Oocyte maturation triggers in high-responders: a report on 1,217 consecutive cycles.

This study was funded by UMIB-Unit for Multidisciplinary Research in Biomedicine (grant numbers UIDB/00215/2020 and UIDP/00215/2020); ITR-Laboratory for Integrative and Translational Research in Population Health (grant number LA/P/0064/2020).

Considering the type of oocyte maturation trigger, cycles were divided into five groups ( Table 1 ). The 232 cycles using an hCG-trigger consisted of 208 fresh ETCs, with luteal support performed with progesterone since oocyte pick-up day (subgroup hCG-fresh), and 18 cycles with intended fresh embryo transfer that were later transformed into freeze-all (FA) cycles (subgroup hCG-FA). In the hCG-trigger group there were 127 FET cycles, 107 with surplus embryos (from fresh ETC) and 20 with embryos from FA cycles. The 473 cycles using an agonist trigger consisted of 327 fresh ETC with luteal support initiated at oocyte pick-up day with a small-dose bolus of hCG/rhCG besides estrogen and progesterone (subgroup Ag-fresh-hCG), 47 cycles intended for fresh embryo transfer, with luteal support initiated at oocyte pick-up day with a small-dose bolus of hCG/rhCG besides estrogen and progesterone, but that were later transformed into FA cycles (subgroup Ag-hCG-FA), and 83 cycles without luteal supplementation that were directed to FA (subgroup Ag-FA). In the agonist group, there were 407 FET cycles, 195 with surplus embryos (from fresh ETC), 83 with embryos from Ag-hCG-FA cycles, and 129 from Ag-FA cycles. Fresh embryo transfers and freeze-all cycles considering the type of oocyte trigger and embryo destination. Demographic and infertility characteristics. Data are presented as mean ± SD or as indicated. Comparison between all groups on the same variable is indicated by a P value. Significance set at P < 0.05. hCG-fresh, cycles using hCG as trigger, with progesterone for luteal support, followed by fresh embryo transfer; hCG-FA, cycles using hCG as trigger, with progesterone for luteal support, followed by freeze-all (freezing of all embryos); Ag-fresh-hCG, cycles using an agonist as trigger, with hCG, estradiol, and progesterone for luteal support, followed by fresh embryo transfer; Ag-hCG-FA, cycles using an agonist as trigger, with hCG, estradiol, and progesterone for luteal support, followed by freeze-all; Ag-FA, cycles using an agonist as trigger, without luteal support, followed by freeze-all; FA, freeze all embryos; ETC, embryo transfer cycles; ages (years); time of infertility (years); HPRL, hyperprolactinemia; PCOS, polycystic ovary syndrome; bAMH, basal anti-Müllerian hormone; bFSH, basal follicle-stimulating hormone; bLH, basal luteinizing hormone; BMI, body mass index; n.a., not applicable. When a low risk of OHSS was perceived, an hCG-trigger was used and a fresh embryo transfer was performed; however, some cases showed later signs of increased OHSS risk and thus embryos went FA. When a higher risk was perceived from the beginning, a GnRHa-trigger was used; in most of the cases there were no signs for development of severe OHSS, and an hCG bolus at oocyte pick-up was additionally added to estrogen and progesterone luteal support and a fresh embryo transfer was performed; however, in some cases, later signs of increased OHSS risk appeared and thus embryos underwent FA. In another series of cases, a very high risk of OHSS was present from the beginning, no luteal support was performed, and embryos underwent FA. The demographic differences found when the five groups were compared were mainly due to differences observed between the hCG-fresh and Ag-fresh-hCG subgroups. In these comparisons it was observed that the use of an agonist trigger was associated with a significantly lower mean female age and lower endometriosis rate, and a significantly higher rate