Does hCG-trigger in the mild stimulation protocol for endometrial preparation have any effect on pregnancy outcome in frozen-thawed embryo transfer?

This retrospective cohort study was designed to investigate the pregnancy outcomes among eligible patients who underwent FET cycles with endometrial preparation using a mild ovarian stimulation protocol between 2020 and 2022 at the Royan Institute (Tehran, Iran). This study approved by the Scientific Board and the Ethics Committee of Royan Institute. All women aged ≤ 40 years who had previously undergone conventional in vitro fertilization (IVF) and/or intracytoplasmic sperm injection (ICSI) with cryopreservation of all or part of numerous top-quality embryos were assessed. The following exclusion criteria were applied: (I) endometrial thickness less than 6 millimeters, (II) patients with moderate and severe endometriosis, (III) patients diagnosed with untreated hydrosalpinx, (IV) treatment cycles using preimplantation genetic screening (PGS) and (V) an abnormal uterine cavity (congenital uterine anomalies, submucosal fibroids, and intrauterine adhesions). The selection of ovarian stimulation protocols, which culminated in oocyte retrieval, was predicated on an array of patient-specific characteristics and baseline ovarian reserve laboratory values. These protocols were implemented in accordance with methodologies delineated in prior literature [ 25 , 26 ]. Embryos of superior quality that were surplus to the immediate requirements were cryopreserved at the cleavage stage (2nd or 3rd day), utilizing a vitrification technique as explicated in the previous study [ 27 ]. On the day of the frozen-thawed ET, the patient’s Cryotop (Kitazato Biopharma, Tokyo, Japan) was retrieved from the liquid nitrogen and the embryos were subjected to a thawing solution of 1 M sucrose in Ham’s F-10 medium (Caisson Laboratories, East Smithfield, USA) with 20% Albuminal-5. Subsequently, they were transferred to dilution solutions of 0.5 M sucrose for 3 min and 0.25 M sucrose for another 3 min. The warmed embryos underwent multiple washes in a washing solution of Ham’s F-10 medium with 20% Albuminal-5 before being transferred to G-1 medium (version 3; Vitrolife, Kungsbacka, Sweden) and then were cultured for 20 min up to 2 h in EmbryoGlue medium (LifeGlobal, Toronto, Canada). ET was thereafter performed using a Labotect catheter (Labotect, Göttingen, Germany) via a standard technique [ 27 , 28 ]. All patients underwent mild stimulation for endometrial preparation according to the following protocol. On the 2nd or 3rd day of menstrual cycle or progesterone withdrawal-induced cycle, patients were administered 5 mg of letrozole (Letrofem ® 2.5 mg, Iran Hormones Company) for 5 days. On either 6th or 7th day of the menstrual cycle, a dose of gonadotropin ampoule (recombinant-follicle stimulating hormone: Cinnal-f, Cinagen, Iran) was administered to induce the growth of 1 to 3 dominant follicles. Ultrasound monitoring was performed every other day after the detection of a dominant follicle ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\ge\:13\:mm)\:$$\end{document} on 8th -10th day of menstrual cycle for all patients. The serum levels of hormones, including LH, progesterone, and estradiol, were measured at each visit in both groups. When the first dominant follicle size reached 18 to 20 mm, and also with a ≥ 150 pg/mL estradiol level, and an endometrial thickness of ≥ 7 mm, ovulation trigger by a recombinant hCG ampoule (Ovitrelle; Merck-Serono, Geneva, Switzerland). The spontaneous LH surge was defined as [ 1 ] LH levels over 15 IU/L or a 2-fold increase above the baseline, and [ 2 ] an apparent elevation of progesterone, despite the fact that LH elevation might not be captured [ 8 ]. In the event of spontaneous