Progesterone Vaginal Gel or Combined Medication for Luteal-Phase Support of Frozen-Thawed Embryo Transfer Cycles: A Single-Centre, Chinese, Randomized, Open-Label, Pilot Study

Frozen embryo transfer (FET) is becoming increasingly common in China and globally [1, 2]. In China, the number of FET cycles performed annually increased by ∼85% from around 82,000 in 2013 to 152,000 in 2016. Over the same period, rates of procedures using fresh embryo transfers (ETs) remained relatively stable, at 71–78,000 in vitro fertilization and 25–28,000 intracytoplasmic sperm injection cycles [1]. Compared with fresh ET, FET can be particularly beneficial for women at risk of ovarian hyperstimulation syndrome and can improve live birth rates in individuals who are high responders to ovarian stimulation [3‒5].

Progesterone (PG) plays a key role in embryo-endometrial synchrony and the maintenance of early pregnancy by enabling the endometrium to become receptive to embryo implantation [6, 7]. This is particularly important for FET as the lack of concurrent ovarian stimulation and resulting lack of corpus luteum development mean that there is no endogenous PG production to facilitate endometrial preparation [7]. Globally, however, there is no consensus as to whether the route or dosage of PG affects clinical outcomes in FET cycles. In one meta-analysis of studies with serum PG levels <10 ng/mL, higher serum PG was associated with higher rates of ongoing pregnancy/live birth (relative risk 1.47; 95% confidence interval [CI] 1.28, 1.70), higher chance of clinical pregnancy (relative risk 1.31; 95% CI: 1.16, 1.49), and lower risk of early pregnancy loss (relative risk 0.62; 95% CI: 0.50, 0.77) [8]. Although this suggests that there may be a minimum clinically effective dose of PG for luteal-phase support (LPS) in FET, further research is needed to confirm this working hypothesis.

In China, combined treatment with vaginal and oral PG (VPG + OPG) is preferred for LPS in hormone replacement therapy (HRT)-FET cycles, with one consensus publication endorsing VPG 90 mg + OPG 20 mg daily [9]. The approach of using VPG 90 mg + OPG 20 mg daily resulted in a clinical-pregnancy rate of 40.1% following FET in a Chinese population [10]. The optimal dose of VPG for LPS in FET is unclear; however, some Chinese retrospective studies advocate the use of 90 mg VPG daily [11‒15]. It remains unclear whether a lower dose of PG is appropriate for FET, or whether a higher dose is associated with better outcomes. Therefore, a pilot study was undertaken to compare VPG monotherapy (90 mg daily), or combination treatment (VPG 90 mg daily + OPG 20 mg daily), for LPS in HRT-FET cycles, by assessing the impact of PG treatment on pregnancy outcomes and safety. The main aim of the study was to provide preliminary evidence to compare the impact of the two PG regimens on efficacy (in particular, ongoing pregnancy rates) and safety (adverse events [AEs], such as bleeding) and clarify the optimal regimen of PG to support pregnancy outcomes.

This was a Chinese, single-centre, open-label, randomized controlled, pilot study that compared the efficacy and safety of VPG alone versus VPG + OPG (ClinicalTrials.gov identifier: NCT03858049; the first trial registration date was February 26, 2019) in participants undergoing HRT-FET cycles. Eligible participants attended the clinic for baseline evaluation and then on endometrial transformation day (Day −5) to measure their endometrial thickness, at which time they were reassessed for study eligibility. Eligible participants then received VPG 90 mg daily until Day −1 to assist in the secretory transformation of the endometrium. On the day of FET (Day 1), participants were randomized to either VPG or VPG + OPG. On the same day, participants in both groups underwent FET with 2 or more Day 5 embryos. All participants received oral oestradiol valerate, dosed according to local clinical practice.

On Day 14 following FET (Day 14), serum β-human chorionic gonadotropin (β-hCG) concentration was measured in each participant for early detection of pregnancy. If the β-hCG pregnancy assessment was positive, it was further confirmed by a clinical-pregnancy assessment on Days 28–35 after FET. Participant follow-up continued until ongoing pregnancy was confirmed on Day 63 after FET and/or at the end of the study (Fig. 1).

