Human Reproduction Update Advance Access originally published online on August 18, 2006
Human Reproduction Update 2006 12(6):651-671; doi:10.1093/humupd/dml038
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Among patients treated for IVF with gonadotrophins and GnRH analogues, is the probability of live birth dependent on the type of analogue used? A systematic review and meta-analysis
1 Unit for Human Reproduction, 1st Department of Obstetrics and Gynaecology, Aristotle University of Thessaloniki, Thessaloniki, Greece, 2 McMaster University, Hamilton and Dalhousie University, Halifax, Nova Scotia, Canada, 3 Centre for Reproductive Medicine, Dutch-Speaking Brussels Free University, Brussels, Belgium and 4 Department of Obstetrics and Gynecology, University Clinic of Schleswig-Holstein, Luebeck, Germany
5 To whom correspondence should be addressed at: Unit for Human Reproduction, Papageorgiou General Hospital, Nea Efkarpia Peripheral Road, Thessaloniki 54603, Greece. E-mail: stratis.kolibianakis{at}otenet.gr
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
This systematic review and meta-analysis aimed to answer the following clinical question: among patients treated for IVF with gonadotrophins and GnRH analogues, is the probability of live birth per randomized patient dependent on the type of analogue used? Eligible studies were randomized controlled trials (RCTs), published as a full manuscript in a peer-reviewed journal, that contained sufficient information to allow ascertainment of whether randomization was true and whether equality was present between the groups compared. A literature search identified 22 RCTs comparing GnRH antagonists and GnRH agonists that involved 3176 subjects. Where live birth was not reported in a study that fulfilled the inclusion criteria, an effort was made to contact the corresponding authors to retrieve the missing information. If this was not possible, the reported outcome measure, clinical pregnancy or ongoing pregnancy was converted to live birth in 12 studies using published data (Arce et al., 2005
). No significant difference was present in the probability of live birth between the two GnRH analogues [odds ratio (OR), 0.86; 95% confidence intervals (CI), 0.72 to 1.02]. This result remains stable in subgroup analysis that ordered the studies by type of population studied, gonadotrophin type used for stimulation, type of agonist protocol used, type of agonist used, type of antagonist protocol used, type of antagonist used, presence of allocation concealment, presence of co-intervention and the way the information on live birth was retrieved. In conclusion, the probability of live birth after ovarian stimulation for IVF does not depend on the type of analogue used for pituitary suppression.
Key words: GnRH agonist / GnRH antagonist / live birth / ovarian stimulation
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The introduction of GnRH antagonists in assisted reproductive technologies (ART) at the beginning of this decade was met with mixed reactions. A series of theoretical and evidence-based advantages supported their use (Tarlatzis et al., 2006
). These included the avoidance of an acute stimulation of endogenous gonadotrophins, a dramatic reduction in the length of analogue treatment because of their ability to inhibit directly the premature LH surge and a reduction in the gonadotrophin requirement used for ovarian stimulation. However, the probability of clinical pregnancy was shown to be significantly lower compared with GnRH agonists (Al-Inany and Aboulghar, 2002).
Although the 5% difference in clinical pregnancy (Al-Inany and Aboulghar, 2002) was marginal in terms of clinical importance, it has generated a lot of concern and resulted in a lower acceptance of GnRH antagonists in ovarian stimulation for IVF (Fauser and Devroey, 2005). Not unexpectedly, data from the German registry (Griesinger et al., 2005) showed that GnRH antagonists as compared with GnRH agonists are being prescribed in patients with a worse prognosis, who are older and have performed more IVF cycles.
A better acceptance of GnRH antagonists in ART as a first-choice analogue instead of GnRH agonists might have occurred if, following the initial comparative trials, it had been made clear that the reported 5% significant difference in the probability of clinical pregnancy in favour of GnRH agonists (Al-Inany and Aboulghar, 2002) was not accompanied by a significant difference in live birth per patient randomized.
In the initial five large randomized controlled trials (RCTs), the difference in live birth rate was 3.8%, still in favour of GnRH agonists, but not statistically significant [95% confidence interval (CI), –8.0 to +0.4] (Figure 1). Thus, with respect to live birth as the clinically relevant outcome measure (Arce et al., 2005), a statistically significant difference between the two analogues was never present.
|
Since 2002, following the initial Cochrane meta-analysis on the use of GnRH analogues in IVF (Al-Inany and Aboulghar, 2002), several studies have been published comparing the two GnRH analogues in the general population and in special groups of patients, such as those with poor ovarian response. There is clearly a need to summarize and critically assess the gathered evidence to confirm or deny the initial conclusions reported.
The current systematic review and meta-analysis aims to answer the following clinical question: among patients treated for IVF with gonadotrophins and GnRH analogues, is the probability of live birth per randomized patient dependent on the type of analogue used?
|
Materials and methods |
|---|
|
TOPAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Identification of studies
A literature search was performed using the bibliographic databases MEDLINE, CENTRAL and EMBASE. Additionally, references of retrieved articles and meeting proceedings were hand-searched. The literature search covered the period until December 2005. The search strategy aimed at identifying RCTs from which comparative data on clinical outcome after GnRH agonist or GnRH antagonist usage for ovarian stimulation could be retrieved. The computerized literature search was performed using the terms ‘cetrorelix’ and ‘ganirelix’ as MeSH headings under ‘ovulation induction’ or ‘hormonal therapy’ or ‘infertility therapy’ and limited to ‘human’, ‘female’ and ‘clinical trial’.