ovulation, the date of embryo transfer (ET) was established according to the identification of day zero. If there was a 30% decrease in estradiol level and any decrease in LH level in comparison with the previous measurement and progesterone level was above 1 ng/ml, that day was considered as day zero [ 9 ]. ET was scheduled on (embryonic age + 1) day after the spontaneous LH surge or on (embryonic age + 2) day after rhCG injection. For cleavage stage embryos (3rd day), the ET was planned on 4th day after the spontaneous LH surge or 5 days after the hCG injection, and for blastocyst stage embryos (5th day), the ET was scheduled on 6th day after the spontaneous LH surge or 7th day after the hCG injection. In the sequential ET, a combination of cleavage the embryo and blastocyst were transferred. Luteal phase support treatment was started on ET day with a daily vaginal progesterone suppository (Cyclogest, Actavis, Barnstaple, UK). The study population were segregated into two distinct groups according to type of ovulation: the spontaneous ovulation group ( n  = 138) and the hCG-trigger group ( n  = 271). The study outcomes were to compare live birth rate (LBR), clinical pregnancy rate (CPR) and implantation between hCG-tigger and spontaneous LH surge groups in the mST-FET cycles. The implantation rate is calculated by dividing the total number of gestational sacs observed in the uterus via ultrasound by the total number of transferred embryos. A clinical pregnancy is identified by the observation of at least one gestational sac containing an embryo and a detectable heartbeat via vaginal sonography at 6–7 weeks after ET. An ectopic pregnancy is diagnosed with a gestational sac outside of the uterus in sonography scanning. An early miscarriage refers to the loss of a clinical pregnancy before 12 weeks of gestation. A live birth is defined as delivery of a viable newborn regardless of the length of gestation [ 29 ]. Statistical analyses were conducted using IBM SPSS Statistics software (version 24.0). Continuous variables were expressed as means ± standard deviations, while categorical variables were expressed as numbers (percentages). The normality of the variables was assessed using the Kolmogorov-Smirnov test and all quantitative variables in this study had a normal distribution. Independent samples t-tests were employed for the comparison of the quantitative variables between the two groups. For categorical variables, chi-square tests were performed. The multivariable logistic regression was performed to detect the most significant variables related to the LBR in the study population. A p-value of less than 0.05 was considered statistically significant.

The present study included 409 patients who had undergone FET cycles with mild stimulation for endometrial preparation. There were no significant differences between the two groups with regard to demographic parameters (Table  1 ). However, duration of mild stimulation and endometrial thickness at day zero or trigger day were significantly higher in the spontaneous ovulation group in comparison with the hCG trigger group ( P  < 0.001). Table 1 The comparison of demographic data and baseline characteristics in mild stimulation + spontaneous ovulation group vs. mild stimulation + hCG trigger group in FET cycles Variables Mild stimulation + Spontaneous Ovulation ( n  = 138) Mild stimulation + hCG trigger ( n  = 271) P -value Age at oocyte retrieval (years) 31.48 ± 5.60 31.79 ± 5.78 0.60 Age at FET (years) 32.84 ± 5.79 33.27 ± 5.77 0.47 Body mass index (kg/m 2 ) 26.04 ± 4.13 26.46 ± 4.41 0.34 No. of Previous pregnancies 0.67 0 77 (55.8%) 160 (59%) 1 30 (21.7%) 60 (22.2%) ≥ 2 31 (22.5%) 51 (18.8%) Type of infertility 0.12 Primary 123 (89.1%) 226 (83.4%) Secondary 15 (10.9%) 45 (16.6%) No. of previous failed