The protocol, informed consent form, and other relevant study documents were approved by the Peking University Third Hospital Ethics Committee (Approval No. D2018096) prior to the initiation of the study. All study participants provided written informed consent before being enrolled.

The study enrolled infertile Chinese women aged 20–38 years who underwent HRT-FET, following both failed fresh and freeze-all ET. All participants were required to have a transitional endometrium of ≥8 mm. Key exclusion criteria included previous failure of three or more cycles of fresh or frozen-thawed ET and having any condition or disease that meant pregnancy was impossible. Detailed inclusion and exclusion criteria are provided in online supplementary Table 1 (for all online suppl. material, see https://doi.org/10.1159/000544065).

Participants in both treatment groups received a 90 mg dose of 8% PG gel (Crinone^®; Merck Serono Co., Ltd., Darmstadt, Germany) intravaginally every morning. Those randomized to VPG + OPG also received a 10 mg PG tablet (Duphaston^®; Abbott, Chicago, IL, USA), administered orally twice daily. All participants received oral oestradiol according to local clinical practice.

Baseline characteristics, including age and age category at FET, as well as body mass index were recorded. Medical history was summarized using Medical Dictionary for Regulatory Activities version 24.0 terminology [16]. FET characteristics were also collected at the baseline visit, which preceded the endometrial transformation visit (Day −5). Concomitant treatments, other than study treatment or oestradiol, and medical procedures that were administered or performed at any time during the study were defined as medications. Treatment compliance was calculated by comparing the number of actual versus planned doses.

The primary endpoint was ongoing pregnancy at 10–12 weeks of pregnancy (8–10 weeks after FET). Ongoing pregnancy was defined as the presence of a viable intrauterine foetus detected by ultrasound examination at 12 weeks of pregnancy.

Secondary endpoints included β-hCG positivity at week 4 of pregnancy (2 weeks post FET); implantation rate, defined as number of observed gestational sacs versus number of embryos transplanted, at 6–8 weeks of pregnancy (4–6 weeks post FET); clinical pregnancy at 6–8 weeks of pregnancy, assessed by ultrasound (4–6 weeks post FET); early pregnancy loss from ET up to 12 weeks of pregnancy; luteal-phase bleeding from ET to β-hCG positivity; and vaginal bleeding following a positive β-hCG test.

The safety endpoints included AEs, deaths, laboratory testing for β-hCG, and vital signs (baseline only). Treatment-emergent AEs (TEAEs) were defined as events with onset dates occurring within the treatment period, which was defined as from the first day of study intervention after randomization until the date of the last dose of study treatment, or the clinical cut-off date, whichever occurred earlier. TEAEs, serious TEAE, TEAEs by worst severity, and TEAEs that caused treatment discontinuation were recorded. Treatment-related AEs (TRAEs) were those events with relationship to study medication listed as missing, unknown, or yes. TRAEs, serious TRAEs, TRAEs by worst severity, TRAEs causing treatment discontinuation, and deaths due to TRAEs were recorded.

Participants were randomized in a 1:1 ratio on the day of ET using an interactive voice response system with the permuted block randomization method (IRTON; Shanghai, China). The automated randomization system ensured that there was no allocation bias in the open-label trial. Randomization was stratified by age group (ages 20–35 years, or age ≥35 years) to ensure a balanced number of participants within each age stratum in the VPG and VPG + OPG treatment groups.

The sample size calculation was based on an assumed ongoing pregnancy rate of 40% for the VPG group and 45% for the VPG + OPG group. With a sample size of 150 per group after randomization, the half-width of the 90% CI of the difference between the VPG and VPG + OPG groups would be 0.09. Assuming a dropout rate of 10% after randomization, 334 participants were planned to be randomized in this study from a single reproductive centre over a 25-month period. The intention-to-treat (ITT) set included all participants who received one or more doses of study treatment after randomization. The ITT set was analysed according to the study intervention initially assigned to the participants. The per-protocol (PP) set was defined as all participants in the ITT set without any important protocol violations. Important protocol deviations were defined as those that may have significantly affected the completeness, accuracy, and reliability of the study data or that may have significantly affected a participant’s rights, safety, or well-being. Important protocol deviations, including deviations from inclusion/exclusion criteria and post-inclusion deviations, were identified and confirmed prior to, or at, the Data Review Meeting. The ITT set was used to assess baseline characteristics, concomitant therapies, and efficacy. Efficacy outcomes were also analysed in the PP set. In the safety set, which included all participants who received one or more doses of study treatment, participants were analysed according to the actual study intervention they received. The safety set was used for the safety analysis, treatment compliance, and treatment exposure.