Criteria for inclusion/exclusion of studies were established before the literature search. Eligible studies were RCTs that contained sufficient information to allow ascertainment of whether randomization was true and whether equality was present between the groups compared, regarding baseline demographic characteristics, gonadotrophin stimulation protocol, number of embryos transferred and luteal phase support administered. An effort was made to contact all authors or sponsors of studies to retrieve missing or additional information. Language of publication or number of patients analysed was not amongst the exclusion criteria. In contrast to the initial plan, all abstracts had to be excluded when it was found that in most of them an assessment of methodological quality and criteria for inclusion/exclusion was impossible. Thus, identified studies were considered for analysis only if they were published as a full manuscript in a peer-reviewed journal. Selection of studies and methodological quality assessment was performed by two of the authors (G.G. and E.M.K.).
The literature search yielded 175 studies that were potentially able to answer the research question. Further evaluation based on study titles and abstracts and/or assessment of full manuscripts resulted in 22 studies that associated the type of analogue used with the probability of pregnancy/live birth. Where live birth was not reported in a study that fulfilled the inclusion criteria, an effort was made to contact the corresponding authors to retrieve the missing information. If this was not possible, the reported outcome measure, clinical pregnancy or ongoing pregnancy was converted to live birth using published data (Arce et al., 2005): probability of live birth; 84% after confirmation of clinical pregnancy with a living fetus at the 7th week of gestation (Franco et al., 2003; Martínez et al., 2003; Loutradis et al., 2004; Sauer et al., 2004; Lee et al., 2005; Malmusi et al., 2005; Schmidt et al., 2005; Xavier et al., 2005) and 92% after ongoing pregnancy with a viable fetus at the 12th week of gestation (Hohmann et al., 2003; Hwang et al., 2004; Bahceci et al., 2005; Cheung et al., 2005).
The following data were recorded from each of the studies: demographic (type of study, country of origin and period of enrolment), methodological (randomization method, allocation concealment, randomization ratio and whether sample size calculation was performed), procedural (whether financial support was declared, number of patients included, baseline characteristics of the patients included, type and protocol of ovarian stimulation, whether down-regulation was confirmed in the agonist group or basal hormonal status was confirmed in the antagonist group, whether LH was assessed during ovarian stimulation, type of gonadotrophin administered, criteria for hCG administration and dose of hCG, type of fertilization, day of embryo transfer and type of luteal support administered) and outcome data [live birth rate, duration of analogue treatment, FSH requirement, duration of gonadotrophin exposure, incidence of LH rises, incidence of LH + progesterone rises, number of oocytes, incidence of ovarian hyperstimulation syndrome (OHSS)]. Where standard deviation was not reported by the authors, it was calculated from the standard error of the mean (SEM). Where SEM was not reported, it was estimated from the range of values reported (Borm and Mannaerts, 2000; European and Middle East Orgalutran Study Group, 2001; Fluker et al., 2001). Data extraction was performed separately by two of the authors (G.G. and E.M.K.). Any disagreement was solved by discussion.
Primary outcome parameter was the live birth rate (described as delivery rate in some studies). Secondary outcome parameters were duration of GnRH analogue exposure (per cycle), gonadotrophin consumption (per cycle), duration of gonadotrophin exposure (per cycle), incidence of LH rise (per patient randomized), incidence of LH and progesterone rise (LH surge) (per patient randomized), number of cumulus–oocyte complexes (COCs) (per cycle) and incidence of OHSS (grade II–III, grade III, associated with hospital admission per patient randomized). Where the information on the incidence of OHSS associated with hospital admission was not available, the outcome used instead for analysis was that of OHSS classified as ‘severe’ or ‘grade III’.
Study features and results were assembled in tabular form, and a formal meta-analysis was performed. The dichotomous data results for each study were expressed as an odds ratio (OR) with 95% CI. These results were combined for meta-analysis with Comprehensive Meta-analysis software (Biostat, Englewood, NJ, USA), using the Mantel/Haenszel method. Live birth rate, LH surge rate, LH rise rate and OHSS rate were calculated for each study per patient randomized. Information was sought from the corresponding authors regarding live birth in patients who were randomized but did not start treatment within the study period. Where this information was not available, these patients were considered as not pregnant.
When the outcome of interest was of a continuous nature, the differences were pooled across the studies that provided information on this outcome parameter, resulting in a weighted mean difference with 95% CI using Revman Software (The Cochrane Collaboration, 2000).
Study-to-study variation was assessed by using the Cochrane Q-test (the hypothesis tested was that the studies are all drawn from the same population, i.e. from a population with the same effect size). A fixed effects model was used where no heterogeneity was present, whereas, in presence of significant heterogeneity, a random effect model was applied. A funnel plot analysis and Egger’s test were performed to detect any publication bias.