ET 1.34 ± 1.36 1.37 ± 1.37 0.82 Cause of infertility 0.091 Unexplained factor 28 (20.3%) 40 (14.8%) Tubal factor 69(50%) 116 (42.8%) PCOS 15 (10.9%) 54 (19.9%) Ovulatory age factor 5(3.6%) 9 (3.3%) Mixed factor 21 (15.2%) 52 (19.2%) Baseline hormone levels LH (IU/L) 6.20 ± 5.68 5.86 ± 3.36 0.45 FSH (IU/L) 6.67 ± 5.53 5.92 ± 2.20 0.12 AMH (ng/ml) 4.57 ± 3.40 5.14 ± 4.40 0.16 COH protocol 0.07 GnRH Antagonist 107 (77.5%) 202 (74.6%) Long GnRH agonist 21 (15.2%) 60 (22.1%) Mild stimulation 10 (7.3%) 9 (3.3%) No. of retrieved oocytes 15.52 ± 7.18 14.87 ± 7.13 0.38 No. of obtained embryos 8.40 ± 3.96 8.41 ± 4.68 0.98 Duration of embryo freezing (month) 15.72 ± 26.78 18.01 ± 22.68 0.36 Duration of mild stimulation 16.73 ± 2.18 14.86 ± 3.88 < 0.001 Total no. of used gonadotropin ampoules 3.56 ± 1.87 3.50 ± 2.45 0.80 Endometrial thickness on trigger day (mm) 10.10 ± 1.95 9.34 ± 1.77 < 0.001 Types of ET 0.62 Easy 133 (96.4%) 257 (94.8%) Difficult 5 (3.5%) 14 (5.2%) Stage of ET 0.058 Cleavage 28 (20.3%) 86 (31.7%) Blastocyst 91 (65.9%) 155 (57.2%) Sequential 19 (13.8%) 30 (11.1%) No. of transferred embryos 2.16 ± 0.65 2.20 ± 0.58 0.52 COS: controlled ovarian stimulation; FET frozen embryo transfer; PCOS: polycystic ovary syndrome; Values are presented as the mean ± Standard deviation and number (percent). * It was obtained by independent t or Chi-square tests and statistically significant level was 0.05 The comparison of demographic data and baseline characteristics in mild stimulation + spontaneous ovulation group vs. mild stimulation + hCG trigger group in FET cycles COS: controlled ovarian stimulation; FET frozen embryo transfer; PCOS: polycystic ovary syndrome; Values are presented as the mean ± Standard deviation and number (percent). * It was obtained by independent t or Chi-square tests and statistically significant level was 0.05 The comparison of the pregnancy outcomes between the two groups indicated no statistically significant differences in terms of implantation, chemical and clinical pregnancy rates as well as miscarriage rate and live birth rates. However, the incidence of ectopic pregnancy was significantly higher in the spontaneous ovulation group in comparison with the hCG trigger group ( P  = 0.03) (Table  2 ). Table 2 The comparison of pregnancy outcomes in mild stimulation + spontaneous ovulation group vs. mild stimulation + hCG trigger group in FET cycles Variables Mild stimulation + Spontaneous Ovulation ( n  = 138) Mild stimulation + hCG trigger ( n  = 271) P -value Implantation rate 0.65 ± 0.32 0.60 ± 0.30 0.31 Chemical pregnancy rate 54 (39.1%) 118 (43.5%) 0.39 Clinical pregnancy rate 51 (37%) 111 (39.7%) 0.53 Ectopic pregnancy rate 3 (5.6%) 0 (0%) 0.03 Early miscarriage rate 1 (1.9%) 10 (8.7%) 0.17 Live birth rate 49 (35.5%) 101(37.3%) 0.74 FET: frozen embryo transfer; hCG: human chorionic gonadotropin The comparison of pregnancy outcomes in mild stimulation + spontaneous ovulation group vs. mild stimulation + hCG trigger group in FET cycles FET: frozen embryo transfer; hCG: human chorionic gonadotropin For a more thorough investigation, a multivariable logistic regression analysis was applied to explore the factors influencing LBR within the studied population ( n  = 409). Potential influential variables such as women’s age at ovum-pick up day, body mass index (BMI), cause of infertility, number of retrieved oocytes, the study groups (hCG trigger versus spontaneous ovulation), endometrial thickness on the trigger day, and number and the stage of the transferred embryo were included in the regression model. The findings revealed that only the stage of the transferred embryo exhibited a significant relationship with LBR (blastocyst vs. cleavage: odds ratio (OR); 2.33, 95% confidence interval (CI):1.41–3.86, P-value = 0.001).