of ovulatory dysfunction and PCOS, and significantly higher mean anti-Müllerian hormone (AMH) levels ( Table 1 ). AMH levels were also higher in the Ag-FA and Ag-hCG-FA subgroups. Most of the cycles in all five groups were performed with ICSI (55.6–75.9%) and ejaculated sperm (72.3–100%). The Ag-fresh-hCG subgroup showed a significantly higher mean number of follicles when compared to the hCG-fresh group, and this number was also higher in Ag-FA and Ag-hCG-FA subgroups. There was a significantly lower mean total gonadotropin dose with the Ag-fresh-hCG subgroup than with hCG-fresh. The differences found in the time of stimulation were due to differences when comparing the Ag-fresh-hCG and the Ag-FA subgroups. Significantly higher mean E2 and P4 levels were observed in the FA subgroups ( Table 2 ). Stimulation characteristics of fresh embryo transfers and freeze-all cycles considering the type of oocyte trigger and embryo destination. Data are presented as mean ± SD. Comparison between all groups on the same variable is indicated by a P value. Significance set at P < 0.05. hCG-fresh, cycles using hCG as trigger, with progesterone for luteal support, followed by fresh embryo transfer; hCG-FA, cycles using hCG as trigger, with progesterone for luteal support, followed by freeze-all (freezing of all embryos); Ag-fresh-hCG, cycles using an agonist as trigger, with hCG, estradiol, and progesterone for luteal support, followed by fresh embryo transfer; Ag-hCG-FA, cycles using an agonist as trigger, with hCG, estradiol, and progesterone for luteal support, followed by freeze-all; Ag-FA, cycles using an agonist as trigger, without luteal support, followed by freeze-all. Estradiol and progesterone measured 1 day before or at trigger day. With hCG-triggering, there were nine OHSS cases (five, 2.4% in subgroup hCG-fresh; four, 22.2% in subgroup hCG-FA), of which three (1.4%) required hospitalization in the subgroup with fresh ETC. With agonist triggering, there were 15 OHSS cases (eight, 2.4% in subgroup Ag-fresh-hCG; four, 8.5% in subgroup Ag-hCG-FA; three, 3.6% in subgroup Ag-FA), of which three (0.9%) required hospitalization in the subgroup with fresh ETC. Severe OHSS only occurred in fresh ETC, with the agonist trigger showing a non-significantly lower rate. There was a significantly higher mean number of COC and mature oocytes in FA subgroups. No significant differences were observed regarding the rates of maturation, fertilization, embryo cleavage, and high-quality embryos at day 3. The differences found in blastocyst rates between subgroups were due to higher blastocyst rates in the three FA subgroups ( Table 3 ). Embryological outcomes of fresh embryo transfers and freeze-all cycles considering the type of oocyte trigger and embryo destination. Comparison between all groups on the same variable is indicated by a P value. Significance set at P < 0.05. hCG-fresh, cycles using hCG as trigger, with progesterone for luteal support, followed by fresh embryo transfer; hCG-FA, cycles using hCG as trigger, with progesterone for luteal support, followed by freeze-all (freezing of all embryos); Ag-fresh-hCG, cycles using an agonist as trigger, with hCG, estradiol, and progesterone for luteal support, followed by fresh embryo transfer; Ag-hCG-FA, cycles using an agonist as trigger, with hCG, estradiol, and progesterone for luteal support, followed by freeze-all; Ag-FA, cycles using an agonist as trigger, without luteal support, followed by freeze-all; COC, cumulus–oocyte complexes (aspirated oocytes); MII, mature oocytes; 2 PB/2 PB, normally fertilized oocytes displaying two pronuclei (PN) and two polar bodies (PB); HQ, high-quality embryos. Statistical comparisons between (A) vs (B), (A) vs (D), (A) vs (E), (C) vs (D), and (C) vs (E) were performed. The Ag-fresh-hCG subgroup showed a significantly