Analysis was carried out to check the stability of the results obtained in subgroups of studies according to the type of population analysed. Four groups of studies were identified—two studies analysed patients with polycystic ovarian syndrome (PCOS) (PCOS group); six studies analysed patients defined as poor responders (poor responder group); four studies used the same criteria as those used in the initial phase III studies (no PCOS, no severe endometriosis, regular cycles, <39 years of age, normal BMI) (phase III trial group, in total nine studies) and, in the remaining five trials, the population analysed could not be classified in one of the above three groups (other group). In addition, subgroup analysis was carried out regarding the type of FSH used (recombinant, urinary or combination), the type of agonist protocol used (long or short), the type of agonist used (Buserelin, Leuprolide, Napharelin or Triptorelin), the type of antagonist used (cetrorelix daily, cetrorelix depot or ganirelix daily) and the type of antagonist protocol used (fixed or flexible). Finally, trials were subgrouped according to randomization quality (allocation concealed or concealment unknown), the presence or not of co-intervention and the way live birth rate was retrieved (reported by the authors or calculated using published data as mentioned above).
A cumulative meta-analysis was also performed to assess the evolving difference in live birth rates. For that purpose, the software program Meta-analysis (Version 1.2.0, 2004, Technopharma S.r.L.) was employed.
It was calculated that the optimal information size required to reject the null hypothesis (no difference in live birth rates between the two GnRH analogues) was 2504 subjects, assuming a clinically important difference of 5%, a baseline live birth rate of 25% and using ß 0.2, 
0.05 and a two-tailed hypothesis test.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Twenty-two RCTs fulfilled the inclusion criteria and were included in the analysis with no disagreement noted between the authors responsible for study selection. Characteristics of the included studies are listed in Tables I and II. All analysed studies were published between 2000 and 2005. The majority of the studies were published in Fertility Sterility (n = 7) and in Human Reproduction (n = 6). The size of the studies ranged from 20 to 730 patients; the median number of patients included was 73, whereas a total of 3176 patients were analysed (GnRH antagonists, n = 1950; GnRH agonists, n = 1226). Seven studies were multi-centre trials, and all others were single-centre studies. In 16 studies, allocation was concealed, whereas, in the remaining studies, concealment of allocation was unclear or not concealed (Bahceci et al., 2005
).
|
|
All the studies were double arm with the exception of four studies (Hohmann et al., 2003; Martínez et al., 2003; Sauer et al., 2004; Lee et al., 2005). In the above studies with multiple arms, the groups compared were combined (Hohmann et al., 2003; Lee et al., 2005), or alternatively, the data were extracted for analysis from the arms that could be compared (Martínez et al., 2003; Sauer et al., 2004).
Financial support by a pharmaceutical company was declared in 11 of the 22 analysed studies. None of the 22 studies was powered to detect differences in the probability of achieving pregnancy. In two studies, the hormonal status of patients randomized before the initiation of stimulation was assessed in the antagonist group, whereas this occurred in the majority of agonist-treated patients.
To inhibit premature LH surges, nine studies used daily ganirelix administration, ten studies used the daily cetrorelix protocol, two studies used the single-dose cetrorelix protocol, and in one study, both the daily and the single-dose cetrorelix protocol were used. In the control groups, 18 studies used a long-agonist protocol, and four studies used a short-agonist protocol.
For ovarian stimulation, three studies used a combination of urinary and recombinant gonadotrophins, seven studies used urinary gonadotrophins, and the remaining 12 studies used recombinant gonadotrophins. Criteria for triggering final oocyte maturation varied markedly across studies and were based on follicular data (n = 14) or on a combination of follicular data and hormonal levels (n = 6), whereas, in two studies, they were not reported. In one study, recombinant hCG was used for triggering final oocyte maturation, whereas the remaining studies used urinary hCG (10 000 IU in 19 studies and 5000–10 000 IU in two studies).
Fertilization methods included both IVF and ICSI, and embryo transfers were performed from 2 to 5 days after oocyte retrieval. Luteal support varied between studies as well as within some multi-centre studies. The majority of the studies analysed used micronized progesterone for luteal support.
In four studies, different interventions were applied in the agonist and the antagonist arm before the initiation of ovarian stimulation: estradiol (E2) pretreatment in the antagonist arm (Franco et al., 2003); oral contraceptive pill (OCP) pretreatment in the agonist arm (Akman et al. 2001; Schmidt et al. 2005); OCP pretreatment in the antagonist arm (Hwang et al., 2004). Moreover, in one study, different criteria were used for triggering final oocyte maturation (Loutradis et al. 2004). These studies were identified as studies with co-intervention, and a subgroup analysis was performed with and without them.
Studies that offered data to answer the research question but did not meet the inclusion criteria are summarized in Tables III and IV. Two studies were excluded because of lack of true randomization (Roulier et al., 2003; Vlaisavljevic et al., 2003).
|
|
The probability of live birth between GnRH agonists and GnRH antagonists was not significantly different (OR, 0.86; 95% CI, 0.72 to 1.02; P = 0.085; heterogeneity, P = 0.99; fixed effect model) (Figure 2). Live birth-rate reduction was –2.7% with GnRH antagonists (95% CI, –5.7 to +0.4; P = 0.087; heterogeneity, P = 0.97, fixed effects model) also not significant. A cumulative meta-analysis on live birth per randomized patient is shown in Figure 3, and a funnel plot of the included studies is shown in Figure 4. No publication bias was present in the studies analysed (Egger indication bias: P = 0.31).
