In a flexible antagonist protocol, earlier, criteria-based initiation of GnRH antagonist is associated with increased pregnancy rates in IVF

Trifon Lainas1,5, John Zorzovilis1, George Petsas1, Georgia Stavropoulou1, Haris Cazlaris2, Vassiliki Daskalaki3, George Lainas4 and Efthymia Alexopoulou4

1 Iatriki Erevna-IVF Unit, 7, Ventiri Str., Athens 11528, 2 University of Thessaly, Medical School, Histology-Embryology Department, Larissa, 3 General Hospital ‘G.GENNIMATAS’, 154, Mesogion Avenue, Athens and 4 Second Radiology Department, University of Athens, Athens, Greece

5 To whom correspondence should be addressed. Email: iatrikiereyna{at}ath.forthnet.gr


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
BACKGROUND: The purpose of the study was to assess ongoing pregnancy rates across groups of patients treated by IVF, which were defined according to criteria aimed at the prevention of premature LH surge and used for initiating GnRH antagonist. METHODS: This is a prospective observational cohort study. During the last 3 years, in IVF-ICSI patients undergoing controlled ovarian stimulation (COS) with the antagonist protocol, the antagonist administration was initiated according to at least one of the following patient-specific criteria: (i) at least one follicle measuring >14 mm; (ii) estradiol levels >600 pg/ml; and (iii) LH levels >10 IU/l. Based upon these criteria, 208 cases of normal responders were analysed and categorized into three groups according to the starting day of the regimen: group D4 (n=40) for day 4, group D5 (n=98) for day 5 and group D6 (n=70) for day 6. The main outcome measure was the ongoing pregnancy rate per started cycle. RESULTS: The total number of patients in the D4 and D5 groups (138 out of 208), who received the antagonist earlier, was considerably larger compared with that of D6 (70 out of 208). Ongoing pregnancy rates were 37.5, 34.7 and 18.6% for groups D4, D5 and D6, respectively. Patients who initiated the GnRH antagonist on days 4 and 5 had statistically significant higher pregnancy rates compared with day 6. Rapid response, causing earlier antagonist administration initiation, according to the proposed criteria for the prevention of premature LH surges, and the absence of premature luteinization, as evidenced by normal progesterone levels on HCG day, were found to be independent positive predictive factors for favourable IVF outcome. CONCLUSIONS: The employment of an algorithm of criteria, aimed at the prevention of premature LH surges in a flexible antagonist protocol, resulted in antagonist initiation earlier than on stimulation day 6 in a significant proportion of patients. In those patients, a higher pregnancy rate was observed.

Key words: criteria for antagonist initiation/flexible antagonist protocol/GnRH antagonist


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Controlled ovarian stimulation (COS) is the primary reason for the improved IVF-embryo transfer (IVF-ET) success rate evidenced during the past two decades (Olivennes et al., 2000Go). A common pitfall of ovarian stimulation is the premature increase in serum levels of LH and progesterone, reported in up to 25% of the patients (Frydman et al., 1991Go). The primary objective of the methods in assisted reproductive technologies (ARTs) is to prevent a premature LH surge. Of the different stimulation regimens, the GnRH agonist down-regulation protocol, widely known as the ‘long protocol’ has come to be considered as the gold standard (Bachelot et al., 1998Go).

On the other hand, GnRH antagonists, such as ganirelix and cetrorelix, provide a new, promising pathway in ART, since antagonists immediately suppress gonadotrophin secretion by blocking the GnRH receptor (Rabinovici et al., 1992Go; Nelson et al., 1995Go), and thus treatment may be restricted to those days when a premature LH surge is likely to occur.

In all studies in which the daily antagonist treatment was established as comparable with the long protocol, the administration of the antagonist regimen was initiated at day 6 of stimulation (Albano et al., 2000Go; European Orgalutran Study Group, 2000Go; European Middle East Orgalutran Study Group, 2001Go; Fluker et al., 2001Go). In all four studies, the multiple low dose (0.25 mg) antagonist regimen was applied. Although the differences in pregnancy rates between the antagonist and long protocols were not statistically significant in any of the four studies, pregnancy rates appeared somewhat lower with the antagonist arms (Shapiro, 2003Go).

A meta-analysis of the above four studies (Al-Inany and Aboulghar, 2002Go), including a fifth study with the single high dose (3 mg) antagonist regimen (Olivennes et al., 2000Go), showed that the probability of clinical pregnancy was lower with GnRH antagonists as compared with GnRH agonist, without any significant difference between the two protocols regarding prevention of premature LH surge and occurrence of severe ovarian hyperstimulation syndrome (OHSS). The study concluded that the clinical outcome may be improved further by developing more flexible antagonist regimens, taking into account individual patient characteristics.

