Abstract

Objective:

The aim of this study was to compare and rank the efficacy and safety of four final oocyte maturation trigger strategies—human chorionic gonadotropin (hCG), gonadotropin-releasing hormone agonist (GnRHa), dual, and double trigger—in predicted high responders undergoing in vitro fertilization (IVF) with GnRH antagonist protocols, using a network meta-analysis (NMA) approach.

Methods:

A systematic search of MEDLINE, EMBASE, CENTRAL, clinical trial registries, and the Cochrane Database of Systematic Reviews was conducted through December 2024. Eligible studies were randomized controlled trials (RCTs) including high responders, defined by elevated antral follicle count, anti-Müllerian hormone, or estradiol levels. Studies using GnRHa triggers followed by fresh embryo transfer were included only if intensive luteal phase support was provided. Oocyte donation cycles, quasi-randomized designs, and trials lacking outcome data were excluded. Data extraction and risk of bias assessment were independently conducted by two reviewers. Study integrity was evaluated using the TRACT checklist. NMA was performed in STATA (v16), and treatment ranking was based on Surface Under the Cumulative Ranking curve (SUCRA).

Results:

Seven high-quality RCTs comprising 632 women were included. There were no significant differences in the number of oocytes retrieved between GnRHa and hCG triggers (mean difference [MD] 1.08, 95% CI –1.06 to 3.22), dual and hCG (MD 0.61, 95% CI –1.53 to 2.74), or GnRHa and dual (MD 1.08, 95% CI –1.06 to 3.22). Similarly, there were no significant differences in mature oocyte yield, clinical pregnancy rate (CPR), or miscarriage rate across comparisons. However, GnRHa trigger significantly reduced the risk of moderate to severe ovarian hyperstimulation syndrome (OHSS) compared with hCG (RR 0.23, 95% CI 0.07–0.82). There were no significant differences in OHSS risk between dual and hCG (RR 0.28, 95% CI 0.05–1.64) or between GnRHa and dual (RR 0.28, 95% CI 0.05–1.64).

Conclusion:

GnRHa, hCG, and dual triggers demonstrate similar efficacy in terms of oocyte yield, maturity, and clinical pregnancy rates in predicted high responders. The GnRHa trigger, however, offers a superior safety profile by significantly lowering the risk of OHSS. Larger multicenter RCTs are required to evaluate live birth outcomes and the potential role of the double trigger in this population.

Systematic review registration:

https://www.crd.york.ac.uk/prospero/, identifier CRD42022351423.

Introduction

Final oocyte maturation is physiologically initiated by the pre-ovulatory surge of luteinizing hormone (LH) (1, 2) which initiates a cascade within granulosa cells and oocytes, loss of oocyte-cumulus gap junctions, meiotic resumption to metaphase II, and granulosa-cell luteinization, while the concomitant progesterone rise reinforces estradiol’s positive feedback to elicit the mid-cycle FSH peak (3, 4), The associated physiological FSH surge, in turn, contributes to up-regulation of LH receptors on granulosa cells and synthesis of a hyaluronic-acid matrix, enabling cumulus expansion and dispersion so that the oocyte–cumulus complex becomes free-floating within the antrum (5–8).

Traditionally, in a cycle of in-vitro fertilization (IVF), human chorionic gonadotropin (hCG) has been a gold standard substitute for the endogenous LH surge due to its structural similarity to LH and its ability to bind and activate the LH/hCG receptors thereby triggering the resumption of meiosis (9). Standard dosing includes urinary hCG 5,000–10,000 IU subcutaneously/intramuscularly or recombinant hCG 250 µg subcutaneously (≈6,500 IU bioactivity), administered ~36 hours before oocyte retrieval (8). However, the extended luteotropic activity caused by hCG administration (4, 10), combined with elevated levels of estradiol and progesterone, can result in ovarian hyperstimulation syndrome (OHSS), particularly in high responders, requiring a postponement of embryo transfer (11–13).