|
|
|
Subgroup analyses of the likelihood of live birth ordered the studies by type of population studied (Figure 5), gonadotrophin type used for stimulation (Figure 6), type of agonist protocol used (Figure 7), type of agonist used (Figure 8), type of antagonist protocol used (Figure 9), type of antagonist used (Figure 10), presence of allocation concealment (Figure 11), presence of co-intervention (Figure 12), the way information on live birth was retrieved (Figure 13). As illustrated in Figures 5–13, the difference in live birth rate between GnRH antagonists and GnRH agonists was not significant in any of the listed subgroup analyses. In none of these subgroup analyses was heterogeneity present, and thus, a fixed effect model was used.
|
|
|
|
|
|
|
|
|
A further analysis that included the two studies that did not meet the inclusion criteria (24 studies, 4309 patients) was not materially changed from the original results (OR, 0.88; 95% CI, 0.76 to 1.02; P = 0.09; heterogeneity, P = 0.99; rate difference, 2.2%; 95% CI, –4.8 to +0.4; P = 0.08; heterogeneity, P = 0.98).
Duration of analogue treatment
As expected, the duration of analogue administration was significantly longer in the agonist group. Seven studies comparing GnRH antagonists with the long GnRH-agonist protocol provided data for this estimation in which a random effects model was used (weighted mean difference (WMD), –19.48 days, 95% CI, –21.05 to –17.91). The inclusion of an additional study comparing a short-agonist protocol with GnRH-antagonist protocol (Martínez et al., 2003) did not change the direction of the original result (WMD, –17.88 days; 95% CI, –20.41 to –15.36).
FSH requirement
The number of FSH ampoules required was not significantly different between the two analogue groups (random effect model; WMD, –3.04 ampoules; 95% CI, –6.27 to +0.19). Seventeen studies offered data for this outcome measure.
Duration of FSH stimulation
The duration of stimulation was significantly longer in the GnRH-agonist group (random effect model; WMD, –1.13 days; 95% CI, –1.83 to –0.44). Fifteen studies offered data for this outcome measure.
COCs retrieved
Significantly more oocytes were retrieved in the agonist group compared with the GnRH-antagonist group (random effect model; WMD, –1.19 COCs; 95% CI, –1.82 to –0.56). Eighteen studies offered data for this outcome measure.
OHSS
The incidence of OHSS associated with hospital admission was significantly lower in the antagonist than in the agonist group (fixed effect model; OR, 0.46; 95% CI, 0.26 to 0.82; P = 0.01). Seven studies offered data for this outcome measure. The incidence of OHSS classified as ‘severe’ (five studies) and that of OHSS classified as grade II–III (three studies) did not differ significantly between the two GnRH analogues compared and was in favour of GnRH-antagonist group, in which the incidence of OHSS was lower (data not shown).
LH surge/rise
The likelihood of LH surges and LH rises was significantly higher with GnRH-antagonist than with GnRH-agonist treatment. For LH rise, the OR was 8.27 (fixed effect model; 95% CI, 3.82 to 17.90; P < 0.001), whereas for LH surge, the OR was 4.05 (fixed effect model; 95% CI, 1.53 to 10.72; P = 0.005). Six studies offered data regarding the incidence of LH rise and LH surge.
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Several reports have been published regarding the optimal outcome measure on which interventions in IVF should be compared (Germond et al., 2004
; Griesinger et al., 2004; Heijnen et al., 2004; Min et al., 2004; Tiitinen et al., 2004; Fauser et al., 2005). Although this issue is still under debate, it is generally agreed that a meaningful outcome measure for patients and clinicians should extend beyond the detection of positive hCG or of clinical and ongoing pregnancy (Arce et al., 2005).
The present meta-analysis compared the effectiveness of GnRH agonist and GnRH antagonists in IVF with respect to the probability of live birth per patient randomized. If live birth was not reported in the studies analysed, the authors were contacted, and information on live birth was retrieved, where this was possible. If the information was not available, the reported outcome measure was converted to a live birth equivalent using the data provided by Arce et al. (2005). This assumes that the probability of live birth after confirmation of clinical pregnancy or ongoing pregnancy does not differ between GnRH agonists and GnRH antagonists. This assumption is supported by evidence from a large set of data reported in the German National IVF Registry (Deutsches IVF Register, 2004).
The rationale for converting reported pregnancy rates to live birth, where such information cannot be retrieved by the authors after communication, is that such a trial can then be included in the analysis. In this way, it is feasible to avoid ignoring those well-designed prospective RCTs, which have used a primary outcome measure other than live birth. It has to be noted that although the importance of using live birth as an end-point in clinical trials in reproductive medicine is well recognized, it is not always feasible to follow patients up to delivery. In the current meta-analysis, 12 of 22 studies would have had to be excluded, if analysis was restricted only to those trials that reported live birth as an outcome measure. On the contrary, focusing on other primary outcome measures such as clinical or ongoing pregnancy might have allowed the inclusion of more trials in the analysis without the need for conversion of the reported pregnancy rates. However, the usefulness of such an approach to the clinician is debatable, as the reported results would still need to be converted to live birth to understand their importance for clinical practice.
The subgroup analysis carried out on the basis of the way live birth was retrieved in the current meta-analysis (reported by the authors or calculated) did not suggest that the conversion of the reported pregnancy rate to live birth using published data (Arce et al., 2005) affected the association between the type of analogue used and the probability of live birth.