There followed a number of studies where various flexible antagonist protocols were compared with the fixed day 6 protocol. In a randomized trial (Ludwig et al., 2002Go), the GnRH administration was delayed until the dominant follicle reached ≥14 mm in diameter. This flexible regimen initiation resulted in a higher number of retrieved oocytes without any difference in clinical outcome as compared with the fixed regimen, despite less consumption of antagonist and recombinant FSH (rFSH).

A flexible approach to the use of the GnRH antagonist protocol according to the leading follicle size ≥14 mm has been compared with the fixed sixth day use in a prospective randomized trial (Escudero et al., 2004Go). No statistically significant difference in outcomes was observed. The starting day of the regimen for the flexible arm was after day 6 of stimulation (7.2±0.2 days).

In another study (Kolibianakis et al., 2003aGo), the flexible arm was based on the criterion of the leading follicle being ≥15 mm at least 5 days after stimulation. In patients with no follicle of ≥15 mm present on day 6 of stimulation, a significantly lower ongoing implantation rate was observed if the flexible scheme was applied as compared with the fixed scheme of administration.

Finally, in a recent study (Mochtar, 2004Go), the effect of an individualized GnRH antagonist protocol on folliculogenesis was examined. For the flexible arm of the study, the starting day of the antagonist was based on the leading follicle size (>15 mm). The authors found a lower, albeit not significantly different, pregnancy rate in the flexible group as compared with the group of patients who started the antagonist in the fixed mode of day 6.

A recent meta-analytic review (Al-Inany et al., 2005Go) included the above four randomized control trials. The analysis concluded that there was no statistically significant difference in pregnancy rate per woman randomized, although there was a trend towards a higher pregnancy rate with the fixed protocol, especially with delayed administration beyond day 8. There was no premature LH surge in any participant in either protocol. However, there was a statistically significant reduction both in number of antagonist ampoules and in amount of gonadotrophin used in the flexible protocol.

During the initial phase of antagonist application in our unit, daily laboratory assessments [LH, estradiol (E2) and progesterone] were recorded. It then appeared that a considerable number of patients showed an abrupt increase of LH prior to the standard day 6 of antagonist administration, accompanied by a proportional increase in levels of E2.

On the basis of the above, in order to a prevent premature LH surge, we reconsidered the day of antagonist initiation and established criteria for the starting day according to follicular size, E2 and LH levels.

The criteria were established following the accumulated knowledge on reproductive physiology. Data from the literature showed that the LH surge is unlikely to occur before either the follicle diameter has reached 15 mm and/or serum E2 levels have reached 600 pmol/l (Cahill et al., 1998Go). There was additional indication that administration of GnRH antagonist when serum E2 exceeds 600 pg/ml prevents premature elevation of LH (Frydman et al., 1991Go). Based on the above, we established the first two criteria; the first one concerning the leading follicle being at least 14 mm in diameter and the second one relating to the E2 serum levels exceeding the value of 600 pg/ml. As a third safety criterion, we used the value of 10 IU/l for LH serum levels (Ferin, 1996Go; European Orgalutran Study Group, 2000Go; Fluker et al., 2001Go).

The purpose of the study was to assess ongoing pregnancy rates across a group of patients treated by IVF, which were defined according to criteria aimed at the prevention of premature LH surge and used for initiating GnRH antagonist.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Study design and patient population
This is a prospective monocentric observational cohort study. Study patients started the antagonist administration when they fulfilled at least one of the following criteria: (i) the presence of at least one follicle measuring >14 mm; (ii) E2 serum levels >600 pg/ml; and (iii) LH serum levels >10 IU/l.

All patients followed a flexible approach, based on the above criteria. Only patients who were fully informed and consented to participate were included. The study was reviewed and approved by the unit's ethics committee.

Two hundred and eight patients treated by IVF/ICSI with the antagonist protocol in our unit during a period of 3 years (July 2001–August 2004) were included in the study. Patients were enrolled in the trial only once. The inclusion criteria were: age of at least 18 but not more than 39 years, a body mass index (BMI) of 18–29 kg/m2, a menstrual cycle ranging from 24 to 35 days and normal basal FSH (<10 IU/l), LH (<10 IU/l) and E2 <80 pg/ml) levels. All poor responder cases, as defined elsewhere (Dor et al., 1992Go; Surrey and Schoolcraft, 2000Go; Copperman, 2003Go; Tarlatzis et al., 2003Go), were excluded. None of the patients received an oral contraceptive on the cycle before ovarian stimulation.

Controlled ovarian stimulation
All patients started rFSH treatment with injections of follitropin-{beta} (Puregon; Organon, The Netherlands) on day 2 or 3 of menses. The starting dose of rFSH was 225 IU per day for all patients. Before antagonist initiation, the dose was adjusted on the third or fourth day of stimulation, depending on the ovarian response as indicated by ultrasound and especially serum E2 levels.

Administration of rFSH continued daily thereafter until and including the day of HCG administration. Ultrasound was performed concomitantly with hormonal assessment in all patients before initiation of stimulation. No follicles >10 mm were present on day 2 or 3 in the patients analysed.

