medical Research Department, for her contribution in preparing the manuscript. References Albano C, Smitz J, Camus M 1997 Comparison of different doses of gonadotropin-releasing hormone antagonist cetrorelix during controlled ovarian hyperstimulation. Fertility and Sterility 67, Albano C, Felberbaum RE, Smitz J et al Ovarian stimulation with HMG: results of a prospective randomised phase III European study comparing the luteinizing hormone-releasing hormone (LHRH)-antagonist cetrorelix and LHRH-agonist buserelin. Human Reproduction 15, Al-Inany H, Aboulghar M 2002 Gonadotrophin-releasing hormone antagonists for assisted conception (Cochrane Review). The Cochrane Library, 1, Oxford: Update Software. Barri NP, Martinez F, Coroleu B, Tur R 2002 The role of GnRH antagonists in assisted reproduction. Reproductive BioMedicine Online 5 (suppl. 1), Bouchard P, Fauser BCJM 2000 Gonadotropin-releasing hormone antagonist: new tools vs. old habits. Fertility and Sterility 73, De la Fuente R, O Brien JM, Eppig JJ 1999 EGF enhances preimplantation developmental competence of maturing mouse oocytes. Human Reproduction 14, Devroey P 2000 GnRH antagonists. Fertility and Sterility 73, Diedrich K, Diedrich E, Santos E et al Suppression of the endogenous LH-surge by the LH-RH antagonist cetrorelix during ovarian stimulation. Human Reproduction 9, Felderbaum R, Diedrich K 2002 Gonadotrophin-releasing hormone antagonists: will they replace the agonists? Reproductive BioMedicine Online 6,

8 308 Fluker M, Grifo J, Leader A et al Efficacy and safety of ganirelix acetate versus leuprolide acetate in women undergoing controlled ovarian hyperstimulation. Fertility and Sterility 75, Ganirelix Dose-Finding Study 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 ). Human Reproduction 13, Hernandez ER 2000 Embryo implantation: the rubicon for GnRH antagonists. Human Reproduction 15, Howles CM 2002 The place of gonadotrophin-releasing hormone antagonists in reproductive medicine. Reproductive BioMedicine Online 4, Suppl. 3, Huirne JAF, Lambalk CB 2001 Gonadotropin-releasing-hormonereceptor antagonist. Lancet 358, Hung Yu E, Chung Ho P 2001 Use of gonadotrophin releasing hormone (GnRH) antagonist (Cetrotide) during ovarian stimulation for in vitro fertilization treatment: multiple doses and single dose. Journal of Obstetrics and Gynecology Research 27, Kol S 2000 Embryo implantation and GnRH antagonists: GnRH antagonist in ART: lower embryo implantation? Human Reproduction 15, Kovacic B, Vlaisavljevic V, Reljic M et al Clinical outcome of day 2 versus day 5 transfer in cycles with one or two developed embryos. Fertility and Sterility 77, Leroy I, d Acremont MF, Brailly-Tabard S et al A single injection of gonadotropin-releasing hormone (LH-RH) antagonist (cetrorelix) postpones the luteinizing hormone (LH) surge: further evidence of the role of LH-RH during the LH surge. Fertility and Sterility 62, Ludwig M, Katalinic A, Diedrich K 2001 Use of GnRH antagonists in ovarian stimulation for assisted reproductive technologies compared to the long protocol. Meta analysis. Archives of Gynecology and Obstetrics 265, Ludwig M, Felberbaum ER, Diedrich K, Lunenfeld B 2002a Ovarian stimulation: from basic science to clinical application. Reproductive BioMedicine Online 5, Suppl. 1, Ludwig M, Katalinic A, Banz C et al. 2002b Tailoring the GnRH antagonist cetrorelix acetate to individual patients needs in ovarian stimulation for IVF: results of a prospective, randomized study. Human Reproduction 17, Nikolettos N, Al-Hasani S, Felberbaum R et al Gonadotropinreleasing hormone antagonist protocol: a novel method of ovarian stimulation in poor responders. European Journal of Obstetrics, Gynecology and Reproductive Biology 97, Olivennes F 2002 GnRH antagonists: do they open new pathways to safer treatment in assisted reproductive techniques? Reproductive BioMedicine Online 5 (suppl. 1), Olivennes F, Fanchin R, Bouchard P et al The single or dual administration of the gonadotrophin-releasing hormone antagonist cetrorelix in an in vitro fertilization-embryo transfer program. Fertility and Sterility 62, Olivennes F, Alvarez S, Bouchard P et al The use of LH-RH antagonist (cetrorelix) in a single dose protocol in IVF embryo transfer: a dose finding study of 3 versus 2 mg. Human Reproduction 13, Olivennes F, Belaisch-Allart J, Emperaire JC, et al Prospective, randomized, controlled study of in vitro fertilization embryo transfer with a single dose of a luteinizing hormone-releasing hormone (LH-RH) antagonist (cetrorelix) or a depot formula of an LH-RH agonist (triptorelin). Fertility and Sterility 73, Reljic M, Vlaisavljevic V, Gavric V et al Number of oocytes retrieved and resulting pregnancy risk factors for ovarian hyperstimulation syndrome. Journal of Reproductive Medicine 44, Tarlatzis BC, Kolibianakis EM Direct ovarian effects and safety aspects of GnRH agonists and antagonists. Reproductive BioMedicine Online 5, Suppl. 1, Vlaisavljevic V, Kovacic B, Gavric Lovrec V et al Simplification of the clinical phase of IVF and ICSI treatment in programmed cycles. International Journal of Gynecology and Obstetrics 69, Vlaisavljevic V, Kovacic B, Reljic M et al Is there any benefit from the culture of a single oocyte to a blastocyst-stage embryo? Human Reproduction 16, Received 4 March 2003; refereed 11 April 2003, accepted 2 June 2003.