lower mean number of transferred embryos and a lower twin pregnancy rate, and significantly higher rates of implantation and CP when compared to the hCG-fresh subgroup, with no significant differences regarding the rates of miscarriage, LBD, and NB, mean gestational age, or newborn weight ( Table 4 ). Clinical outcomes of fresh embryo transfers and freeze-all cycles considering the type of oocyte trigger and embryo destination. Statistical comparison between hCG-fresh and Ag-fresh-hCG is indicated by a P value. Significance set at P < 0.05. hCG-fresh, cycles using hCG as trigger, with progesterone for luteal support, followed by fresh embryo transfer; Ag-fresh-hCG, cycles using an agonist as trigger, with hCG, estradiol, and progesterone for luteal support, followed by fresh embryo transfer; n.a., not applicable. In the FA subgroups, the tendentially higher rates of implantation and CP were not found to be significantly superior when compared to the other subgroups. No significant differences were observed for the number of transferred embryos, rates of LBD and NB, mean gestational age, or newborn weight ( Table 5 ). Frozen-thawed embryo transfer considering embryo origin with cumulative clinical outcomes. Data are presented as mean ± SD or as indicated. Comparison between all groups on the same variable is indicated by a P value. Significance set at P < 0.05. hCG-fresh, cycles using hCG as trigger, with progesterone for luteal support, followed by cryopreservation of surplus embryos and posterior frozen embryo transfer; hCG-FA, cycles using hCG as trigger, with progesterone for luteal support, followed by freeze-all (freezing of all embryos) and posterior frozen embryo transfer; Ag-fresh-hCG, cycles using an agonist as trigger, with hCG, estradiol, and progesterone for luteal support, followed by cryopreservation of surplus embryos and posterior frozen embryo transfer; Ag-hCG-FA, cycles using an agonist as trigger, with hCG, estradiol, and progesterone for luteal support, followed by freeze-all and posterior frozen embryo transfer; Ag-FA, cycles using an agonist as trigger, without luteal support, followed by freeze-all and posterior frozen embryo transfer; ETC, fresh embryo transfer cycle; FET, frozen embryo transfer; n.a., not applicable. Indicates the sum of hCG-fresh and hCG-FA embryos as well as the sum of Ag-fresh-hCG, Ag-hCG-FA and Ag-FA embryos. Statistical comparisons between (A) and (C), and (B) and (C) were done. Statistical comparisons between (A) and (C), and (C) and (E) were done. The Ag-fresh-hCG subgroup presented significantly higher rates of CP and NB when compared to the hCG-fresh subgroup. No significant differences were observed regarding the rates of LBD ( Table 5 ). In the hCG-fresh subgroup, the CP (51.9 fresh vs 40.2% FET) and LBD (43.3 fresh vs 33.6% FET) rates were compared between fresh and frozen-thawed embryo transfers, with significant differences being observed for CP ( P = 0.048) but not for LBD. In the Ag-fresh-hCG subgroup, the CP (60.9 fresh vs 44.1% FET) and LBD (48.9 fresh vs 35.4% FET) rates were compared between fresh and frozen-thawed embryo transfers, with significant differences being observed for CP ( P < 0.001) and LBD ( P = 0.003). Although the clinical pregnancy rate (odds ratio (OR) = 1.44, 95% confidence interval (CI): 1.013–2.045, P = 0.049) and the cumulative pregnancy rate (OR = 1.60, 95% CI: 1.100–2.317, P = 0.014) were significantly higher with the Ag-fresh-hCG protocol than with the hCG-fresh protocol, after adjusting for possible confounders such as female age, AMH, PCOS, estradiol, and number of follicles, these outcomes were no longer associated with the protocol. No associations with the protocol were found for live birth, independently of adjusting for the above-described factors, either for fresh ETC or FET cycles.