The follow-up included estimation of E2, LH and progesterone, and ultrasound assessment of endometrium and follicular growth, starting on day 4, and was continued daily, up to and including the day of HCG administration. Blood samples were taken early in the morning and all assessments were performed immediately. This is a vital step, since the decision regarding antagonist initiation was based on follicular size, LH and E2 values, in correlation with the preceding day's measurements. This ensured the absence of a lag in the initiation of the antagonist administration, since the danger of an LH upsurge might have been imminent.

Daily antagonist administration of ganirelix 0.25 mg (Orgalutran, Organon, The Netherlands) was initiated if at least one of the above-mentioned criteria was met, and was continued up to and including the day of HCG administration.

When at least three follicles were ≥17 mm, 10 000 IU of HCG (Pregnyl; Organon, The Netherlands) were administered i.m. Transvaginal ultrasound-guided oocyte retrieval was performed 36 h later. Ultrasound guidance was used for all embryo transfers. Luteal phase support with 600 mg of micronized progesterone (Utrogestan Laboratoires Besins-International S.A., France) was given from the day of embryo transfer.

Also, in accordance with the unit's established standard of care since 1998, patients at high risk of developing OHSS (Brannstrom and Enskog, 2002Go) received treatment with methylprednisolone (Lainas et al., 2002Go). For the purposes of the current study, we adopted the modified system of OHSS classification previously described (Schenker and Weinstein, 1978Go).

Ultrasound and laboratory assays
All ultrasound measurements were performed using a 7.5, 6 or 5 MHz vaginal probe (Sonoline, Adara, Siemens). E2, progesterone and LH levels were measured on an Immulite analyser using the corresponding commercially available kits (DPC, Los Angeles, CA). Analytical sensitivity was 0.1 mIU/ml for LH, 15 pg/ml for E2 and 0.2 ng/ml for progesterone. Intra- and inter-assay precision at the concentrations of most relevance to the current study (expressed as coefficients of variation) were 5.9 and 8.1% for LH, 6.3 and 6.4% for E2, and 7.9 and 10% for progesterone.

Outcome measures and statistical analysis
The primary outcome measure was the ongoing pregnancy rate per started cycle.

Additionally, biochemical and clinical pregnancy rates were estimated. Ongoing pregnancy was defined as the presence of gestational sac with fetal heartbeat detection at 12 week of gestation; clinical pregnancy was defined as the presence of a gestational sac with positive heartbeat at 6 weeks; and biochemical pregnancy as positive HCG test 2 weeks after ET.

Secondary outcome measures were the number of oocytes retrieved, the number of 2PN zygotes, the duration of stimulation, and the E2, LH and progesterone concentration on the day of HCG administration. The incidence of OHSS was also recorded. Comparison groups were patients who started antagonist on day 4 (group D4), on day 5 (group D5) and on day 6 (group D6).

Proportional differences of the main outcome measures of pregnancy rate and the distribution of OHSS among the groups were analysed with the {chi}2 for trends statistic. In the case of OHSS, Somer's d statistic was also employed. Other categorical variables (distributions of causes of infertility, previous IVF attempts and of the criteria of the antagonist initiation between the groups) were analysed with the Pearson's {chi}2 statistic. Measures concerning scale variables were compared by the one-way analysis of variance (ANOVA) method, followed by post hoc Bonferroni tests for multiple comparisons. Finally, in order to elucidate which factors are most responsible for favourable IVF outcome, a number of logistic regression models were applied. The dependent variable was ongoing pregnancy and the parameters examined were baseline characteristics (age, BMI and years of infertility), baseline hormonal profile (FSH, LH and E2), antagonist initiation day, total and mean daily FSH dosage, and secondary end-points (stimulation period, progesterone and E2 levels on HCG day). At the first step, all covariates were entered into the logistic regression model one by one. At the second step, all covariates were forced to enter the regression model. At the third step, a stepwise approach was applied. Also, at this and the next step, the antagonist initiation day was converted to a categorical variable, with day 6 serving as the baseline. Finally, at the fourth step, the regression model was calculated only with the covariates that were found to have a significant effect on any the first three steps. All tests were two-tailed and the level of statistical significance was set at 0.05.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The final population for statistical analysis included 208 patients who were categorized into three distinct groups based on the day of antagonist administration: group D4, 40 patients (19.2%) on day 4; group D5, 98 patients (47.1%) on day 5; and group D6, 70 patients (33.7%) on day 6. This means that, following the patient-specific criteria mentioned above, the majority (138 out of 208 or 66.3%) of patients started antagonist treatment before day 6. Only 14 patients started the antagonist after day 6 (eight on day 7, three on day 8 and three on days 9/10). These patients were not included in the analysis.

