The early luteal hormonal profile in IVF patients triggered with hCG

Abstract

STUDY QUESTION

What is the early luteal phase hormonal profile in patients undergoing ovarian stimulation for IVF/ICSI followed by hCG trigger and a freeze-all strategy without luteal phase support?

SUMMARY ANSWER

The peak concentration of progesterone occurred 4 days after oocyte pick-up (OPU + 4), with an average 35% fall from OPU + 4 to OPU + 6, and progesterone levels before and 12 h after hCG administration predicted levels during the early luteal phase.

WHAT IS KNOWN ALREADY

The luteal phase during IVF differs from that during normal cycles, particularly with respect to the serum progesterone level profile. This can cause asynchrony between the embryo and the endometrium, potentially resulting in implantation failure and poor reproductive outcomes.

STUDY DESIGN, SIZE, DURATION

This prospective study included 161 women with normal ovarian reserve receiving GnRH antagonist co-treatment during ovarian stimulation with FSH who were followed up to 6 days after OPU in a single IVF cycle.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Women aged 18–42 years undergoing IVF with ovarian stimulation using FSH were included. Ovulation was triggered with recombinant hCG 250 μg. Hormone levels were determined from blood samples taken on the day of trigger, before hCG, at 12, 24 and 36 h after hCG and at 1, 2, 3, 4, 5 and 6 days after OPU. The primary endpoint was early luteal phase serum concentrations of progesterone, LH, estradiol and hCG.

MAIN RESULTS AND THE ROLE OF CHANCE

One outlier with a pre-hCG serum progesterone level of 11.42 ng/mL was excluded, so all analyses included 160 subjects. Progesterone levels began to increase 1 day after OPU, peaked 4 days after OPU (114 ng/mL), then declined from OPU + 5 onwards. Peak progesterone levels were at OPU + 4, OPU + 5 or OPU + 6 in 38.8, 29.4 and 13.8% of patients, respectively. Approximately two-thirds of patients had a fall in serum progesterone from OPU + 4 to OPU + 6. Pre-hCG progesterone levels correlated significantly with those at 24 h after hCG (r² = 0.28; P < 0.001), which in turn correlated significantly with progesterone at OPU + 4 (r² = 0.32; P < 0.001). LH peaked (4.4 IU/L) 12 h after hCG trigger, persisting for 24 h but was barely elevated compared with physiological levels. Serum estradiol peaked twice: at 24 h post-trigger and at OPU + 4. Highest hCG levels (130 mIU/mL) occurred at 24 h post-injection. The best correlations between the number of follicles ≥11 mm and serum progesterone level were seen at 24 and 36 h after hCG and OPU + 1.

LIMITATIONS, REASONS FOR CAUTION

The influence of different profiles of serum progesterone on reproductive outcomes could not be determined because a freeze-all strategy was used in all patients. In addition, data were not available to relate serum hormone level findings with endometrial histology or endometrial receptivity analysis to clearly identify the relationship between serum hormones and the window of implantation.

WIDER IMPLICATIONS OF THE FINDINGS

Detailed information about early luteal phase hormone levels could be used to optimize and individualize luteal phase support to improve reproductive outcomes.

STUDY FUNDING/COMPETING INTEREST(S)

This study was funded by My Duc Hospital, Ho Chi Minh City, Vietnam. All authors state that they have no conflicts of interest to disclose.

TRIAL REGISTRATION NUMBER

NCT02798146; NCT03174691.

Introduction

In IVF cycles, the luteal phase starts from the time of early corpus luteum formation after ovulation trigger to implantation, or to menstruation if implantation does not occur. The luteal phase of IVF cycles is shorter than that of the natural cycle due to supraphysiological progesterone and estradiol levels during the early luteal phase induced by the hCG trigger (Jones, 1996; Fatemi et al., 2007; Yanushpolsky, 2015). In particular, progesterone significantly reduces the production of luteinizing hormone (LH) via negative feedback mechanisms on the hypothalamus and pituitary. As LH activity is crucial for the function of the corpus luteum, the significant reduction in this gonadotropin following trigger will result in a relative corpus luteum malfunction, necessitating luteal phase support at least until pregnancy is well detected 12–14 days after embryo transfer (Edwards et al., 1980; Tavaniotou et al., 2001; van der Gaast et al., 2002; Tavaniotou and Devroey, 2006). Thus, the early luteal progesterone profile in IVF differs markedly from the progesterone profile of the natural, unstimulated cycle, in which the peak of progesterone is reached around 6–8 days after ovulation—the time of expected implantation (Ross et al., 1970; Lenton et al., 1982; Navot et al., 1991; Reed and Carr, 2018). In contrast, the premature early luteal phase rise in progesterone appearing after ovarian stimulation with exogenous gonadotropins and hCG trigger advances the window of implantation, causing asynchrony between the embryo and the endometrium, which may result in implantation failure and poor reproductive outcomes (Ubaldi et al., 1997; Bourgain et al., 2002; Kolibianakis et al., 2002).