In 1990, Gonen et al. (14) demonstrated that ovulation can also be induced using a gonadotropin-releasing hormone (GnRH) agonist trigger. By displacing the GnRH antagonist from the receptor, the GnRHa trigger causes a flare-up of both LH and FSH, similar to the natural cycle, and therefore regarded as a more physiological trigger for ovulation (5, 11). In 2000, Itskovitz-Eldor et al. (15) reported the first series of patients at high risk of developing severe OHSS who underwent controlled ovarian hyperstimulation (COH) using a GnRH antagonist protocol with a GnRHa trigger for final follicular maturation. While 50% of the patients achieved conception, none developed clinical or biochemical signs of OHSS. This is because, unlike hCG, GnRH agonists have a shorter half-life, leading to rapid luteolysis and significantly reducing OHSS risk (16). As a result, the use of GnRHa trigger, within a GnRH antagonist cycle, has become a cornerstone in achieving OHSS-free IVF practice (16).

Building on these observations, multiple studies have compared hCG with GnRHa for final oocyte maturation in IVF (17, 18). Across reports, the number of oocytes retrieved, the proportion of metaphase II oocytes, and the yield of top-quality embryos were generally comparable and in several studies favored the GnRHa trigger (19, 20). However, their use has been associated with significantly lower clinical pregnancy rates and higher first-trimester pregnancy losses (12, 21). This was due to the associated luteal phase insufficiency, marked by lower progesterone levels, which can compromise the peri-implantation period and result in higher early pregnancy loss rates following fresh embryo transfer (22–27). Various strategies were therefore explored to optimize reproductive outcomes to modify luteal phase support by enhancing corpus luteum function (25, 28). One proposed approach involves the co-administration of GnRHa with a standard hCG bolus (5,000–10,000 IU) before oocyte retrieval to enhance final follicular maturation, with the aim of improving oocyte competence, embryo quality, and overall IVF outcomes while addressing the luteal insufficiency associated with GnRHa administration alone (11, 28).

The dual trigger involves the co-administration of GnRH-a with low-dose hCG (≈1,000–2,500 IU) given simultaneously at trigger (29) and the double (staggered) trigger administers GnRH-a ~40 h and hCG ~34 h before retrieval, a timing variant conceived to optimize final maturation and cumulus expansion in selected contexts (24). Their mechanisms and clinical advantages however differ. The dual trigger believed to more closely mimic the physiological mid-cycle surge, has been reported to result in improved oocyte maturation, embryo quality, and pregnancy outcomes without increasing the risk of OHSS (30, 31) while the extended LH surge provided by the double trigger staggered approach has been reported to result in an improvement in follicular synchronization, thereby enhancing oocyte yield and maturity in patients with previous suboptimal responses or empty follicle syndrome (32, 33).

Although prior individual randomized controlled trials have performed pairwise comparison assessing the effectiveness of individual protocols such as hCG, GnRHa, dual, or double triggers to each other, there has not been a comprehensive analysis simultaneously evaluating all four trigger protocols in high responder patients. Evidence syntheses have also typically been pairwise, often mixing heterogeneous patient populations or not isolating predicted high responders. As a result, there is a lack a comprehensive, hierarchical comparison of all four trigger protocols specifically in high-risk patients and there remains a lack of consensus on the optimal triggering protocol for high responders where trigger choice has direct implications for safety, embryo transfer strategy, and time to pregnancy.

In light of these considerations, the objective of our study is to address this gap in research literature and conduct a systematic review of existing high integrity evidence and perform a network meta-analysis to simultaneously assess the efficacy and safety of the hCG trigger alone, GnRHa trigger alone, dual trigger, and double trigger protocols in predicted high responder patients.

MethodsSearch and study selection

We adhered to the PRISMA extension statement for reporting systematic reviews with network meta-analyses of healthcare interventions in reporting this systematic review (34). The protocol was registered on PROSPERO: International Prospective Register of Systematic Reviews, with registration number CRD42022351423.

We conducted a systematic search, from their inception up to December 30, 2024, across several databases including MEDLINE, EMBASE, Cochrane Database of Systematic Reviews and Cochrane Central Register of Controlled Trials (CENTRAL) which included trials registered on clinicaltrials.gov as well as the trial registry of the World Health Organization and the Cochrane Gynaecology and Fertility specialised register. Two authors (Y.B. and J.M.N.D) examined the identified studies independently for compliance with inclusion criteria and selected eligible studies. They resolved disagreements by discussing with a third author (S.S.) and documented the selection process with a ‘PRISMA’ flowchart.