In the present study, we found that the pregnancy outcomes in the mST-FET cycles with and without hCG triggering were similar. To the best of our knowledge, there is no similar study to compare present finding in mST-FET cycles; however, the effect of hCG triggering in the natural-FET cycles have been reported previously [ 30 ]. The administration of hCG to induce ovulation during the planning of a FET, showed a viable option for terminating the follicular phase and scheduling embryo transfer independently of the natural occurrence of the endogenous LH surge. This approach offers several advantages, including decrease in rate of visits, ultrasound scans and hormonal blood tests, making it more economical and convenient for both the patients and the clinician [ 8 , 31 – 35 ]. However, emerging evidence suggests that the administration of hCG during the late follicular phase initiates a series of events in the endometrium that would naturally commence several days later with the presence of an endogenous LH surge. These events appear to have a detrimental impact on the ongoing pregnancy rate [ 24 , 36 ]. To address the pros and cons of using hCG as a triggering option in ovulation, several investigations have been conducted. The findings of Gao and colleagues suggested that using an hCG trigger in modified NC (mNC)-FET could potentially result in a lower occurrence of LBR when compared to NC-FET among women with regular menstrual cycles [ 13 ]. However, Mackens et al. conducted an randomized clinical trial and found no significant difference between NC-FET and mNC-FET in clinical outcomes, including CPR and ongoing pregnancy rate (OPR) [ 32 ]. Other studies have reported similar results in both groups in terms of embryo implantation rate, miscarriage rate, LBR, and multiple CPR [ 35 , 37 , 38 ]. On the other hand, Farid Mojtahedi et al. reported a significant higher rate of implantation in mNC-FET (using hCG as ovulation trigger) in comparison with true natural-FET [ 35 ]. Levi Setti et al., indicated a remarkable difference in pregnancy outcomes in terms of CPR and LBR in mNC-FET over NC-FET or AC-FET [ 39 ].The current research did not focus on the natural cycle. However, our findings indicate that in mST-FET cycles, similar to certain studies conducted on natural cycles, there is no significant difference between the use of HCG for triggering ovulation and spontaneous ovulation. In fact, greater endometrial thickness was achieved in the spontaneous ovulation group in comparison with the hCG trigger group, nevertheless this improvement did not affect pregnancy outcomes which is in agreement with the recent observations by Farid Mojtahedi et al. and Lee et al. [ 35 , 38 ]. In a study bearing resemblance to our current investigation, Bilgory et al. conducted an evaluation of all NC-FET and mST-FET cycles involving the use of letrozole. This evaluation was performed within a cohort of 346 patients and was retrospective in nature [ 8 ]. They reported that the chemical pregnancy, live birth and miscarriage rates were all comparable between groups and concluded that hCG triggering even after LH elevation would not affect the pregnancy outcomes. Whereas, our study compromised of just mST-FET cycles. In accordance with Bilgory et al.’s findings, chemical pregnancy, live birth and miscarriage rate were all the same between our groups. The only difference between our study and theirs is the significant higher rate of ectopic pregnancy (EP) in our spontaneous ovulation group. In line with our results, Levi Setti and co-workers presented a higher EP rate in NC-FET versus mNC-FET cycles [ 39 ]. Conflicting information exists on whether luteal phase support in fresh cycle or endometrial preparation in FET cycles can affect the EP development [ 40 ]. A hypothesis for the occurrence of EP has been attributed to the induction of anomalous uterine contractions. These contractions could potentially instigate a retrograde migration of embryos from the uterine cavity into the Fallopian tube, culminating in ectopic implantation [ 41 ]. Sahin and colleagues highlighted that the decrease in EP rate following hCG triggering could be attributed to improvement in the luteal phase support and implantation [ 42 ]. There have been only limited studies in this area. Further studies are needed to draw conclusions. Clayton et al. have reported that tubal factor is the main risk factor of ectopic pregnancy in IVF-ET procedures [ 43 ]. Three cases of ectopic pregnancy, each associated with infertility attributed to tubal factors, were observed exclusively within the spontaneous ovulation group. We concluded that tubal factor that may only be observed with hydrosalpingography (HSG) is the main culprit for EP. Hence, neither HCG trigger nor its absence are held responsible for this event. The main limitation in our study is its retrospective nature. The strength of our study is an appropriate scale sampling in the mST-FET cycles. Insufficient research has been conducted on the impact of hCG triggering in mST-FET cycles. It is important to highlight that the present study applied a multivariable regression model to analyze the variables influencing the LBR in the mST-FET cycles. As it is determined, in our study population, the stage of ET was the only major factor which significantly related to the LBR; so that blastocyst-stage ET was associated with a 2.3-fold increase in the probability of live birth in compared to cleavage-stage. Recent research has focused on predicting LBRs after FET cycles. Liang et al. found that major factors including maternal age, BMI, basal LH and FSH levels, and insulin resistance index were key predictors of LBR in a regression model [ 44 ]. Meanwhile, Pan and colleagues highlighted that the duration of infertility, endometrial thickness, and number of transferred embryos were crucial variables for LBRs in young women after cleavage-FET [ 45 ]. Additionally, Fang and co-workers reported that advanced paternal age and high maternal basal LH hormone levels adversely affected LBR and they also noted that LBRs significantly improved with artificial cycle preparation of the endometrium and 5-day blastocyst embryo transfer [ 46 ]. However, further research is needed to explore the various factors influencing LBRs after FET cycles, particularly in different endometrial preparation protocols. According to our study, pregnancy outcomes such as CPR, and LBR are comparable in the mST-FET cycles with or without the hCG-trigger. Based on the findings from multivariate regression analysis, the sole significant predictive factor for the live birth rate was the transfer of blastocyst embryos. However, it is recommended that these results be examined and discussed in future prospective studies with a larger sample size, considering the lack of comparable research in this field.