The current systematic review and meta-analysis suggests that the probability of live birth is not dependent on the type of GnRH analogue used for suppression of premature LH rise/surge. The finding is stable in several subgroup analyses performed (Figures 5–13), and no statistical heterogeneity was present in the studies analysed regarding live birth rate. The non-significant difference in live birth rate should not be unexpected, because no difference between the two analogues has been demonstrated in terms of embryo (Kol et al., 1999; Seelig et al., 2002; Zikopoulos et al., 2004) or endometrium quality (Simon et al., 2005).
If the observed non-significant difference in live birth rates of 2.7% represents the true difference present in the population, it is likely that this will become significant (P < 0.05) with the addition of more comparative studies. However, even in this case, the clinical importance of such a difference in live birth rates between the two analogues is probably negligible, especially in view of the non-outcome benefits associated with antagonists (Tarlatzis et al., 2006).
A question remaining is whether the live birth rate achieved with GnRH antagonists might in the future supersede that achieved with GnRH agonists. Several prospective studies have been published in recent years with the aim to gather knowledge that could be used to optimize the antagonist protocol (for review see Tarlatzis et al., 2006). However, the stimulation protocols used in the studies included in the meta-analysis did not take advantage of such published experience. It is expected that the availability of evidence-based information on the optimal use of GnRH antagonists will enhance the probability of pregnancy associated with their application in ovarian stimulation.
A significantly higher incidence of premature LH rise/surge in GnRH-antagonist cycles in the present meta-analysis was an unexpected finding. This is in contrast to the widely accepted notion of a comparative efficacy of GnRH analogues in suppressing the premature LH rise/surge (Al-Inany and Aboulghar, 2002). However, the question asked in the Cochrane meta-analysis regarding premature LH inhibition (Al-Inany and Aboulghar, 2002) was whether there is a difference between the two analogues in their efficacy in suppressing the premature LH surge. Although interesting, this question is not the one the clinician is really interested in. The clinical question of interest is whether there is a difference between the two analogue protocols in their effectiveness in suppressing the premature LH surge. Obviously, regarding GnRH antagonists, adequate suppression of premature LH surge is highly dependent on the time GnRH antagonist is initiated, because from that point onwards, the probability of an LH rise/surge is attenuated. Because a significant proportion of the LH rises/surges reported occurred before antagonist initiation, antagonist initiation earlier than day 6 of stimulation is worth evaluation (Kolibianakis et al., 2003, 2004).
It should be noted that the conclusions from the present meta-analysis regarding the LH rise/surge incidence originate from only six studies. Only 13 of 22 studies described the way LH was assessed during ovarian stimulation, whereas in the remaining studies LH assessment was not described (n = 8) or was not performed (n = 1). In addition, in the studies that described LH assessment, it was not always clear what was the timing of LH assessment in relation to antagonist administration. Ideally, antagonist administration should occur immediately after blood is collected for hormonal analysis (Griesinger et al., 2006). In a different case, an LH rise might be missed. Although the adverse effect of premature LH rise (that is suppressed by antagonist initiation) on the probability of pregnancy is not clear, it cannot be excluded that optimization of the GnRH-antagonist protocol probably lies also in the improvement of LH control. It is also necessary that future antagonist studies report detailed data on the incidence of LH rise/surges, whereas the issue of the optimal dose with which GnRH antagonists should be used might need further investigation (Messinis et al., 2005).
The current meta-analysis confirmed existing knowledge regarding the comparison between the two GnRH analogues. Besides a dramatic reduction in the duration of analogue treatment, a significantly reduced duration of FSH stimulation was present in the antagonist group. On the contrary, significantly more oocytes were retrieved in the agonist as compared with antagonist group. In all of these analyses, statistical heterogeneity was detected, and a random effect model was used. The source of heterogeneity is not clear, although it might lie in the clinical heterogeneity among the studies compared. However, the fact that most of the studies for all the above outcomes analysed were in the same direction with the pooled effect is reassuring for the conclusions drawn.
The application of GnRH antagonist in ovarian stimulation for IVF was associated with a significantly lower probability of OHSS associated with hospital admission. The criteria for hospital admission, however, were not mentioned always by the authors. In addition, for this calculation, where data were not provided specifically regarding hospital admission associated with OHSS, the occurrence of OHSS classified as ‘severe’ or as grade III was considered as equal to OHSS associated with hospital admission. Thus, this outcome measure (OHSS associated with hospital admission) should be viewed with caution, and more studies with accurate reporting on OHSS incidence are necessary. Nevertheless, a lower incidence of hospital admission associated with OHSS is not unexpected, because significantly less COCs were retrieved in the antagonist group after a significantly shorter duration of stimulation. Both of the above factors have been related to OHSS occurrence (Delvigne and Rozenberg, 2001, 2002).
An economic analysis will assist in clinical decision making regarding the type of analogue to be preferred in ovarian stimulation for IVF. However, until this is available, the current meta-analysis provides evidence that the choice of GnRH analogue for inhibiting the premature LH surge does not alter significantly the probability of live birth.
|
Acknowledgements |
|---|
|
TOPAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
We thank the following persons for their support in this work: M.A.Akman (Istanbul, Turkey); A.Badrawi (Cairo, Egypt); M.Bahceci (Istanbul, Turkey); M.L.Check (Camden, NJ, USA); L.P.Cheung (Hong Kong, China); H.A.Goolsby (Organon, USA); F.Hohmann (Rotterdam, the Netherlands); K.Gordon (Organon, USA); B.Mannaerts (Organon, the Netherlands); H.Sindermann (Zentaris, Germany); K.Stefanidis (Athens, Greece); R.Marci (Rome, Italy); D.W.Schmidt (Farmington, CT, USA) and V.Vlaisavljevic (Maribor, Slovenia).
|
References |
|---|
|
TOPAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Akman MA, Erden HF, Tosun SB, Bayazit N, Aksoy E and Bahceci M (2001) Comparison of agonistic flare-up-protocol and antagonistic multiple dose protocol in ovarian stimulation of poor responders: results of a prospective randomized trial. Hum Reprod 16,868–870.