According to the determinations of the National Law of Medically Assisted Procreation (Law of 2006) and guidelines of the National Council for Medically Assisted Procreation (CNPMA-2021), the use of patient clinical databases for research may be used without additional ethical approval, as long as data are used under strict individual anonymity and after informed and written consent from the patients. The use of these data was further authorized by the Academic and Hospital Ethics Committee with project number: 2024/CE/P23 (P434/2024/CETI). This work did not involve human or animal experiments, and thus the provisions of the Declaration of Helsinki as revised in Tokyo 2004 do not apply to this work. In this retrospective observational cohort study, the studied population consisted of 652 patients with high-responder criteria selected from 11 consecutive years (2012–2023). The selected population of infertile women considered high-responders consisted of women <38 years old under the additional presence of at least one of the following criteria: E2 levels ≥3,000 pg/mL, last ultrasound with ≥20 follicles (each ≥11 mm at oocyte maturation trigger day), or ≥13 cumulus–oocyte complexes (COC) ( Gat et al. 2015 , Sousa et al. 2015 , Drakopoulos et al. 2023 , Feferkorn et al. 2023 ). Selected patients were divided according to the oocyte maturation trigger: 226 cycles were triggered with hCG and 457 cycles were triggered with a GnRHa. Of the cases with an hCG-trigger, there were 127 frozen-thawed embryo transfer (FET) cycles, and of the cases with a GnRHa-trigger, there were 407 FET cycles. The hCG-trigger option was selected when, besides age and ovulatory dysfunction, clinicians did find a non-suspicious number of follicles and E2 levels, and in these cases a fresh embryo transfer was issued; however, in a few cases, although initially sensing as secure the use of an hCG-trigger, development of clinical and ultrasound risk criteria led to a freeze-all approach. The GnRHa/agonist-trigger option was selected when, besides age and ovulatory dysfunction, clinicians did find a high number of follicles and E2 levels; in most cases patients received luteal additional hCG supplementation followed by fresh embryo transfer; however, some cases that had been primed with additional luteal hCG supplementation then changed to FA under new ultrasound evidences; and in other cycles, where the risk was perceived to be extremely high before oocyte pick-up (age, PCOS, elevated AMH (>3.4 ng/mL), or E2 ≥ 3,000 pg/mL), clinicians decided immediately for an FA approach. Data on demographic, infertility, and stimulation characteristics, and on embryological, clinical, and newborn outcomes were evaluated in detail, with cumulative clinical outcomes being additionally provided. The methodology for karyotyping ( Rooney & Czepulkowski 1997 ), Y-chromosome screening ( Gonçalves et al. 2017 , Krausz et al. 2024 ), and preimplantation genetic testing ( Lopes et al. 2018 , Giuliano et al. 2023 ) followed described protocols. Semen analysis ( WHO 2021 ), semen preparation and freezing, and testicular sperm retrieval and freezing ( Madureira et al. 2014 ) were performed as previously described. In this study, no exclusion criteria were applied to male causes, and thus fresh and frozen ejaculated sperm or testicular sperm were used. Regarding sperm quality, the majority of cases presented altered semen parameters. Embryological and clinical definitions followed the International Glossary ( Zegers-Hochschild et al. 2017 ). In IVF cases, as cumulus cells are not removed, the maturation rate (MII/COC) could only be evaluated at the time of pronuclei observation. In these cases, the total number of MII corresponded to normal and abnormal fertilized oocytes. In this study, the primary outcome was live birth (LB) and the secondary outcomes were the rates of OHSS, CP, newborn (NB), and respective cumulative clinical outcomes. Patients underwent COS with a GnRH-antagonist protocol (0.25 mg, cetrorelix, Cetrotide, Merck Serono, Switzerland; or ganirelix, Orgalutran, Organon, Netherlands). For stimulation, recombinant follicle-stimulating hormone (rFSH) was used (Puregon, Organon; or Gonal-F, Merck Serono, Darmstadt, Germany). Ovulation trigger was performed 36 h before follicle aspiration, either with hCG (5.000–10.000 IU, Pregnyl, Organon), recombinant hCG (rhCG) (250 μg/0.5 mL, Ovitrelle, Merck Serono), or with a GnRHa (0.2 mg, triptorelin, Ipsen Pharma Biotech, France). Estradiol, progesterone (P4), and LH serum levels were evaluated on the day or one day before oocyte maturation triggering ( Huirne et al. 2007 , Pinto et al. 2009 ). More specifically, although adapted to each individual, generally the following protocol was applied: on the second day of menses, rFSH