The search was specifically focused on RCTs which compared different triggers for final oocyte maturation—hCG, GnRHa, dual trigger, or double trigger—in patients defined as high responders and undergoing ovarian stimulation for IVF/ICSI within GnRH antagonist cycles. Both free text and index terms were utilized for the search, as detailed in Supplementary Appendix 1.

Study eligibility

Only randomized trials were eligible. Cross-over, quasi-randomized, and non-randomized trials were excluded, as well as studies involving oocyte donation transfer cycles. Studies were restricted to patients predicted to be high responders defined as per at least one of the following criteria: a normal FSH level with a history of PCOS as per Rotterdam criteria (35, 36), number of antral follicles more than 20, anti-Mullerian hormone level >3.5 ng/mL, number of follicles >18 mm more than 20, or estradiol level more than 4000 pg/mL (37). These definitions were used in place of any newer criteria which were not adopted when the previous RCTs were published. Where necessary, additional trial details or protocols were sought to determine potential study eligibility.

Studies involving GnRH antagonist protocols and utilizing GnRHa trigger alone followed by fresh transfers were included only if they incorporated intensified luteal progesterone support (LPS) according to the previously defined criteria and consisted of one of the following: micronized progesterone vaginally, estradiol supplementation trans-dermally or vaginally (19, 23, 26, 38) daily low dose hCG administration during the luteal phase without exogenous progesterone (39), a single bolus of hCG given after oocyte retrieval (40, 41), recombinant LH (Luveris, Merck-Serono) starting on the day of oocyte retrieval up to day 10 after oocyte retrieval (32).

Risk of bias, trustworthiness, and overall certainty of evidence assessments

Two authors (Y.B. and J.M.N.D.) independently gathered data from the included studies and assessed the risk of bias utilizing the Cochrane Collaboration’s risk of bias assessment tool. Any discrepancies were resolved through discussion with a third author (S.S.). We evaluated selection bias, performance bias, detection bias, attrition bias, reporting bias, and other biases.

Identified RCTs were further evaluated for trustworthiness using the Trustworthiness in Randomized Controlled Trials (TRACT) checklist (42) which assessed 19 items organized into seven domains that are applicable to every RCT: 1) Governance, 2) Author Group, 3) Plausibility of Intervention Usage, 4) Timeframe, 5) Drop-out Rates, 6) Baseline Characteristics, and 7) Outcomes [32]. Only studies that met high integrity standards according to the TRACT checklist were included. Authors were contacted for clarification regarding these discrepancies and reasons for exclusion included lack of prospective clinical trial registration. The overall quality of evidence was also assessed using the web application Confidence in Network Meta-Analysis (CINeMA), which is based on the Grading of Recommendations, Assessment, Development, and Evaluations framework (43). (Supplementary Table 1).

Outcomes

In addition to reporting the recommended core outcome set, developed for infertility research (44, 45), such as clinical pregnancy, live birth and pregnancy loss, we also included non-core outcome outcomes such as number of oocytes collected, number of mature oocytes retrieved and rate of OHSS per patient randomized per stimulation cycle. We aimed to extract data for other core outcomes such as gestational age at delivery; birthweight; neonatal mortality; major congenital anomaly and time to live birth but these were not reported.

Clinical pregnancy was defined as a pregnancy diagnosed by ultrasonographic visualization of one or more gestational sacs or definitive clinical signs of pregnancy (46). Live birth was defined as a birth in which a fetus is delivered with signs of life after complete expulsion or extraction from its mother (46). Miscarriage was defined as spontaneous loss of an intrauterine pregnancy prior to twenty-two completed weeks of gestational age (46). A mature oocyte was defined as an oocyte at metaphase of meiosis II, exhibiting the first polar body and with the ability to become fertilized (46). Ovarian hyperstimulation was defined as an exaggerated systemic response to ovarian stimulation characterized by a wide spectrum of clinical and laboratory manifestations (46).