Al-Inany H and Aboulghar M (2002) GnRH antagonist in assisted reproduction: a Cochrane review. Hum Reprod 17,874–885.[Abstract]
Albano C, Felberbaum RE, Smitz J, Riethmuller-Winzen H, Engel J, Diedrich K and Devroey P (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 buserelin. European Cetrorelix Study Group. Hum Reprod 15,526–531.
Arce JC, Nyboe Andersen A and Collins J (2005) Resolving methodological and clinical issues in the design of efficacy trials in assisted reproductive technologies: a mini-review. Hum Reprod 20,1757–1771.
Badrawi A, Al-Inany H, Hussein M, Zaki S and Ramzy A (2005) Agonist vs. antagonist in ICSI cycles: a randomized trial and cost-effectiveness analysis. MEFS J 10,49–54.
Bahceci M, Ulug U, Ben-Shlomo I, Erden HF and Akman MA (2005) Use of a GnRH antagonist in controlled ovarian hyperstimulation for assisted conception in women with polycystic ovary disease: a randomized, prospective, pilot study. J Reprod Med 50,84–90.[Web of Science][Medline]
Barmat LI, Chantilis SJ, Hurst BS and Dickey RP (2005) A randomized prospective trial comparing gonadotropin-releasing hormone (GnRH) antagonist/recombinant follicle-stimulating hormone (rFSH) versus GnRH-agonist/rFSH in women pretreated with oral contraceptives before in vitro fertilization. Fertil Steril 83,321–330.[CrossRef][Web of Science][Medline]
Borm G and Mannaerts B (2000) Treatment with the gonadotrophin-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. The European Orgalutran Study Group. Hum Reprod 15,1490–1498.
Check ML, Check JH, Choel JK, Davies E and Kiefer D (2004) Effect of antagonists vs agonists on in vitro fertilization outcome. Clin Exp Obstet Gynecol 31,257–259.[Medline]
Cheung LP, Lam PM, Lok IH, Chiu TT, Yeung SY, Tjer CC and Haines CJ (2005) GnRH antagonist versus long GnRH agonist protocol in poor responders undergoing IVF: a randomized controlled trial. Hum Reprod 20,616–621.
The Cochrane Collaboration (2000) Review manager (RevMan) [4.1]. The Cochrane Collaboration, Oxford.
Delvigne A and Rozenberg S (2001) Preventive attitude of physicians to avoid OHSS in IVF patients. Hum Reprod 16,2491–2495.
Delvigne A and Rozenberg S (2002) Epidemiology and prevention of ovarian hyperstimulation syndrome (OHSS): a review. Hum Reprod Update 8,559–577.
Deutsches IVF Register (2004) Annual report from the German IVF registry 2004. Available from http://www.deutsches-ivf-register.de
European and Middle East Orgalutran Study Group (2001) Comparable clinical outcome 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.
Fauser BC and Devroey P (2005) Why is the clinical acceptance of gonadotropin-releasing hormone antagonist cotreatment during ovarian hyperstimulation for in vitro fertilization so slow? Fertil Steril 83,1607–1611.[CrossRef][Web of Science][Medline]
Fauser BC, Devroey P and Macklon NS (2005) Multiple birth resulting from ovarian stimulation for subfertility treatment. Lancet 365,1807–1816.[CrossRef][Web of Science][Medline]
Fluker M, Grifo J, Leader A, Levy M, Meldrum D, Muasher SJ, Rinehart J, Rosenwaks Z, Scott RT Jr, Schoolcraft W et al. (2001) Efficacy and safety of ganirelix acetate versus leuprolide acetate in women undergoing controlled ovarian hyperstimulation. Fertil Steril 75,38–45.[CrossRef][Web of Science][Medline]
Franco JG Jr, Baruffi LRL, Petersen CG, Mauri CG, Felipe V and Contart P (2003) Comparison of ovarian stimulation with recombinant FSH after 2nd phase protocols with GnRH analogs (I-Estradiol + Ganirelix versus II-Nafarelin). J Brasileiro Reprod Assist 7,27–33.
Germond M, Urner F, Chanson A, Primi MP, Wirthner D and Senn A (2004) What is the most relevant standard of success in assisted reproduction? The cumulated singleton/twin delivery rates per oocyte pick-up: the CUSIDERA and CUTWIDERA. Hum Reprod 19,2442–2444.
Griesinger G, Dafopoulos K, Schultze-Mosgau A, Felberbaum R and Diedrich K (2004) What is the most relevant standard of success in assisted reproduction? Is BESST (birth emphasizing a successful singleton at term) truly the best? Hum Reprod 19,1239–1241.