is initiated. The starting dose is calculated according to the following parameters: if the patient presents an antral follicle count inferior to ten, the starting rFSH dose is 150–225 U; if the patient’s antral follicle count is equal to or greater than 10, the starting dose is 100–150 U. Nevertheless, if the patient’s BMI is equal to or greater than 30 kg/m 2 , the starting dose is 225 U, independent of the antral follicle count. After 4 days of rFSH medication, an ultrasound is performed: if the patient has at least one follicle with 13 mm or more, GnRH-antagonist is initiated. After another 4–6 days, another ultrasound is performed: if the patient has at least three follicles with 17 mm or more, the oocyte maturation trigger is given to the patient; if this criterion is not met, ultrasounds are performed daily until follicle size is achieved. Follicles with ≥11 mm were aspirated under ultrasound-guided ovarian puncture. Gamete and embryo handling were performed with media from Origio (Denmark) and Vitrolife (Sweden) in a heated airflow (K-system, Kivex Biotec, Denmark; Cooper Surgical, Denmark). Cumulus-oocyte complexes were collected in SynVitro Flush Medium (without heparin, Origio) in 90 mm Falcon embryo culture plates (Corning, USA) and left in sequential Fert Medium (Origio) for 3 h in 1-well Falcon tissue culture plates (Corning) in ESCO incubators (MIRI Multiroom Incubator, Esco Medical, Singapore), until insemination or denudation. In in vitro fertilization (IVF) cycles, insemination was performed in sequential Fert Medium, and observation of pronuclei was carried out after 16–18 h. Zygotes were placed in sequential Cleavage Medium (Origio). In intracytoplasmic sperm injection (ICSI) cycles, denudation was performed in 60 mm Falcon embryo culture plates containing sperm preparation medium (SPM) under oil (Ovoil, Vitrolife), first enzymatically in SPM with Cumulase (Origio), and then mechanically in SPM with Handling micropipettes (Vitrolife). After denudation, oocytes were placed in sequential Cleavage Medium inside an ESCO incubator. Microinjection was performed in an inverted microscope with a thermal stage (37°C) and Narishige micromanipulators, using Origio micromicropipettes (Cooper Surgical). ICSI was performed using the strong dislocation of the cytoplasm ( Tesarik & Sousa 1995 ). Embryos were cultured in sequential Cleavage Medium up to the morning of day 3, and then in sequential Blast Medium (Origio) up to embryo transfer or freezing, in Embryoslide culture dishes (Vitrolife) inside an Embryoscope apparatus (Vitrolife). The grade of cleavage embryos and blastocysts followed Istanbul consensus ( Balaban et al. 2011 ). Ultrasound-guided embryo transfer was performed with a Sure View Wallace Embryo Replacement Catheter or Wallace malleable stylet (Smiths Medical Int, UK). For embryo freezing, blastocysts were kept in Embryoslide culture dishes containing Sequential Blast Medium. Expanded blastocysts (BL4, BL5) were submitted to laser blastocoel collapse before freezing (in the same dish and medium). Collapse was performed with a Saturn active laser system (Laser Saturn 5 Active, Research Instruments, UK) operated in the inverted microscope. Blastocysts were vitrified and warmed according to the Kitazato protocol (Kitazato Corp, Japan). Retracted or slightly expanded blastocysts were submitted to assisted hatching using a laser beam to drill about 1/4 of the zona pellucida. Blastocysts were then incubated for 2 h in the same 5-well embryo culture dish containing Sequential Blast Medium, under oil, in an ESCO incubator until transfer. Luteal supplementation in fresh embryo transfers of hCG-triggered cycles was accomplished with intravaginal administration of progesterone (Cyclogest, LD Collins, UK; or Progeffik, EFFIK, Portugal), 800 mg daily since the day of oocyte retrieval. Luteal support in fresh embryo transfers of GnRHa-triggered cycles was additionally associated with oral estradiol (Estrofen 2 mg, Isdin, Novo Nordisk, Denmark), 2 mg 12/12 h for the same period as progesterone, and a single injection of hCG (1,500 IU) or of rhCG (60 μg/0.5 mL) on the day of oocyte pick-up ( Radesic & Tremellen 2011 ). The endometrium was prepared with estradiol (Estrofen, Isdin), 6 mg (2 mg morning, 4 mg night) since day 2 of menses. Acetylsalicylic acid, 150 mg (Tromalyt, Rottapharm, Ireland), one per day, was added concomitantly with estradiol. About 10–18 days later, if the endometrium was ≥7 mm and there were no follicles ≥12 mm, intravaginal administration of progesterone (Cyclogest, LD Collins; or Progeffik, EFFIK), 800 mg daily, was added. Embryo transfer was performed on day 5 after the beginning of progesterone (day 0). Under ultrasound surveillance, when a preovulatory follicle (≥16 mm) was observed