Table I shows the distribution of patients fulfilling the different criteria for antagonist initiation. The ultrasound criterion on its own was able to qualify two-thirds of the population, but it also appeared in combination with the E2 criterion. In total, nine out of 10 women fulfilled the ultrasound criterion, either as the only criterion or in combination with E2. However, there still remained 10% of the women who had elevated E2 levels without a concomitant increase in the leading follicle diameter. Finally, an increase in LH levels beyond 10 IU/l was observed only in one woman and again in conjunction with a concurrent upsurge of E2. None of the women in the total sample surpassed the 10 IU/l value of LH after the antagonist initiation and up to the HCG day.


View this table:
[in this window]
[in a new window]
 
Table I. Absolute and percentage distribution of patients fulfilling the different criteria for antagonist initiation

 
The distribution of the criteria did not differ among the groups in a significant manner, although there was an increased proportion of women in group D6 who simultaneously met both the ultrasound and E2 criteria.

Baseline characteristics and hormonal profile of the patients are listed in Table II. With regards to the characteristics, ANOVA tests revealed that age, BMI and duration of infertility did not differ significantly among the three groups. Likewise, {chi}2 comparisons showed that the causes of infertility and the proportion of patients who had previous attempts at IVF treatment were homogeneously distributed among the groups.


View this table:
[in this window]
[in a new window]
 
Table II. Baseline characteristics and hormonal profile of the study groups

 
With regards to baseline hormone values, ANOVA tests showed that the mean values of most hormones were equal among groups, with the exception of FSH and E2. Bonferroni post hoc pairwise comparisons showed that in FSH, statistically significant differences were between D4 and D5, while in E2 they were between D6 compared with D4 and D5. The cancellation rate was similar among the study groups (two out of 40 for group D4, two out of 98 for group D5, one out of 70 for group D6). The groups were also comparable with regards to the number of embryos transferred (group D4, 3.4±1.5; group D5, 3.8±1.4; and group D6, 3.5±1.5, P=0.288). However, there were differences in the total FSH dosage and the mean FSH dosage per day. The total FSH dosage was 2116±681 for day 4, 2624±913 for day 5 and 3176±1220 for day 6. All three groups were different from each other. The mean FSH dosage per day was 251±90 for day 4, 289±93 for day 5 and 318±103 for day 6. In this case, statistically significant differences exist between day 4 and day 6 only. The association of the three groups with the outcome measures is summarized in Table III. Differences in the pregnancy rates between the groups were tested with the {chi}2 for trend statistic. The ongoing pregnancy rate per started cycle was 37.5% for group D4, 34.7% for group D5 and 18.6% for group D6. Also, biochemical and clinical pregnancy rates per started cycle were significantly greater in groups D4 and D5 compared with group D6 (see Figure. 1). The likelihood of an ongoing pregnancy per started cycle in group D5 was 2.3 times greater than in group D6 [95% confidence interval (CI) 1.1–4.8], whereas in group D4 the likelihood was increased to 2.6 (95% CI 1.1–6.3) compared with that of group D6. Even higher are the odds ratios of D5 (2.74 with 95% CI 1.33–5.65) and D4 (3.09 with 95% CI 1.29–7.31) with respect to the baseline value of D6. Only three women (21.4%) who started the antagonist later than day 6 had a positive IVF outcome. Incidentally, they all belonged to day 7, which brings the pregnancy rate for the specific day to 37.5%.


View this table:
[in this window]
[in a new window]
 
Table III. Associations of pregnancy rates and outcome variables with the study groups

 


View larger version (30K):
[in this window]
[in a new window]
 
Figure 1. Comparison of pregnancy rates of groups D4, D5 and D6 in the antagonist protocol. Odds ratios for ongoing pregnancy rates: day 4 versus day 6, 3.09 with 95% CI 1.29–7.31; day 5 versus day 6, 2.74 with 95% CI 1.33–5.65.

 
In relation to secondary end-points, duration of stimulation was significantly higher in group D6 than in groups D4 and D5. Oocytes were retrieved in significantly higher numbers in group D6 than in group D4. On the contrary, no statistically significant differences between the groups were observed in the number of eggs fertilized, as well as in the LH, E2 and progesterone levels on HCG day.

With regards to the occurrence of OHSS, severe hyperstimulation did not develop in any group. However, the {chi}2 for trends statistic showed there was a dependence of the appearance of mild and moderate OHSS on the antagonist initiation day. The odds ratio of the appearance of mild or moderate OHSS on day 5 versus day 4 was 2.66 (95% CI 0.87–8.30), while for day 6 versus day 4 the odds ratio rose to 3.21 (95% CI 1.01–11.1). This was corroborated further by the significance of Somer's statistic (P=0.034). However, the low value of Somer's d (0.134) indicates that the relationship between the appearance of OHSS and the antagonist initiation day is a fairly weak one.

Figure 2 shows the hormone profile of the three groups during stimulation. The mean values of the three hormones on the HCG day are shown in Table III. LH levels evidenced a significant decrease the day following antagonist initiation, i.e. on day 5 for group D4, day 6 for group 5 and day 7 for group 6. Furthermore, groups D5 and D6 evidenced a significant increase of LH levels just before antagonist initiation.