Continuous and qualitative variables were summarized using descriptive statistics and counts/percentages, respectively. No formal statistical hypothesis testing was performed. For each primary and secondary efficacy endpoint, point estimate and 95% CI for each study intervention group were calculated using the Clopper-Pearson method. Point estimate and 90% CI for difference between two study intervention groups were calculated using the stratified Miettinen-Nurminen method [17]. It was computed with the SAS^® 9.4 (SAS Institute, Cary, NC) FREQ procedure, using the riskdiff (COMMON) option in the Table statement and specified alpha = 0.1.

The study started on June 4, 2019, and was stopped on March 15, 2021. The study design is illustrated in Figure 2. Of the 208 women screened, 172 were randomized to VPG 90 mg daily (n = 86) or VPG 90 mg daily + OPG 20 mg daily (n = 86). Thirty-six participants were not randomized because they withdrew consent or did not fulfil the study eligibility criteria. All randomized participants received one or more doses of study medication.

Among the 172 randomized participants in the ITT set, 16 (18.6%) in the VPG and 9 (10.5%) in the VPG + OPG group discontinued the study early because of AEs. Two participants discontinued for other reasons: one refused the blood β-hCG test during menstruation (VPG group) and one had a VPG overdose (VPG + OPG group).

Of the 86 participants in each treatment group (ITT set), the study was completed by 69 participants (80.2%) in the VPG group and 76 (88.4%) in the VPG + OPG group. During the study, a tendency towards the use of increasing dosages of PG for LPS was observed. After a comprehensive assessment by the study sponsors, it was decided to terminate the study early; consequently only 172 of the planned 334 participants were randomized. Because this early termination resulted in a smaller sample size than planned, this trial should be considered a pilot study.

All participants were Chinese, with no marked differences in body mass index (Table 1). The mean age at ET was 30.8 years across both treatment groups. In terms of age categories, 14.0% and 9.3% in the VPG and VPG + OPG groups, respectively, were ≥35 years old at FET.

Medical history profiles were similar between treatment groups across most categories, except for one spontaneous abortion in those with a previous pregnancy (19.6% and 7.1% with VPG and VPG + OPG, respectively), mean duration of infertility (60.0 and 50.5 months with VPG and VPG + OPG, respectively), male-only infertility (10.5% and 20.9% with VPG and VPG + OPG, respectively), and “other” causes of female infertility (19.8% and 9.3% with VPG and VPG + OPG, respectively) (Table 2).

Polycystic ovary syndrome (47.7% and 46.5% with VPG and VPG + OPG, respectively), fallopian tube obstruction (9.3% and 14.0% with VPG and VPG + OPG, respectively), and uterine fibroids/leiomyoma (5.8% and 15.1% with VPG and VPG + OPG, respectively) were the most commonly reported reproductive system disorders.

One participant (1.2%) in the VPG treatment group received two embryos while all remaining participants received one embryo (Table 3). The proportion of “good-quality embryos” graded AA, AB, BA, or BB according to the Gardner embryo grading system [18] was higher in the VPG + OPG group (72%) compared with the VPG group (58%).

There was no major difference in the proportion of women taking concomitant medications between the VPG group (74/86 participants, 86.0%) and the VPG + OPG group (75/86 participants, 87.2%). The top three concomitant medications coded included multivitamins (45.3% and 32.6% with VPG and VPG + OPG, respectively), folic acid (19.8% and 31.4% with VPG and VPG + OPG, respectively), and acetylsalicylic acid (20.9% and 17.4% with VPG and VPG + OPG, respectively).