and the endometrium thickness was ≥7 mm, E2, LH, and P4 levels were measured, and a single injection of rhCG (250 μg/0.5 mL) was administered. Two days after ovulation trigger, micronized progesterone (Cyclogest or Progeffik), 400 mg daily, was initiated. Embryo transfer was performed on day 7 after ovulation trigger. Implantation was confirmed by serum βhCG values ≥ 20 mIU/mL, measured 12 days after embryo transfer. CP was established by ultrasound at 6 weeks of gestation (embryo sac). Progesterone/estradiol were continued until 8–10 weeks of gestation. Quantitative variables were described as mean ± standard deviation, and were analyzed using the independent samples t -test when only two groups were compared, or one-way ANOVA (analysis of variance) when the five groups were compared. Categorical variables were described as percentages and compared using the Chi-square test. Whenever the necessary conditions were not met, not applicable (n.a.) was reported in the tables instead of the P -value. Pairwise comparisons were performed when significant differences were found in the global analysis of the five groups. Pairwise comparisons of means were performed using the Tukey test, and pairwise comparisons of proportions were performed using the Bonferroni correction. Although not common, it is possible that the global test was significant but the pairwise comparison test could not identify where the differences lay. Associations of the different outcomes with protocol (Ag-fresh-hCG versus hCG-fresh) were investigated using logistic regression. The outcomes considered for each cycle were CP and live birth delivery. As a first step, the univariable association between the trigger and each outcome was investigated. As a second step, the association between each outcome and the protocol was obtained by controlling for female age, anti-Müllerian hormone (AMH), PCOS, estradiol, and number of follicles in a multivariable model. All these variables were considered in the multivariable model, and non-significant variables ( P > 0.05) were dropped from the final model. Unadjusted odds ratio and adjusted odds ratios with 95% confidence intervals were estimated. The subsets fresh embryo transfer cycles (ETC) and frozen-thawed embryo transfer (FET) were considered. The same methodology was considered for the global analysis of all treatments, Ag-fresh-hCG and hCG-fresh, irrespective of the number of cycles performed, considering the outcomes of having at least one CP and having at least one LBD. Having at least one CP or LBD represents the likelihood of achieving a CP or LBD, respectively, when starting a first cycle, considering that freeze-all can occur and that frozen ETCs may be performed. All statistical tests were two-tailed, and the significance level considered was P < 0.05. Statistical analysis was carried out through the IBM SPSS Statistics 29 program.

The present manuscript describes a common situation observed in many ART centers, where a large span of time was needed to accept the GnRHa as trigger, where it is still a matter of debate the use of a freeze-all policy, and where different clinicians with distinct convictions work together. In the absence of a definite consensus for high-responders in the studied period, clinicians relied on proposals from most credited authors on the subject, basing the ultimate decision on each patient’s characteristics and response to ovarian stimulation. In the present retrospective analysis, we could distinguish five different approaches that depended on the assessed perception risk for the development of severe OHSS. Nearly one third of the cases underwent hCG-trigger as, besides age and ovulatory dysfunction, clinicians did find a non-suspicious number of follicles and E2 levels, and thus a fresh embryo transfer was issued; however, in a few cases, although initially sensing secure the use of an hCG-trigger, development of clinical and ultrasound risk criteria led to the use of an FA approach. The same reasoning applied to cases where a GnRHa/agonist-trigger was used, this time also considering a high number of follicles, with most cases then receiving luteal additional hCG supplementation followed by fresh embryo transfer ( Humaidan & Haahr 2023 ); however, some cases that had been primed with luteal additional hCG supplementation then changed to FA under new ultrasound evidence; and in other cycles, where the risk was perceived to be extremely high before oocyte pick-up (age, PCOS, elevated AMH (>3.4 ng/mL), E2 ≥ 3,000 pg/mL), clinicians decided immediately for an FA approach. With this reasoning, severe OHSS cases only developed in fresh ETCs. As of the three cases with severe OHSS in hCG-trigger cycles only age and PCOS were simultaneously present, and as of the three cases with