View larger version (16K):
[in this window]
[in a new window]
 
Figure 2. Hormone profile of the three groups during stimulation.

 
Finally, it was found that high progesterone levels on the HCG day are implicated in decreased pregnancy rates. The mean progesterone value on the HCG day for women who subsequently achieved pregnancy (1.03±0.41 ng/l) was significantly lower than the same value for women who did not achieve pregnancy (1.26±0.69 ng/l). However, progesterone values on the HCG day are not correlated either with the FSH dosage or with the duration of stimulation. They also do not correlate, as has already been shown (see secondary end-points in Table III), with the day of antagonist initiation.

Table IV shows the results of logistic regression models with ongoing pregnancy as the dependent variable and antagonist initiation day, baseline characteristics and hormonal profile and secondary end-points as the independent variables. Model 1 shows the odds ratios [expressed as exp(b)] with their 95% CIs with the related significance value when all covariates were forced to enter the regression model. The findings show that neither the baseline characteristics nor the baseline hormonal profile bears any predictive value on pregnancy outcome, despite the observed between-group differences in baseline FSH and E2. Additionally, logistic regression tests failed to provide evidence for the dependence of the IVF outcome on either the total or the mean daily FSH dosage. The same applies to secondary end-points such as stimulation period and E2 levels on the HCG day.


View this table:
[in this window]
[in a new window]
 
Table IV. Results of logistic regression models with ongoing pregnancy as the dependent variable

 
On the contrary, the results revealed that pregnancy outcome is dependent on two factors: progesterone levels on the HCG day and the rapid response as expressed by the women's compliance with the proposed criteria, according to which the antagonist was initiated. Both odds ratios are below unity, which means that slower response during stimulation and elevated progesterone levels on the HCG day reduce the probability of favourable IVF outcome.

Model 2 shows the results of the stepwise logistic regression including the two significant factors found in the previous steps. In this step, the antagonist initiation day was converted to a categorical variable, with day 6 serving as the baseline. The odds ratios of day 4 and 5 versus day 6 are naturally the ones already reported.


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Although several studies have indicated a slight reduction in pregnancy rate with GnRH antagonists as compared with agonists, it has been noted that this problem could be faced by developing flexible antagonist regimens designed for individual patients (Felderbaum and Diedrich, 2003Go). Is there evidence to support the idea that ‘tailored’ GnRH antagonist protocols would lead to better clinical results?

Several studies have been performed to compare a fixed (on day 6 of stimulation) with a flexible approach of GnRH antagonist initiation. In particular, the meta-analytic review of Al-Inany et al. (2005)Go concluded that although data are not yet conclusive, all four randomized conrolled trials (RCTs) included were in favour of the fixed protocol. Consequently there is no agreement yet as to whether a flexible approach can improve outcome with respect to the administration in a fixed manner.

The most appropriate moment to start GnRH antagonist administration has already been the subject of several studies. In all the above studies, antagonist was initiated in the flexible arm on day 6 of stimulation or later. The present study provides data on the use of a flexible GnRH antagonist protocol according to criteria that result in an earlier than day 6 antagonist initiation in the majority of the cases analysed. In this flexible approach, the ongoing pregnancy rate for patients that received the antagonist earlier was higher.

The above statement does not necessarily imply that starting the GnRH administration on day 4 or 5 improves cycle outcome. Likewise, it cannot be asserted that those patients that started GnRH administration according to the established criteria on day 6 or later would have had a better outcome if the antagonist was initiated earlier.

What the present findings show is that rapid response is a positive predictor of a favourable IVF outcome. The main idea of the present study was to prevent early LH surges for those patients with a rapid follicular growth, or with elevated E2 levels, given the existence of a positive feedback between LH and the increasing levels of E2 during follicular development (Ferin, 1996Go). In view of the above, in order to prevent premature LH surges in rapid responders, we proposed an algorithm for antagonist initiation, if at least one of three criteria (follicular size >14 mm, E2 >600 pg/ml and LH >10 IU/l) was met.

The above line of reasoning in not new. As early as 2001, the European Orgalutran Study Group pointed out that for women with a more rapid follicular growth, early LH rises may be prevented by starting ganirelix treatment on day 5 instead of day 6 of stimulation, while the North American Ganirelix Study Group (Fluker et al., 2001Go) noted that earlier administration of the antagonist may improve IVF outcome by limiting LH surges in more rapid responders.

An equally important issue is that early antagonist administration in the cycle decreases the exposure of the genital tract to LH and E2 in the early follicular phase. This high hormonal exposure has been associated with a reduced chance of pregnancy in IVF/ICSI treatment using the antagonist protocol (Kolibianakis et al., 2003aGo). Also, it has been suggested that the initiation of the antagonist regimen on day 1 compared with day 6 of stimulation is related to a lower exposure to LH and estradiol, without any affect to follicular development (Kolibianakis et al., 2003bGo).