During IVF treatment, the early luteal phase LH activity deficit induced by high early luteal phase progesterone will be partly compensated for by the bolus of human chorionic gonadotropin (hCG) used for triggering of final oocyte maturation due to the long half-life of hCG (Weissman et al., 1996a; Fauser et al., 2002; Yding Andersen and Vilbour Andersen, 2014). In addition, the luteal phase will be covered by exogenous progesterone supplementation, and after successful implantation, the embryo itself will provide supportive LH activity, securing the function of the corpus luteum (Belaisch-Allart et al., 1983; Shapiro et al., 2012; Yding Andersen and Vilbour Andersen, 2014). However, it has been suggested that there is an hCG/LH deficiency period during the early luteal-peri-implantation phase after hCG trigger, during which corpus luteum stimulation is suboptimal (Yding Andersen and Vilbour Andersen, 2014).

In general, there is a limited amount of published data about the early luteal steroid profile after hCG trigger (Ragni et al., 2001; Fauser et al., 2002; Beckers et al., 2003). This information is important to allow provision of the most optimal luteal phase support strategies during IVF, with the goal of increasing live birth rates. However, existing studies have been limited by small patient populations and a low number of blood samples for steroid level determination during the early luteal phase (Ragni et al., 2001; Fauser et al., 2002; Beckers et al., 2003).

To improve our understanding of the early luteal phase after hCG trigger, this study was performed exploring early luteal phase hormonal profiles in patients undergoing ovarian stimulation for IVF/ICSI followed by hCG trigger and freeze-all.

Materials and Methods

Study design

This single-center prospective study was conducted at IVFMD, My Duc Hospital, Ho Chi Minh City, Vietnam, over the period June 2016 to July 2017. The initial study protocol (NCT02798146) included a target sample size of 30 patients; however, this was later extended so that an additional 130 patients were recruited (NCT03174691). Both phases of the study received ethical approval from the Ethical Board of My Duc Hospital (approval numbers 01/16/ĐĐ-BVMĐ and 10/17/ĐĐ-BVMĐ, dated 31 May 2016 and 10 May 2017, respectively). All patients provided written informed consent prior to enrolment in the study, which was conducted in accordance with the ICH Harmonised Tripartite Guideline for Good Clinical Practice and the ethical principles of the Declaration of Helsinki. The protocol was identical during the two study phases, and results were combined for all analyses.

Study subjects

Subjects aged 18–42 years with a body mass index (BMI) <28 kg/m² who were undergoing IVF followed by a freeze-all cycle after hCG trigger were eligible for the study and were consecutively enrolled in the study. Additional inclusion criteria were normal ovarian reserve, defined as anti-Müllerian hormone (AMH) level >1.25 ng/mL or antral follicle count (AFC) ≥6, measured within the 2 months prior to starting stimulation; GnRH antagonist co-treatment during ovarian stimulation; ability to comply with the requirements of the study protocol; and provision of informed consent. Patients were excluded if they had previously had a poor response to stimulation (≤3 oocytes) or a hyper-response (>20 follicles at ≥14 mm) after high-dose FSH stimulation, if they were participating in another clinical trial or if they had a chronic medical condition (e.g. diabetes mellitus, Crohn’s disease, thyroid disease, hepatitis B, sexually transmitted diseases).

Ovarian stimulation, monitoring, trigger and oocyte retrieval

Stimulation, monitoring and oocyte retrieval were performed using a GnRH antagonist (Orgalutran, Merck Sharp and Dohme, The Netherlands) protocol. FSH (Puregon, Merck Sharp and Dohme, The Netherlands) was started on the second day of menstruation; the starting dose was 150–300 IU/day, depending on age, AMH level and response to FSH in any prior IVF cycle. Ultrasound and blood tests were performed starting from stimulation day 5 or 6; the number of follicles of both ovaries with a diameter of ≥11 mm was recorded. When the mean diameter of at least two leading follicles was 17 mm, ovulation was triggered with a s.c. bolus of 250 μg recombinant human chorionic gonadotropin (r-hCG) (Ovitrelle, Merck Serono, Germany), followed by oocyte retrieval at 36 h later after hCG trigger.