severe OHSS in GnRHa-trigger cycles only age, PCOS, and higher number of follicles were simultaneously present, this suggests that under the presence of at least two clinical criteria (such as age and PCOS) it could be best to apply preventive measures. In the literature, the use of a GnRHa-trigger instead of an hCG-trigger to prevent OHSS has not shown consensual results. In a study with high-responders, authors offered evidence that GnRHa-triggering with luteal hCG support followed by fresh embryo transfer was associated with absence of OHSS ( Fusi et al. 2015 ), whereas in other studies severe OHSS could not be abolished ( Engmann et al. 2008 , Radesic & Tremellen 2011 , Iliodromiti et al. 2013 , Christopoulos et al. 2016 ). In the present report, GnRHa-triggered cycles with luteal hCG support followed by fresh embryo transfer, although not abolishing OHSS, enabled obtaining lower severe OHSS rates (1.4%-hCG/0.9%-GnRHa). However, when a GnRHa-trigger with luteal hCG support or an hCG-trigger were followed by freeze-all, there were no cases with severe OHSS. This is in line with previous observations that indicated total avoidance of severe OHSS with a freeze-all policy ( Devroey et al. 2011 , Humaidan et al. 2011 , 2012 , 2013 a , b , Humaidan & Haahr 2023 ). The use of a GnRHa-trigger has also been associated with similar clinical outcomes to those obtained with hCG triggering ( Engmann et al. 2008 , Humaidan et al. 2011 , Abdulkhalikova et al. 2022 ), with the same being observed in high-responders using GnRHa-triggering with luteal hCG ( Fusi et al. 2015 ). Similarly, embryological outcomes were found to be similar with both triggers, suggesting that the GnRHa-trigger has no detrimental effect on embryo quality ( Humaidan et al. 2010 , Martazanova et al. 2022 ). Other authors referred to higher clinical outcome rates in GnRHa-trigger cycles, which were attributed to the simultaneous surge of FSH and LH, which enabled better oocyte and embryo quality ( Humaidan et al. 2010 , Kol & Humaidan 2010 ), and to the influence of the hCG bolus and E2 in the luteal phase support, which counteracted the negative impact of premature luteolysis ( Humaidan et al. 2010 , Radesic & Tremellen 2011 ). In the present work, higher rates of COC, MII, and blastocysts were observed only in FA cases, and higher implantation and CP rates in agonist-trigger cycles with fresh embryo transfer. The impact of FET is yet unclear, as some studies revealed lower CP rates ( Pinborg 2012 ) and higher abortion rates ( Tomás et al. 2012 ), whereas others found no detrimental impact ( Shapiro et al. 2013 ). In the present population, although we also did not find differences between groups after FET cycles, the cumulative CP and NB rates were higher with Ag-fresh-hCG. Nevertheless, there were no significant differences after multivariable logistic regression analysis. The strengths of this study include presenting a large patient series with well-defined criteria for high-responders, different trigger comparisons, inclusion of fresh and frozen cycles, cumulative rates, and detailed embryological, clinical, and newborn outcomes. The main limitation is the retrospective nature of the study, allied to the heterogeneity observed among the studied group. Finally, it is important to frame the present results in their particular context, as, since 2012, several technological advances have been implemented in ART. Therefore, the conditions to which patients were submitted could have been partially different, considering when the cycle was performed. The current tendency in high-responder patients is to engage in GnRHa-triggering, with luteal hCG supplementation, followed by a freeze-all approach ( Humaidan & Haahr 2023 ). The present report describes the outcomes in high-responder patients when hCG was still used as trigger, and when a GnRHa-trigger was introduced in association with hCG luteal support and fresh embryo transfer or with freeze-all. Four main conclusions and advice come from the observed results: i) the use of an agonist-trigger provides similar embryological and clinical outcomes compared with the classic hCG-trigger, with the advantage of obtaining a decrease in severe OHSS rates in cases using a GnRHa-trigger in association with hCG luteal support and fresh embryo transfer; ii) the option to freeze all embryos elicited total avoidance of severe OHSS, without compromising clinical outcomes after frozen embryo transfer; iii) the cumulative CP and NB rates are superior in agonist-trigger cycles; iv) the clinical presence of young ovarian age and PCOS should be a signal to undertake a freeze-all approach. Results thus indicate that, in high-responders, the best treatment option is the agonist-trigger protocol followed by a freeze-all approach.