As outlined in our PROSPERO registration, live birth was initially identified as a primary outcome. However, due lack of reporting and available evidence, it was not possible to perform a comprehensive network meta-analysis on this outcome. Our analysis focused on laboratory outcome (oocyte yield, mature oocyte yield), clinical outcomes (clinical pregnancy and miscarriage rates) and rate of OHSS.

Data synthesis and analysis

Statistical analysis for pairwise was performed using STATA version 16 (STATA Corp., College Station, TX, USA). We used network plots to visually represent all direct comparisons among the randomized controlled trials (RCTs) included in our study (47). To assess potential inconsistency, we employed a design-by-treatment interaction model (48). If no inconsistency was identified, we performed pairwise and network meta-analyses using a random-effects model in STATA, utilizing the ‘network’ command for network meta-analysis and other STATA commands for visualizing and reporting results (49). Effect estimates are presented as risk ratios (RR) for dichotomous outcomes and mean differences (MD) for continuous outcomes, with 95% confidence intervals (CI).

We used the Surface Under the Cumulative Ranking (SUCRA) methodology to establish a hierarchy among the interventions considered in our study (50). Notably, a higher SUCRA value indicates a more favorable treatment rank. For example, if the SUCRA for treatment A is 100%, all other treatments are considered inferior to A, indicating A as the best.

In addition, a post-hoc power analysis was carried out to assess the detectability of the observed differences given the sample sizes available. This analysis was performed using G*Power (43), employing a two-tailed z-test for the difference between two independent proportions. We set the significance level (α) at 0.05 and used the actual sample sizes from our included studies as the basis for the analysis. The analysis computed the critical z-value and the achieved power (1 - β error probability), which reflects the probability of correctly detecting an effect of the magnitude observed if it truly exists. A z-value exceeding ±1.96 was considered statistically significant at the 5% level (two-tailed), indicating that the observed difference between intervention groups would be unlikely to have occurred by chance.

ResultsStudy selection and characteristics

The initial electronic database search yielded 6,179 records. After excluding 3,043 duplicate records, 3,136 titles and abstracts were screened for potential inclusion. 3,022 were further excluded and 114 full text articles were further assessed for eligibility. Nine RCTs, reporting data from 961 participants met our inclusion criteria for integrity assessment (Figure 1) (22, 23, 37, 38, 51–55).

PRISMA 2020 flow diagram showing identification, screening, eligibility, integrity assessment, and inclusion of randomised controlled trials for the systematic review and network meta-analysis.

Reports with concerns about trustworthiness included two full-text publications. We contacted the authors of these studies for clarification and they were ultimately excluded from the final analysis due to a lack of prospective clinical trial registration and lack of information to assess discrepancies of  > 15% between the intended sample size in the trial registration compared to the actual sample size achieved in the trial (53, 55).

Seven RCTS, representing 632 IVF/ICSI participants were finally included in our systematic review, pair wise meta-analysis and network meta-analysis (22, 23, 38, 51, 52, 54, 56). Publication dates ranged from 2006 to 2022, four studies published within the last 10 years (37, 38, 41, 54), conducted in six countries. The study characteristics of the included studies are listed in Table 1.