The findings of our study show that practically two-thirds of the patients started the antagonist administration before the standard day 6 according to the strategy described. For the majority of patients, the criterion for the initiation of the antagonist was the diameter of the leading follicle >14 mm, while the E2 criterion proved to be not redundant. The fact that only one woman fulfilled the LH criterion (in combination with the E2 criterion) does not disqualify the specific criterion. The LH criterion should remain as a safety precaution, since antagonist administration is targeted towards the prevention of LH upsurges. Nevertheless, the first two criteria for antagonist initiation have proved to warrant this prevention of LH surges.

The fact that, in the present study, there was only a flexible protocol is clearly a limitation. This is because we focused our attention on the comparison of pregnancy rates in IVF cycles between groups of patients, which were defined according to specific criteria used for initiating GnRH antagonist. To the best of our knowledge, this is the first study that performs such comparisons. A second novelty is that our proposed criteria resulted in an earlier than day 6 antagonist initiation in the majority of the cases analysed.

In all the studies comparing the fixed with a flexible protocol, the pregnancy rates of the flexible protocol were computed for the total sample irrespective of the antagonist initiation day. In the study by Mochtar (2004)Go, if each initiation day is considered separately, the percentage of ongoing pregnancies for starting days 4 and 5 is higher than for the following days. Also, there seemed to be a declining trend in pregnancy rate in the flexible group from stimulation day 7 onward. Specifically for GnRH antagonist start days 4 and 5, the percentage of ongoing pregnancies was 36%; for day 6, 20%; for day 7, 24.2%; for day 8, 12.5%; and for days 9 and 10, 0%. Despite the different study design, the results in pregnancy rates for the flexible arm of the trial were similar to ours.

The lower pregnancy rate in group D6 may be explained by several factors. One factor may be the prolonged exposure to LH and E2 that has been associated with reduced chance of pregnancy. This may be attributed either to an effect of the early follicular phase endocrine environment on oocyte/embryo quality or to endometrium receptivity (Kolibianakis et al., 2003aGo). Secondly, delayed antagonist administration has been associated with increased endometrial advancement at oocyte retrieval, which in extreme cases may result in a decreased probability of pregnancy (Kolibianakis et al., 2002Go). Also, the same author concluded that on the basis of the available studies, it can be supported that prolongation of the follicular phase decreases the probability of pregnancy (Kolibianakis et al., 2005Go). Finally, it may be that patients that finally initiated GnRH antagonist on day 6 or later had a slower response that required higher gonadotrophin dosages and thus a lower prognosis independently of the moment of GnRH administration.

Based on our criteria, the population of patients in the D6 and D4+D5 groups seems to be distinct despite the homogeneous baseline characteristics among the groups. This distinction in the response of the groups, as statistical analysis has shown, could not be predicted by the observed statistical differences in baseline FSH and E2 values. Moreover, these differences are inconsequential from the clinical point of view, since all baseline values were within the accepted normal limits allowing for the commencement of the IVF trial.

The slower response of D6 should not be confused with poor response. As a matter of fact, the number of oocytes retrieved for D6 patients was significantly greater than for D4 and D5. This means that D6 patients were not poor responders but, on the contrary, were better responders than D4 and D5. However, a lower fertilization rate in D6 resulted in the same number of fertilized eggs as D4 and D5.

In summary, based on the collected data, before the commencement of the IVF trial, there was no way to foretell that a specific woman would be a rapid responder and that this would increase her chances of a favourable outcome. In addition, it is undistinguishable whether the intervention (antagonist administration on day 4, 5 or 6) or underlying characteristics of the patients (patients reacting differently to ovarian stimulation and thus fulfilling the antagonist criteria earlier or later) cause the differences observed in pregnancy rate.

Our data need support from RCTs to clarify: (i) whether D4 and D5 patients had better outcome due to the intervention (early antagonist administration), among other factors; and (ii) whether earlier administration of the antagonist would also benefit D6.

The number of patients who started the antagonist regimen later than day 6 of stimulation according to our criteria was too small to allow us to draw any conclusions. Our comment on this is that most of them actually started receiving the antagonist on day 7 of stimulation, where most of them had a polycystic ovary ultrasound appearance (Rotterdam PCOS Consensus Workshop Group, 2004Go).

The findings concerning secondary end-points, i.e. shorter duration of stimulation and reduced occurrence of OHSS, as well as reduced total FSH consumption show that the implementation of the protocol in a flexible manner enhances the well known advantages and benefits of the GnRH antagonists. This is especially true in those cases where clinical criteria demand an early but timely antagonist initiation.