Ovarian hyperstimulation syndrome (OHSS), the most frequent assisted reproductive treatment (ART) complication ( Aboulghar & Mansour 2003 ), occurs as an excessive response to controlled ovarian stimulation (COS) ( Grossman et al. 2010 ) and can lead to cycle cancellation, prolonged bed rest, hospitalization, and death ( Delvigne & Rozenberg 2003 ). The main identified risk factors for OHSS are young age, low body mass index (BMI), polycystic ovary syndrome (PCOS), higher doses of exogenous gonadotropins administered, high absolute or rapidly rising serum estradiol (E2) levels, high number of developing follicles or retrieved oocytes, and previous OHSS episodes ( Practice Committee of the American Society for Reproductive Medicine 2024 ). Although patients who exhibit a high response to COS present high pregnancy rates ( Kligman & Rosenwaks 2001 ), they also have a substantially higher risk for OHSS ( Smitz et al. 1990 ). Several strategies have been implemented to lower/eliminate OHSS, such as using a gonadotropin-releasing hormone agonist (GnRHa) instead of human choriogonadotropin (hCG) as an oocyte maturation trigger, as well as the freeze-all approach ( Devroey et al. 2011 ). The GnRHa, when used as a trigger in GnRH antagonist stimulation cycles, displaces the GnRH antagonist from the hypothalamic receptor and elicits an increase in the pituitary liberation of luteinizing hormone (LH) and follicle stimulating hormone (FSH) (the flare-up effect). This dual surge is more physiological than the LH-like only effect of the hCG-trigger, as the additional FSH surge further promotes LH receptors in granulosa cells and oocyte nuclear maturation. However, after this surge, due to its short life, there is an abrupt decrease in LH levels, which results in corpus luteum involution, with a subsequent decrease in steroid hormone production that then causes endometrium insufficiency. This obliges rescuing the endometrium with a low dose of hCG (1,500 IU) at oocyte pick-up in addition to estradiol and progesterone until the trophoblast begins secretion of hCG. Under low LH levels, granulosa cells of the corpus luteum do not produce enough estradiol and progesterone, and that is why the endometrium must be primed with these steroids. On the other hand, decreased LH stimulation decreases the synthesis of the vascular endothelial growth factor, thus decreasing the occurrence of OHSS ( Radesic & Tremellen 2011 , Humaidan et al. 2012 , 2016 , Humaidan & Haahr 2023 ). The use of a GnRHa to trigger the endogenous LH surge ( Gonen et al. 1990 ) was then shown to reduce the OHSS risk by inducing a quick and reversible luteolysis ( Castillo et al. 2020 ). However, the use of a GnRHa-trigger also results in luteal phase deficiency, with lower E2 and progesterone levels and shorter luteal phase duration ( Humaidan et al. 2005 , Tannus et al. 2017 ), being observed to elicit lower implantation and clinical pregnancy (CP) rates ( Humaidan et al. 2005 , Kolibianakis et al. 2005 ). Few studies have compared hCG and GnRHa triggers in patients classified as high-responders. To provide new solid information, this study evaluated a large cohort of high-responders, comparing two oocyte maturation triggers, GnRHa with hCG at oocyte retrieval day, to hCG, presenting detailed OHSS, embryological, clinical, newborn, and cumulative clinical outcomes.

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the work reported.

The data underlying this article are available from the corresponding author on reasonable request.

MG contributed to data acquisition, statistical analyses, and draft writing. MC contributed to data acquisition and conceptualization. JTS contributed to conceptualization and female evaluation. JS and PV contributed to data acquisition. CO contributed to female evaluation. MFC did statistical assistance. AB contributed to patient recruitment. MS contributed to supervision and final manuscript writing. All authors approved the final manuscript.