Author (year)CountryDesign/populationTrigger 1 (intervention)Dose/timingOutcomeN/mean and SDTotal number of participant in arm 1Trigger 2 (control)Dose / TimingOutcomenTotal number of participant in arm 2Aghahosseini et al. (2017) (37)IranRCT – PCOS / high riskDual triggerBuserelin 0.5 mL + hCG 1 500 IU (simultaneous administration)CPR1840GnRHa onlyBuserelin 0.5 mL + hCG 1 500 IU 35 h later for luteal supportCPR1940Number of oocytes26.5 (8.9)Number of oocytes24.4 (8.5)OHSS13OHSS5Babayof et al. (2006) (22)IsraelRCT – PCO patientsGnRHa triggerTriptorelin 0.2 mg (Decapeptyl)OHSS015hCG trigger250 IU OvitrelleOHSS413Miscarriage2Miscarriage2Number of mature oocytes18 (2.9)Number of mature oocytes16 (2.2)Elgindy et al. (2018) (41)EgyptRCT – High responder ICSIGnRHa triggerTriptorelin 0.3 mg + 1 500 IU hCG at retrieval for luteal supportOHSS1095hCG trigger10 000 IU + standard luteal supportOHSS2395CPR51CPR53Number of mature oocytes14.9
(5.2)Number of mature oocytes15.6
(5.2)Engmann et al. (2008) (23)USARCT – High risk (PCOS / previous high response)GnRHa triggerLeuprolide 1 mgCPR1730hCG trigger3 000–10 000 IUCPR1529Number of oocytes20.2 (9.9)Number of oocytes18.8 (6.3)OHSS0OHSS10Miscarriage3Miscarriage4Number of mature oocytes24 (16.3)Number of mature oocytes19 (6.3)Engmann et al. (2019) (38)USADouble-blind RCT – High respondersDual triggerGnRHa (Leuprolide 1 mg) + hCG 1 000 IU (simultaneous administration)CPR1526GnRHa onlyLeuprolide acetate 1 mg
hCG 1 500 IU 35 h later for luteal supportCPR3331Number of oocytes18.4 (5.9)Number of oocytes18.4 (7.7)Miscarriage5Miscarriage6Humaidan et al. (2013) (51)Denmark / Belgium / GreeceMulticentre RCT – patients at high OHSS riskGnRHa triggerBuserelin 0.5 mg + 1 500 IU hCG post-retrieval for luteal supportCPR2160hCG trigger5 000 IUCPR1758Number of oocytes13.7 (5.9)Number of oocytes13.5 (5.7)OHSS0OHSS2Miscarriage4Miscarriage4Valipour et al. (2022) (54)IranRCT – High risk (antagonist cycles)Dual triggerDecapeptyl 0.2 mg + hCG 2 500 IU (simultaneous)CPR3250hCG trigger10 000 IUCPR1650Number of oocytes10.4 (4.1)Number of oocytes8.7 (3.3)OHSS0OHSS3

Summary of included randomised controlled trials comparing ovulation-trigger strategies in high-responder IVF patients.

SD, standard deviation; GnRHa, gonadotropin-releasing hormone agonist; hCG, human chorionic gonadotropin; OHSS, ovarian hyperstimulation syndrome; PCO(S), polycystic ovary (syndrome); CP, clinical pregnancy rate; OPR, ongoing pregnancy rate; LBR, live birth rate; NS, not significant.

Four studies totaling 395 participants compared GnRHa trigger to hCG trigger (22, 23, 41, 51), one study totaling 100 participants compared dual trigger to hCG trigger (54) and two studies consisting of 137 women compared GnRHa trigger to dual trigger (37, 38). There was no published RCT assessing the use of double trigger in women defined as being high responders.

Risk of bias assessment

The risk of bias observed in the included studies ranged from low to high based on the Cochrane risk-of-bias tool. The quality assessment of the included studies is illustrated in Supplementary Figure 1.

Network and pairwise meta-analysesLaboratory outcomesNumber of oocytes

Five studies representing 414 participants reporting on number of oocytes retrieved were identified as being of high integrity and included (23, 37, 51, 54, 57). Pair wise meta-analysis, network plot and the results of the network meta-analysis are displayed in Supplementary File 1.

GnRH agonist trigger vs hCG trigger

Two studies compared GnRH agonist vs. hCG trigger in 177 participants (23, 51). There was insufficient evidence of a difference based on pairwise meta-analysis (MD 0.44, 95% CI: -1.44 to 2.31, I2 = 0%) and network meta-analysis (MD 1.08, 95% CI: -1.06 to 3.22). Post hoc power analysis yielded a critical z-value of approximately 1.97 and an achieved power of only 7.41%.

Dual trigger vs hCG trigger

One study compared dual vs. hCG trigger in 100 participants (54). There was insufficient evidence of a difference based on pairwise analysis (MD 1.70, 95% CI:0.24 to 3.16) and network meta-analysis (MD 0.61. 95% CI -1.53 to 2.74).