Eight participants, four (4.7%) per treatment group, underwent one or more concurrent medical procedures during the study, mostly resulting from serious AEs (SAEs). These included investigations (one participant [1.2%] in each treatment group) and surgical/medical procedures (4 participants [4.7%] in each treatment group). Procedures included uterine curettage, uterine dilation and curettage, laparoscopy, lysis of pelvic adhesions, right salpingectomy, medical abortion, and bilateral salpingectomy.

All participants received one or more doses of study medication. Compliance with study treatment was ≥80% for 100% of women (n = 86) in the VPG group and 98.8% (n = 85) of those in the VPG + OPG group.

Results for the primary and secondary efficacy endpoints were similar between the ITT analysis (presented below) and the PP analysis (data not shown).

A numerically higher ongoing pregnancy rate was observed with VPG + OPG (25/86 participants, 29.1%) versus VPG alone (16/85 participants, 18.8%), with treatment difference of 8.4% (90% CI: −2.2%, 19.0%) in the ITT set (Table 4).

There was a treatment difference of 0.9% (90% CI: −10.8%, 12.7%) in β-hCG positivity between the VPG + OPG group (37/86 participants, 43.0%) and the VPG-only group (32/85 participants, 37.6%).

In the VPG + OPG group, 29/86 (33.7%) of women achieved successful implantation compared with 18/86 (20.9%) of women in the VPG-only group, equating to a 10.4% (90% CI: −0.5%, 21.3%) treatment difference.

There was a treatment difference for clinical pregnancy, with a rate of 10.1% (90% CI: −0.8%, 21.1%) favouring VPG + OPG (29/86 participants, 33.7%) versus VPG alone (18/85 participants, 21.2%).

Four participants (13.8%) in the VPG + OPG group and two (11.1%) in the VPG group experienced early spontaneous abortion. The between-group treatment difference was 0.1% (90% CI: −18.2, 18.5).

The overall incidences of TEAEs, moderate and severe TEAEs, SAEs, and bleeding rates were comparable between both treatment groups (Table 5). A total of 22/86 participants (25.6%) in the VPG group and 19/86 (22.1%) in the VPG + OPG group experienced TEAEs, and no events were considered to be related to study treatment. In terms of TEAE severity, 5/86 participants (5.8%) in the VPG group and 4/86 (4.7%) in the VPG + OPG group experienced moderate TEAEs. Additionally, 1/86 participants (1.2%; decreased embryo viability) in the VPG group and 3/86 participants (3.5%; two decreased embryo viability, one spontaneous abortion) in the VPG + OPG group experienced severe TEAEs. All severe TEAEs led to the discontinuation of the relevant study medication, and all of them met the SAE criteria. Severe TEAEs were resolved by the time of data cut-off in all cases. Only four participants (4.7%) in each treatment group experienced serious TEAEs, and no events were considered to be treatment related. TEAEs leading to discontinuation of one or more study treatments were experienced by 12/86 participants (14.0%) in the VPG group and 8/86 (9.3%) in the VPG + OPG group. More cases of biochemical pregnancy were reported in the VPG compared with the VPG + OPG group (9.3% and 3.5%, respectively). There were no deaths.

Luteal-phase bleeding was experienced by 1/85 participants (1.2%) in the VPG group and 3/86 participants (3.5%) in the VPG + OPG group, in the time between ET and β-hCG positive assessment. Vaginal bleeding was reported by 6/85 participants (7.1%) in the VPG group and 9/86 (10.5%) in the VPG + OPG group.

This pilot study found no significant difference between VPG and VPG + OPG in terms of ongoing pregnancy rate (primary endpoint), although a numerically higher ongoing pregnancy rate was observed in the group that received VPG 90 mg and OPG 20 mg daily versus VPG 90 mg daily alone. Similarly, no differences were observed between treatments for the secondary outcomes, although the VPG + OPG group was also associated with numerically higher secondary efficacy outcomes, including rates of β-hCG positivity, implantation, and clinical pregnancy. The safety profiles of both treatment regimens were comparable.