In the recent years, there has been a remarkable rise in the employment of frozen-thawed embryo transfer (FET) and this can be attributed to the expansion of its indications, facilitated in part by advancements in vitrification techniques [ 1 ]. Mounting evidence of successful rate of FET outcomes, shifted clinical practice towards “freeze-all” strategy [ 2 ]. Singletons born following FET brought hopeful outcomes, including a decrease rate of low-birth-weight newborns, preterm delivery in comparison with singletons born after fresh ET. Research indicates that FET cycles are frequently associated with a reduced incidence of placenta previa, placental abruption, perinatal mortality, and ovarian hyperstimulation syndrome in comparison with fresh ET [ 1 – 3 ]. One downside is that it has an increased risk of large for gestational age newborns (LGA) and macrosomia (> 4500 g) when compared to both fresh cycles and spontaneous conception [ 4 ]. To maximize the likelihood of a successful pregnancy, it is essential to synchronize the development of the embryo and endometrium [ 5 ]. Various protocols are commonly employed to prepare the endometrium for FET. These may include natural cycles (NC), either pure or modified, stimulated cycles involving ovulation induction, and programmed or artificial cycles (ACs) requiring hormonal replacement therapy (HRT). The former two options involve the presence of a corpus luteum, while the latter involves its absence [ 6 , 7 ]. Some studies have been expressed that the absence of a corpus luteum would lead to hypertensive disorders of pregnancy [ 1 , 8 , 9 ]. The AC-FET procedure comprises a sequential regimen of estrogen and progesterone to mimic a normal endometrial cycle. Estradiol is initially introduced to induce proliferative endometrium and to suppress the growth of a dominant follicle. Estrogen is then continued until the endometrial thickness reaches 7–9 mm on vaginal ultrasonography, at which point progesterone is initiated to facilitate the secretory phase [ 7 ]. NC-FET protocol can be implemented through either spontaneous luteinizing hormone (LH) surge detection or administration of human chorionic gonadotropin (hCG) to determine an ET timing. This approach represents the simplest endometrial preparation method. A mST-FET cycle involves inducing ovulation through the administration of drugs such as clomiphene citrate or letrozole, with or without gonadotropins. Therefore, the endometrium is prepared through endogenous estrogen and progesterone [ 7 ]. Letrozole is an aromatase inhibitor that can elevate serum gonadotropin-releasing hormone levels to induce follicular development [ 8 – 11 ]. A stimulated cycle with letrozole could enhance ανβ3 integrin expression which is an indicator of endometrial receptivity and is crucial for the success of IVF treatment, compared to natural cycles [ 12 ]. Recently, potential benefits of using letrozole in comparison to natural or HRT cycles for pregnancy outcomes have been indicated. These benefits include a significant increase in the rate of clinical pregnancy and live birth, and a notable decrease in miscarriage rate, among women who underwent early cleavage or blastocyst ET using letrozole in FET approach in comparison with those who underwent natural or HRT cycles [ 13 – 17 ]. In light of these findings, we designed our present study based on mST-FET cycle with letrozole. Compared to natural and stimulated cycles, emerging evidences have indicated that artificial FET cycles exhibit a greater risk of hypertensive disorders, such as preeclampsia [ 18 , 19 ]. The presence of the corpus luteum significantly reduces the likelihood of pregnancy complications, such as hypertensive disorders, in NC-FET and mST-FET cycles [ 18 , 20 ]. The corpus luteum secretes vasoactive hormones, such as relaxin, which are not replicated in artificial cycles, potentially compromising maternal cardiovascular adaptations to pregnancy and resulting in an increased likelihood of preeclampsia [ 1 , 21 , 22 ]. Furthermore, some studies have demonstrated that true NC-FET without hCG trigger (tNC-FET) is more effective than mNC-FET [ 13 , 23 , 24 ]. The effects of hCG trigger on stimulated cycle FET with letrozole (L-FET) and NC-FET have been examined previously, revealing that the live birth rate was not impacted by the hCG trigger [ 8 ]. Therefore, the purpose of this study was to assess the outcomes of triggered or spontaneous ovulation in mildly stimulated FET cycle with letrozole plus low-dose gonadotropins (mST-FET).