GnRH agonist trigger vs dual trigger

Two studies of 137 participants compared GnRH agonist vs. dual trigger (37, 57). There was insufficient evidence of a difference based on pairwise analysis (MD 0.97, 95% CI: -1.62 to 3.56, I2 = 0%) and network meta-analysis (MD 1.08. 95% CI -1.06 to 3.22). Post hoc power analysis yielded a critical z-value of approximately 1.97 and an achieved power of only 11.7%.

In the network meta-analysis, GnRH agonist trigger ranked highest for number of oocytes (SUCRA 76.1%), followed by dual trigger (51.5%) and hCG trigger (22.5%).

Number of matured oocytes

Six studies representing 532 participants reporting on number of mature oocytes retrieved were identified as being of high integrity and included (22, 23, 37, 38, 41, 51). Figure 2 presents the network plot for number of mature oocytes retrieved and the results of the network meta-analysis. Pair wise meta-analysis is displayed in Supplementary File 1.

Network meta-analysis of number of mature oocytes in predicted high responders undergoing IVF antagonist cycles. (A)Network forest plot showing mean differences (MD) with 95% confidence intervals for number of mature oocytes between interventions. Positive values favour the first-listed intervention. (B)Network geometry illustrating direct and indirect evidence comparing final oocyte-maturation triggers: human chorionic gonadotropin (hCG), gonadotropin-releasing hormone agonist (GnRHa), and dual trigger (GnRHa low-dose hCG). Node size is proportional to the number of participants assigned to each intervention, and line thickness represents the number of trials providing direct comparisons. (C)SUCRA ranking curves displaying the relative probability of each intervention ranking among the most effective for number of mature obcytes.

GnRH agonist trigger vs hCG trigger

Four studies compared GnRH agonist vs. hCG trigger in 395 participants (22, 23, 41, 51). There was insufficient evidence of a difference based on pairwise meta-analysis (MD 1.11, 95% CI: -1.74 to 3.97, I2 = 76%, p=0.02) and network meta-analysis (MD -1.26, 95% CI: -4.03 to 1.52). Post hoc power analysis yielded a critical z-value of approximately 1.96 and an achieved power of only 12.5%.

Dual trigger vs hCG trigger

Through indirect evidence only, there was no difference in number of mature oocytes retrieved (MD 1.43, 95% CI: -1.47 to 4.33).

GnRH agonist trigger vs dual trigger

Two studies compared GnRH agonist vs. dual trigger in 137 participants (37, 38). There was insufficient evidence of a difference based on pairwise meta-analysis of these studies (MD 6.01, 95% CI: -2.40 to 14.41, I2 = 80%, p=0.03) and network meta-analysis (MD -1.26, 95% CI: -4.03 to 1.52). Post hoc power analysis yielded a critical z-value of approximately 1.97 and an achieved power of only 29.1%.

In the network meta-analysis, dual trigger ranked highest for number of mature oocytes (SUCRA 82.0%), followed by GnRH agonist (SUCRA 40.4%) and hCG trigger (SUCRA 27.5%).

Clinical outcomesClinical pregnancy rates

Five studies representing 414 participants reporting on clinical pregnancy rates (CPR) (23, 37, 51, 54, 57). Figure 3 presents the network plot for clinical pregnancy rates and the results of the network meta-analysis. Pair wise meta-analysis is displayed in Supplementary File 1.

Network meta-analysis of clinical pregnancy rates in predicted high responders undergoing IVF antagonist cycles. (A)Network forest plot showing risk ratios (RR) with 95% confidence intervals for clinical pregnancy across interventions. Positive values favour the first-listed intervention. (B)Network geometry illustrating the evidence base comparing human chorionic gonadotropin (hCG), gonadotropin-releasing hormone agonist (GnRHa), and dual trigger (GnRHa + low-dose hCG) for final oocyte maturation. Node size corresponds to the total number of participants randomised to each intervention, and edge thickness represents the number of direct comparisons available between treatments. (C)SUCRA ranking curves displaying the relative probability of each intervention ranking among the highest for clinical pregnancy ra