Griesinger G, Felberbaum R and Diedrich K (2005) GnRH antagonists in ovarian stimulation: a treatment regimen of clinicians’ second choice? Data from the German national IVF registry. Hum Reprod 20,2373–2375.
Griesinger G, Dawson A, Schultze-Mosgau A, Finas D, Diedrich K and Felberbaum R (2006) Assessment of luteinizing hormone level in the gonadotropin-releasing hormone antagonist protocol. Fertil Steril 85,791–793.[CrossRef][Web of Science][Medline]
Heijnen EM, Macklon NS and Fauser BC (2004) What is the most relevant standard of success in assisted reproduction? The next step to improving outcomes of IVF: consider the whole treatment. Hum Reprod 19,1936–1938.
Hohmann FP, Macklon NS and Fauser BC (2003) A randomized comparison of two ovarian stimulation protocols with gonadotropin-releasing hormone (GnRH) antagonist cotreatment for in vitro fertilization commencing recombinant follicle-stimulating hormone on cycle day 2 or 5 with the standard long GnRH agonist protocol. J Clin Endocrinol Metab 88,166–173.
Hwang JL, Seow KM, Lin YH, Huang LW, Hsieh BC, Tsai YL, Wu GJ, Huang SC, Chen CY, Chen PH et al. (2004) Ovarian stimulation by concomitant administration of cetrorelix acetate and HMG following Diane-35 pre-treatment for patients with polycystic ovary syndrome: a prospective randomized study. Hum Reprod 19,1993–2000.
Kol S, Lightman A, Hillensjo T, Devroey P, Fauser B, Tarlatzis B, Mannaerts B and Itskovitz-Eldor J (1999) High doses of gonadotrophin-releasing hormone antagonist in in-vitro fertilization cycles do not adversely affect the outcome of subsequent freeze-thaw cycles. Hum Reprod 14,2242–2244.
Kolibianakis EM, Albano C, Camus M, Tournaye H, Van Steirteghem AC and Devroey P (2003) 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.
Kolibianakis EM, Zikopoulos K, Smitz J, Camus M, Tournaye H, Van Steirteghem AC and Devroey P (2004) Administration of gonadotropin-releasing hormone antagonist from day 1 of stimulation in in vitro fertilization. Fertil Steril 82,223–226.[CrossRef][Web of Science][Medline]
Lee TH, Wu MY, Chen HF, Chen MJ, Ho HN and Yang YS (2005) Ovarian response and follicular development for single-dose and multiple-dose protocols for gonadotropin-releasing hormone antagonist administration. Fertil Steril 83,1700–1707.[CrossRef][Web of Science][Medline]
Loutradis D, Stefanidis K, Drakakis P, Milingos S, Antsaklis A and Michalas S (2004) A modified gonadotropin-releasing hormone (GnRH) antagonist protocol failed to increase clinical pregnancy rates in comparison with the long GnRH protocol. Fertil Steril 82,1446–1448.[CrossRef][Web of Science][Medline]
Malmusi S, La Marca A, Giulini S, Xella S, Tagliasacchi D, Marsella T and Volpe A (2005) Comparison of a gonadotropin-releasing hormone (GnRH) antagonist and GnRH agonist flare-up regimen in poor responders undergoing ovarian stimulation. Fertil Steril 84,402–406.[CrossRef][Web of Science][Medline]
Marci R, Caserta D, Dolo V, Tatone C, Pavan A and Moscarini M (2005) GnRH antagonist in IVF poor-responder patients: results of a randomized trial. Reprod Biomed Online 11,189–193.[Web of Science][Medline]
Martínez F, Coroleu B, Marqués L, Parera N, Buxaderas R, Tur R and Barri PN (2003) Comparación del "Protocolo Corto" versus "Antagonistas" con o sin Citrato de Clomifeno para estimulación en FIV de pacientes con "baja respuesta". Rev Ibero Am Fertil 2025–2360.
Messinis IE, Loutradis D, Domali E, Kotsovassilis CP, Papastergiopoulou L, Kallitsaris A, Drakakis P, Dafopoulos K and Milingos S (2005) Alternate day and daily administration of GnRH antagonist may prevent premature luteinization to a similar extent during FSH treatment. Hum Reprod 20,3192–3197.
Min JK, Breheny SA, MacLachlan V and Healy DL (2004) What is the most relevant standard of success in assisted reproduction? The singleton, term gestation, live birth rate per cycle initiated: the BESST endpoint for assisted reproduction. Hum Reprod 19,3–7.
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][Web of Science][Medline]
Roulier R, Chabert-Orsini V, Sitri MC, Barry B and Terriou P (2003) Depot GnRH agonist versus the single dose GnRH antagonist regimen (cetrorelix, 3 mg) in patients undergoing assisted reproduction treatment. Reprod Biomed Online 7,185–189.[Medline]
Sauer MV, Thornton MH 2nd, Schoolcraft W and Frishman GN (2004) Comparative efficacy and safety of cetrorelix with or without mid-cycle recombinant LH and leuprolide acetate for inhibition of premature LH surges in assisted reproduction. Reprod Biomed Online 9,487–493.[Web of Science][Medline]
Schmidt DW, Bremner T, Orris JJ, Maier DB, Benadiva CA and Nulsen JC (2005) A randomized prospective study of microdose leuprolide versus ganirelix in in vitro fertilization cycles for poor responders. Fertil Steril 83,1568–1571.[CrossRef][Web of Science][Medline]
Seelig AS, Al-Hasani S, Katalinic A, Schopper B, Sturm R, Diedrich K and Ludwig M (2002) Comparison of cryopreservation outcome with gonadotropin-releasing hormone agonists or antagonists in the collecting cycle. Fertil Steril 77,472–475.[CrossRef][Web of Science][Medline]
Simon C, Oberye J, Bellver J, Vidal C, Bosch E, Horcajadas JA, Murphy C, Adams S, Riesewijk A, Mannaerts B et al. (2005) Similar endometrial development in oocyte donors treated with either high- or standard-dose GnRH antagonist compared to treatment with a GnRH agonist or in natural cycles. Hum Reprod 20,3318–3327.