Another interesting point that emerged from the analysis is the association between pregnancy rates and progesterone levels on the day of HCG. The implication of premature luteinization, as evidenced by high progesterone levels, towards lower implantation and pregnancy rates has already been reported in several studies. Specifically, in a study by Bosch et al. (2003)Go, it was concluded that premature luteinization during GnRH antagonist IVF-ET cycles is a frequent event that is associated with lower pregnancy and implantation rates. Progesterone elevations are not related to serum LH levels and may reflect the mature granulosa cell's response to high FSH exposure. A progesterone level ≥1.2 ng/l was the best cut-off criterion for an adverse outcome. The frequency of women above this cut-off point was 38.3% and they had a significantly lower pregnancy rate per transfer in comparison with women with progesterone levels below the cut-off point (25.8 versus 54.0%). Likewise, the implantation rate (13.8 versus 32.0%) was significantly lower in the premature luteinization group.

What is important from the perspective of the present study is that, since antagonist initiation day and progesterone levels on the HCG day both entered the logistic regression model, these two factors influence the outcome of IVF independently of each other.

In conclusion, the employment of an algorithm of criteria, aimed at the prevention of premature LH surges in a flexible antagonist protocol, resulted in administration of the antagonist earlier than on stimulation day 6 in a significant proportion of patients. In those patients, a higher pregnancy rate was observed.


    Acknowledgements
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The authors thank the study participants, Carol Barnes for valuable editorial help, Niki Kokkona for thorough collection of data, Antigoni Bonti for continuous reference survey, and Roula Karoussou and Kondylia Boura for meticulous data entry and preparation of the manuscript.


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Albano C, Felberbaum RE, Smitz J, Riethmuller-Winzen H, Engel J, Diedrich K and Devroey P on behalf of the European Cetrorelix Study Group (2000) Ovarian stimulation with hMG: results of a prospective randomized phase III European study comparing the luteinizing hormone-releasing hormone (LHRH)-antagonist cetrorelix and the LHRH agonist buserellin. Hum Reprod 15, 526–531.[Abstract/Free Full Text]

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

Al-Inany HG, Aboulghar M, Mansour R and Serour GI (2005) Optimizing GnRH antagonist administration: meta-analysis of fixed vs flexible protocol. Reprod Biomed Online 10, in press.

Bachelot A, Pouly JL, Devecchi Quenard A and de Mouzon J (1998) 1997 FIVNAT general balance. Contracept Fertil Sex 26, 463–465.[ISI][Medline]

Bosch E, Valencia I, Escudero E, Crespo J, Simon C, Remohi J and Pellicer A (2003) Premature luteinization during gonadotropin releasing hormone antagonist cycles and its relationship with in vitro fertilization outcome. Fertil Steril 80, 1444–1449.[CrossRef][ISI][Medline]

Brannstrom M and Enskog A (2002) Leukocyte networks and ovulation. J Reprod Immunol 57, 47–60.[CrossRef][ISI][Medline]

Cahill DJ, Wardle PG, Harlow CR and Hull MGR (1998) Onset of the preovulatory luteinizing hormone surge: diurnal timing and critical follicular prerequisites. Fertil Steril 70, 56–59.[CrossRef][ISI][Medline]

Copperman AB (2003) Antagonists in poor responder patients. Real-life clinical applications of GnRH antagonists: individualization of fertility treatment. Fertil Steril 80 (Suppl 1), 16–24.[ISI][Medline]

Dor J, Seidman DS, Ben-Shlomo I, Levran D, Karasik A and Mashiach S (1992) The prognostic importance of the number of oocytes retrieved and estradiol levels in poor and normal responders in in-vitro fertilization (IVF) treatment. J Assist Reprod Genet 9, 228–232.[CrossRef][ISI][Medline]

Escudero E, Bosch E, Crespo J, Simón C, Remohi J and Pellicer A (2004) Comparison of two different starting multiple dose gonadotropin-releasing hormone antagonist protocols in a selected group of in vitro fertilization-embryo transfer patients. Fertil Steril 81, 562–566.[CrossRef][ISI][Medline]

European Middle East Orgalutran Study Group (2001) Comparable clinical outcomes using the GnRH antagonist ganirelix or a long protocol of the GnRH agonist triptorelin for the prevention of premature LH surges in women undergoing ovarian stimulation. Hum Reprod 16, 644–651.[Abstract/Free Full Text]

European Orgalutran Study Group, Borm G and Mannaerts B (2000) Treatment with the gonadotropin-releasing hormone antagonist ganirelix in women undergoing ovarian stimulation with recombinant follicle stimulating hormone is effective, safe and convenient: results of a controlled, randomized, multicentre trial. Hum Reprod 15, 1490–1498.[Abstract/Free Full Text]

Felberbaum R and Diedrich K (2003) Gonadotropin-releasing hormone antagonists: will they replace the agonists? Reprod Biomed Online 6, 43–53.[Medline]

Ferin MJ (1996) The menstrual cycle: an integrative view. In Adashi EY, Rock JA, and Rosenwaks Z (eds) Reproductive Endocrinology, Surgery, and Technology, Vol. 1. Lippincott–Raven, New York, pp. 104–121.