Early luteal phase blood sampling

Blood samples (2 mL) for determination of hormone levels were collected on the day of trigger, prior to the injection of hCG at approximately at 8 pm, 12 h after hCG injection (approximately 8 a.m.), 24 h after hCG injection (approximately at 8 p.m.), 38 h after hCG injection (2 h after oocyte pick-up [OPU]; approximately 10 a.m.) and at 1, 2, 3, 4, 5 and 6 days after OPU—all at approximately 8 a.m. All samples were processed immediately and stored at −20°C. Serum hormone levels were determined using electrochemiluminescence immunoassay (ECLIA; Roche Cobas E 801, Roche Diagnostics, Germany). Lower level of quantification, inter-assay variability and intra-assay variability were 0.1 mIU/mL, 2–5% and 2–5% for LH; 0.1 mIU/mL, 2–5% and 2–5% for beta hCG; 0.5 ng/mL, 2–6% and 2–4% for progesterone; and 5 pg/mL, 2–6% and 2–4% for estradiol.

Endpoints

The primary endpoint was serum concentrations of progesterone, LH, estradiol and hCG during the early luteal phase. The correlation between the number of follicles ≥11 mm and serum progesterone levels during the early luteal phase was determined as a secondary endpoint.

Statistical analysis

No formal sample size calculation was performed due to the scarcity of existing data on early luteal phase hormone levels. In the analysis of study data, continuous variables are presented as mean ± standard deviation (SD) and were compared using the Student’s t test. Categorical data are expressed as frequencies and percentages and were compared using the chi-square test. Data were analyzed using SPSS version 20 software. All tests are two tailed, and a P value of <0.05 was considered statistically significant.

Results

Patients

A total of 161 patients were enrolled in the study, but one outlier with a pre-hCG progesterone level of 11.42 ng/mL was excluded, meaning that 160 patients were included in the analysis (age 25–41 years, BMI 15.6–27.1 kg/m²) (Table I). The most common indications for IVF were male factor or unexplained infertility, and a freeze-all strategy was used most often due to patient preference or an unfavorable endometrium (defined as endometrial thickness <7 mm in a previous IVF cycle or the presence of fluid in cavity) (Table I).

Stimulation outcomes

Outcomes after ovarian stimulation, including number of follicles, oocytes, embryos and good quality embryos, were consistent with clinical practice at our center (Table II). The luteal phase duration was 8–13 (mean 10.0) days, and one patient developed moderate ovarian hyperstimulation syndrome (Table II).

Serum progesterone

Overall, 8/160 patients (5%) had a pre-hCG serum progesterone level ≥3 ng/mL. Serum progesterone levels started to increase 1 day after OPU, peaked at 4 days after OPU (at 113.64 ng/mL), then started to decline from 5 days after OPU onwards (Table III; Fig. 1). Overall, peak progesterone level occurred at OPU + 2 in 2 patients (1.3%), at OPU + 3 in 28 patients (17.5%), at OPU + 4 in 62 patients (38.8%), at OPU + 5 in 47 patients (29.4%) and at OPU + 6 in 22 patients (13.8%). The majority of patients (≈65%) had a fall in serum progesterone from OPU + 4 to OPU + 6.

There was a significant correlation between pre-hCG progesterone levels and those at 24 h after hCG administration (r² = 0.28; P < 0.001), and between serum progesterone at 24 h after hCG and serum progesterone at 4 days after OPU (r² = 0.32; p < 0.001) (Fig. 1A); correlations were also seen between serum progesterone and serum hCG (Fig. 1).

Serum LH

Peak LH levels (4.4 IU/L) occurred at 12 h after the hCG trigger injection and lasted for 24 h (Fig. 2).

Serum hCG

Levels of hCG were highest at 24 h after injection (peak 130 mIU/mL) and then declined gradually until OPU + 6 (Fig. 2).

Serum estradiol

Serum estradiol levels showed two peaks: one 24 h after trigger and another lower peak 4 days after OPU (Fig. 2).

Correlation between serum progesterone and number of follicles

There were weak correlations between the number of follicles ≥11 mm and serum progesterone level from before hCG injection to 6 days after OPU (Fig. 3). The best correlations were seen at 24 and 36 h after hCG injection and 1 day after OPU (Fig. 3).