Tarlatzis B, Fauser BC, Kolibianakis EM, Diedrich K and Devroey P (2006) GnRH antagonists in ovarian stimulation for IVF. Hum Reprod Update 12,333–340.
Tiitinen A, Hyden-Granskog C and Gissler M (2004) What is the most relevant standard of success in assisted reproduction?: The value of cryopreservation on cumulative pregnancy rates per single oocyte retrieval should not be forgotten. Hum Reprod 19,2439–2441.
Vlaisavljevic V, Reljic M, Lovrec VG and Kovacic B (2003) Comparable effectiveness using flexible single-dose GnRH antagonist (cetrorelix) and single-dose long GnRH agonist (goserelin) protocol for IVF cycles—a prospective, randomized study. Reprod Biomed Online 7,301–308.[Medline]
Xavier P, Gamboa C, Calejo L, Silva J, Stevenson D, Nunes A and Martinez-de-Oliveira J (2005) A randomised study of GnRH antagonist (cetrorelix) versus agonist (busereline) for controlled ovarian stimulation: effect on safety and efficacy. Eur J Obstet Gynecol Reprod Biol 120,185–189.[CrossRef][Web of Science][Medline]
Zikopoulos K, Kolibianakis EM, Camus M, Tournaye H, Van den Abbeel E, Joris H, Van Steirteghem A and Devroey P (2004) Duration of gonadotropin-releasing hormone antagonist administration does not affect the outcome of subsequent frozen-thawed cycles. Fertil Steril 81,473–475.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  P. Devroey, B.C.J.M. Fauser, K. Diedrich, and on behalf of the Evian Annual Reproduction (EVAR) Approaches to improve the diagnosis and management of infertility Hum. Reprod. Update, July 1, 2009; 15(4): 391 - 408. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Allegra, A. Marino, F. Coffaro, A. Lama, G. Rizza, P. Scaglione, F. Sammartano, A. Santoro, and A. Volpes Is there a uniform basal endometrial gene expression profile during the implantation window in women who became pregnant in a subsequent ICSI cycle? Hum. Reprod., June 20, 2009; (2009) dep222v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B.C.J.M. Fauser, B.M.J.L. Mannaerts, P. Devroey, A. Leader, I. Boime, and D.T. Baird Advances in recombinant DNA technology: corifollitropin alfa, a hybrid molecule with sustained follicle-stimulating activity and reduced injection frequency Hum. Reprod. Update, May 1, 2009; 15(3): 309 - 321. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Nelson, R. W. Yates, H. Lyall, M. Jamieson, I. Traynor, M. Gaudoin, P. Mitchell, P. Ambrose, and R. Fleming Anti-Mullerian hormone-based approach to controlled ovarian stimulation for assisted conception Hum. Reprod., April 1, 2009; 24(4): 867 - 875. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Devroey, M. Aboulghar, J. Garcia-Velasco, G. Griesinger, P. Humaidan, E. Kolibianakis, W. Ledger, C. Tomas, and B. C.J.M. Fauser Improving the patient's experience of IVF/ICSI: a proposal for an ovarian stimulation protocol with GnRH antagonist co-treatment Hum. Reprod., April 1, 2009; 24(4): 764 - 774. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Hayden GnRH analogues: applications in assisted reproductive techniques Eur. J. Endocrinol., December 1, 2008; 159(suppl_1): S17 - S25. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
The Corifollitropin Alfa Dose-finding Study Group A randomized dose-response trial of a single injection of corifollitropin alfa to sustain multifollicular growth during controlled ovarian stimulation Hum. Reprod., November 1, 2008; 23(11): 2484 - 2492. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. S. Macklon, M. H. van der Gaast, A. Hamilton, B. C. J. M. Fauser, and L. C. Giudice The Impact of Ovarian Stimulation With Recombinant FSH in Combination With GnRH Antagonist on the Endometrial Transcriptome in the Window of Implantation Reproductive Sciences, April 1, 2008; 15(4): 357 - 365. [Abstract] [PDF]
|
|
|
|
|
|
J.A. Huirne, R. Homburg, and C.B. Lambalk Are GnRH antagonists comparable to agonists for use in IVF? Hum. Reprod., November 1, 2007; 22(11): 2805 - 2813. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C.A. Venetis, E.M. Kolibianakis, E. Papanikolaou, J. Bontis, P. Devroey, and B.C. Tarlatzis Is progesterone elevation on the day of human chorionic gonadotrophin administration associated with the probability of pregnancy in in vitro fertilization? A systematic review and meta-analysis Hum. Reprod. Update, July 1, 2007; 13(4): 343 - 355. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||