Fluker M, Grifo J, Leader A, Levy M, Meldrum D, Muasher SJ, Rinehart J, Rosenwaks Z, Scott RT Jr, Schoolcraft W, Shapiro DB and Gordon K; the North American Ganirelix Study Group (2001) Efficacy and safety of ganirelix acetate versus leuprolide acetate in women undergoing controlled ovarian hyperstimulation. Fertil Steril 75, 38–45.[CrossRef][ISI][Medline]

Frydman R, Cornel C, de Ziegler D, Taieb J, Spitz IM and Bouchard P (1991) Prevention of premature luteinizing hormone and progesterone rise with a gonadotropin-releasing hormone antagonist. Nal-Glu, in controlled ovarian hyperstimulation. Fertil Steril 56, 923–927.[ISI][Medline]

Kolibianakis E, Bourgain C, Albano C, Osmanagaoglou K, Smitz J, Van Steirteghem AC and Devroey P (2002) Effect of ovarian stimulation with recombinant follicle-stimulating hormone, gonadotropin releasing hormone antagonists, and human chorionic gonadotropin on endometrial maturation on the day of oocyte pick-up. Fertil Steril 78, 1025–1029.[CrossRef][ISI][Medline]

Kolibianakis E, Albano C, Kahn J, Camus M, Tournaye H, Van Steirteghem AC and Devroey P (2003a) Exposure to high levels of luteinizing hormone and estradiol in the early follicular phase of gonadotropin-releasing hormone antagonist cycles is associated with a reduced chance of pregnancy. Fertil Steril 79, 873–880.[CrossRef][ISI][Medline]

Kolibianakis EM, Albano C, Camus M, Tournaye H, Van Steirteghem AC and Devroey P (2003b) Initiation of gonadotropin-releasing hormone antagonist on day 1 as compared to day 6 of stimulation: effect on hormonal levels and follicular development in in vitro fertilization cycles. J Clin Endocrinol Metab 88, 5632–5637.[Abstract/Free Full Text]

Kolibianakis EM, Tarlatzis BC and Devroey P (2005) GnRH antagonists in IVF. Reprod Biomed Online 10, in press.

Lainas T, Petsas G, Stavropoulou G, Alexopoulou E, Iliadis G and Minaretzis D (2002) Administration of methylprednisolone to prevent ovarian hyperstimulation syndrome in patients undergoing in vitro fertilization. Fertil Steril 78, 529–533.[CrossRef][ISI][Medline]

Ludwig M, Katalinic A, Banz C, Schroder AK, Loning M, Weiss JM and Diedrich K (2002) Tailoring the GnRH antagonist cetrorelix acetate individual patients' needs in ovarian stimulation for IVF: results of a prospective randomized study. Hum Reprod 17, 2842–2845.[Abstract/Free Full Text]

Mochtar MH on behalf of The Dutch Ganirelix Study Group and (2004) The effect of an individualized GnRH antagonist protocol on folliculogenesis in IVF/ICSI. Hum Reprod 8, 1713–1718.[CrossRef]

Nelson LR, Fujimoto VY, Jaffe RB and Monroe SE (1995) Suppression of follicular phase pituitary gonadal function by a potent new gonadotropin-releasing hormone antagonist with reduced histamine-releasing properties (ganirelix). Fertil Steril 63, 963–969.[ISI][Medline]

Olivennes F, Belaisch-Allart J, Emperaire JC, Dechaud H, Alvarez S, Moreau L, Nicollet B, Zorn JR, Bouchard P and Frydman R (2000) 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). Fertil Steril 73, 314–320.[CrossRef][ISI][Medline]

Rabinovici J, Rothman P, Monroe SE, Nerenberg C and Jaffe RB (1992) Endocrine effects and pharmacokinetic characteristics of a potent new gonadotropin-releasing hormone antagonist (ganirelix) with minimal histamine-releasing properties: studies in menopausal women. J Clin Endocrinol Metab 75, 1220–1225.[Abstract]

Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group (2004) Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertil Steril 81, 19–25.[CrossRef][ISI][Medline]

Schenker J and Weinstein D (1978) Ovarian hyperstimulation syndrome: a current survey. Fertil Steril 30, 255–268.[ISI][Medline]

Shapiro DB (2003) GnRH antagonists in normal-responder patientsl. Real-life clinical applications of GnRH antagonists: individualization of fertility treatment. Fertil Steril 80 (Suppl 1), 8–15.

Surrey ES and Schoolcraft WB (2000) Evaluating strategies for improving ovarian response of the poor responder undergoing assisted reproductive techniques. Fertil Steril 73, 667–676.[CrossRef][ISI][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 4, 2005; resubmitted on April 27, 2005; accepted on April 29, 2005.





This Article
Abstract
Full Text (PDF )
All Versions of this Article:
20/9/2426    most recent
dei106v1
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Request Permissions
Google Scholar
Articles by Lainas, T.
Articles by Alexopoulou, E.
PubMed
PubMed Citation
Articles by Lainas, T.
Articles by Alexopoulou, E.