Discussion

This study, to the best of our knowledge, is the first study to provide detailed information on luteal phase progesterone concentrations after a bolus trigger of hCG in a large number of patients who did not receive any exogenous luteal phase support. The data clearly demonstrated that peak progesterone concentrations occur 4 days after OPU, which is earlier than in the natural menstrual cycle (Ross et al., 1970; Lenton et al., 1982; Navot et al., 1991; Reed and Carr 2018). In addition, on day OPU + 6, the progesterone concentration had already decreased by 35% compared with the peak concentration on day OPU + 4 and continued to decline steadily. This clearly demonstrates that the course of progesterone levels during hCG-triggered cycles is asynchronous with the time course for normal implantation on day OPU + 7.

In this context, it is noteworthy that more than 40% of patients had a >50% decrease in progesterone concentration between OPU + 4 and OPU + 6 (data not shown). Although exogenous progesterone supplementation in women receiving IVF/ICSI treatment will ameliorate this drop in serum progesterone, our findings highlight the requirement for studies examining how the probability of achieving pregnancy in fresh cycles is affected by the timing and magnitude of the reduction in progesterone concentrations.

The current study also demonstrates that late follicular phase progesterone concentrations are significantly associated with progesterone concentrations measured at all time points until day OPU + 1. This confirms that follicular LH sensitivity is reflected by late follicular phase progesterone concentrations, which then determine the early luteal phase response, as has recently been suggested (Wang et al., 2019). Progesterone concentrations at hCG + 12 h or hCG + 24 h associated better with progesterone concentrations at all other time points than did the late follicular progesterone concentration before hCG trigger. Progesterone concentration at hCG + 12 h or hCG + 24 h can thus be used to determine the course of serum progesterone levels during the rest of the measured luteal phase. Thus, blood sampling at either or both of these times might be sufficient in clinical practice. The underlying mechanism for the significant associations between progesterone concentrations prior to OPU and those at later time points was not investigated in this study, but may reflect differences in the turnover and clearance of hCG. The study findings suggest that individualization of luteal phase support may be best performed based on progesterone concentrations measured at hCG + 12 h or hCG + 24 h.

There is an extensive body of literature suggesting that women with increased late follicular phase progesterone concentrations have a reduced pregnancy rate when embryos are replaced in the fresh IVF cycles (Xu et al., 2012; Venetis et al., 2013). The significant associations between pre-hCG progesterone concentrations and the progesterone concentrations at OPU + 12 h, OPU + 24 h and OPU + 36 h (but not at later time points) suggests that the negative effect may be exerted during this early period of the luteal phase and may reflect that a very steep rise in serum progesterone in the early luteal phase (or high concentrations compared with the natural menstrual cycle) has a negative effect or advances the endometrium and reduces implantation (Friis Wang et al., 2019). If confirmed in additional studies, these data favor the use of an agonist trigger that induces a more physiological mid-cycle surge of gonadotropins with a slower increase in progesterone concentrations.

It is interesting to notice that the peak progesterone concentration during hCG trigger was more than 10 times higher than that observed during the natural menstrual cycle (approximately 10 ng/mL) (Groome et al., 1996). This fits very nicely with the number of follicles exceeding 14 mm in diameter at OPU, which averaged 10 in this study and demonstrates that each corpus luteum during ovarian stimulation produces a similar amount of progesterone as that produced by only one corpus luteum during the natural menstrual cycle.

The present study demonstrated that a bolus of 250 μg r-hCG induced a peak serum hCG level of 130 mIU/mL at 12 h after injection. The profile of serum beta-hCG levels after SC administration of a 6500 IU dose in our study including only normo-gonadotropic women was similar to that reported after SC administration of hCG 10 000 IU in an early pharmacokinetic study conducted in two hypogonadal women (Weissman et al., 1996b). Both studies showed that, despite high levels of hCG in the early luteal phase, levels of hCG decline and almost disappear around the mid-luteal phase (at day OPU + 6), confirming that the drive for progesterone synthesis at that crucial time of implantation is almost absent. Thus, although an injection of 10 000 IU will maintain higher levels of hCG at a given time point compared with a lower dose, hCG doses used in clinical practice will invariably be too low on OPU + 6 to provide full support of progesterone output.

Progesterone is the most important hormone of the luteal phase for implantation. In our study, serum progesterone levels started to decline 5 days after OPU, which is earlier than in natural cycles (Groome et al., 1996). Overall, in terms of serum progesterone, the findings of the present study support current clinical practice of providing luteal phase support with progesterone during peri-implantation in stimulated IVF cycles when an hCG trigger is used for final oocyte maturation (Smitz et al., 1992; Tavaniotou et al., 2000; Lawrenz et al., 2019; Mohammed et al., 2019). For luteal phase support, it would appear more appropriate to focus on addressing LH deficiency by supplementing with LH, hCG or GnRHa (Bar-Hava et al., 2016; Bar Hava et al., 2017; Papanikolaou et al., 2011; Yding Andersen and Vilbour Andersen, 2014). In addition, an appropriate dose of FSH for ovarian stimulation is required to avoid FSH-induced up-regulation of LH receptors, which appears to contribute to high early progesterone levels and subsequent implantation failure (Filicori et al., 2002; Friis Wang et al., 2019). Thus, we consider that it might be beneficial to individualize not only the stimulation with exogenous gonadotropins but also the luteal phase support in IVF patients.

LH stimulates the corpus luteum to produce progesterone in addition to sex steroids, growth factors and cytokines (Wang et al., 1976; Devoto et al., 2009). In the present study, LH levels peaked within 12 h of trigger but were barely elevated compared with physiological levels, suggesting that endogenous LH plays little, if any, role in providing the luteal transforming signal in connection with an exogenous bolus trigger of hCG. Furthermore, the progesterone peak was not reached until 4 days after OPU. Therefore, if LH or hCG is administered for luteal phase support, the question is when supplementation should be initiated. Since the early luteal phase rise in progesterone is pronounced, it may be argued that exogenous LH or hCG should only be administered when the corpus luteum demise starts, which occurs approximately on OPU + 3 or OPU + 4 in GnRHa-triggered cycles (Kol and Humaidan, 2013; Vuong et al., 2016). In primates, the corpus luteum may still be rescued after 3 days of LH deprivation (Hutchison and Zeleznik, 1985) and after 7 days of LH deprivation as shown in an early pilot study including three hypogonadotropic-hypogonadism patients (Weissman et al., 1996b). However, the effect of delaying luteal phase support on the reproductive outcome is not yet determined and the present study suggests that endometrium damage might already take place during the early progesterone rise shortly after hCG trigger.

The correlation between the number of follicles of ≥11 mm and the serum progesterone level was investigated in the present trial because it has previously been hypothesized that the higher the number of follicles, the higher the progesterone level (Friis Wang et al., 2019; Yding Andersen et al., 2011). The presence of a correlation would be clinically relevant because appropriate FSH dosing is required to avoid hyper-responses, which lead to an excessive number of follicles that, in turn, has been associated with a rise in serum progesterone and, therefore, a poor reproductive outcome.

Overall, our data contribute to the body of evidence suggesting that steroid levels determine luteal phase characteristics. A key strength of the study is its prospective design, including a large sample size. In addition, this is the first study to address early luteal phase hormone levels in detail. Ten blood samples were collected from each patient to more accurately describe the hormone profile during the luteal phase. However, we acknowledge that serum progesterone levels need to be interpreted taking into account the immunoassay used for the study, which may limit comparisons between studies and the generalizability of our findings. The current study was not able to determine the impact of the different courses of progesterone concentrations on reproductive outcome because all patients followed a ‘freeze-all’ strategy. This is a limitation of the present study, but the magnitude and timing of differences in progesterone concentrations are now clearly documented, explaining the asynchrony in endometrial receptivity that has occurred during IVF for many years. Unfortunately, we did not have data to correlate serum hormone level findings with endometrial histology or endometrial receptivity analysis to clearly identify the relationship between serum hormones and the window of implantation. Such an analysis would be a useful inclusion in future studies.

In conclusion, this study provides detailed information about the early luteal phase after hCG trigger in IVF/ICSI cycles. Later cycle progesterone levels correlate well with those obtained during the luteal phase. In addition, progesterone levels before and 12 h after hCG administration predict levels during the early luteal phase. Taken together, these findings suggest that late follicular progesterone levels determine the likelihood of achieving pregnancy. Although further work is needed to determine how the results can be specifically applied in clinical practice, data from this study could be used in future research to determine the optimal protocols for luteal phase support and how to individualize luteal phase support to improve reproductive outcomes.

Authors’ roles

L.N.V. was involved in the study design, execution, analysis, manuscript drafting, critical discussion and final approval of the manuscript. T.M.H. was involved in the study design, execution, critical discussion and final approval of the manuscript. T.D.P. was involved in the analysis, critical discussion and final approval of the manuscript. V.N.H. was involved in the execution, critical discussion and final approval of the manuscript. C.Y.A. was involved in the critical discussion and final approval of the manuscript. P.H. was involved in the study design, critical discussion and final approval of the manuscript.

Funding

My Duc Hospital, Ho Chi Minh City, Vietnam.

Conflict of interest

All authors state that they have no conflicts of interest to disclose.

References