In FET-HRT cycles, endogenous PG is not produced because of the lack of corpus luteum development [19]. Therefore, LPS is required, including exogenous PG administration to support the secretory endometrium and promote embryo implantation [19]. Although the evidence supporting optimal PG dosing in FET is limited, other PG studies have shown similar findings to the numerically higher (albeit non-significant) ongoing pregnancy rate with a higher dosage of PG reported here. For example, several studies have reported that higher PG doses or PG supplementation were associated with improved pregnancy and live birth rates versus lower doses or non-supplementation in HRT-FET cycles [20‒25]. In a retrospective study that used the same type of VPG as was administered in the present study, 180 mg daily compared with 90 mg daily was associated with improved pregnancy (38.4% versus 26.7%, respectively; p = 0.021) and live birth (20.5% versus 8.7%, respectively; p = 0.002) rates [20]. Overall, these findings suggest that the total PG dosage, rather than the route of administration, is key in terms of pregnancy outcomes.

In terms of secondary efficacy outcomes, the present findings of numerically improved (not statistically significant) β-hCG positivity, implantation, and clinical pregnancy, as well as reduced early pregnancy loss, observed with the higher- versus lower-dose PG regime align with previous data. In a randomized trial of HRT-FET cycles, a higher-dose PG regimen (VPG 400 mg daily and intramuscular PG 50 mg every 3 days compared with VPG 400 mg daily alone) was associated with improved β-hCG positivity (62% vs 55%, respectively; p = 0.02), a higher proportion of women achieving clinical pregnancy (54% vs 37%, respectively; p < 0.0001), and a reduced rate of total pregnancy loss (26% vs 50%, respectively; p < 0.0001) [22]. The lower rates of β-hCG positivity and clinical pregnancy in the current study compared with the work of Devine et al. [22] likely reflect differences in the daily PG dose, which was higher in that study, and the quality of transferred embryos. In the present study, 72% of embryos in the VPG + OPG group and 58% in VPG group had an inner cell mass and trophectoderm grade of at least BB, compared with 100% in the earlier study [22]. Therefore, it is possible the relatively reduced embryo quality in the current study resulted in lower β-hCG positive and pregnancy rates.

The safety profiles of VPG and VPG + OPG observed in this study were consistent with the safety profile of VPG reported previously [25‒27]. A study that examined VPG 90 mg daily as LPS for fresh embryo in vitro fertilization cycles reported somewhat higher rates of overall TEAEs compared with the present study (48.6% vs 25.6%, respectively) and severe TEAEs (6.8% vs 1.2%, respectively) [27]. There were notable similarities between our study and the earlier work, such as rates of TEAEs leading to study termination (14.0% and 11.7%, respectively) and vaginal bleeds (7.1% and 8.5%, respectively) [27]. It is possible that the open-label design of the current study may have increased the risk of bias for subjective endpoints, such as the reporting of AEs. For instance, known side effects may be preferentially reported by patients and examined by clinicians. However, the current study found no major differences in TEAEs between the VPG versus VPG + OPG groups, and the primary endpoint (ongoing pregnancy) was non-subjective. Future related research should involve blinded trials with independent clinicians to reduce potential bias.

Strength of the present study is that it provides further information on the impact of PG dose on the clinical outcomes of FET cycles, for which there are limited data to influence evidence-based treatment. Participants were treated according to local clinical practice, so the results are generalizable to individuals receiving LPS for HRT-FET cycles in China, and for countries that use similar protocols for HRT-FET, noting the use of oral PG may be more prevalent in China compared with its limited use reported in other countries [28]. However, there may also be bias from being a single-centre study, such as protocol or ethnicity bias, and ongoing related research should involve multicentre trials. There were several other limitations of the current study. As mentioned earlier, there were differences between treatment groups in the number of good-quality embryos, with a lower proportion of good-quality embryos in the VPG group (58%) compared with the VPG + OPG group (72%). Differences in embryo quality may have affected implantation rates and the occurrence of biochemical pregnancy in the VPG versus VPG + OPG group as embryo quality has been shown to impact these criteria in other studies [29, 30]. Second, the ongoing pregnancy rates in this study were 18.8% and 29.1% for the VPG group and VPG + OPG group, respectively. This is relatively low for blastocyst-stage transfers. In comparison, another study reported ongoing pregnancies in around half of all cases following FET [31]. The difference in ongoing pregnancy rates between the present and past studies may reflect different approaches to ET. In the current study, good embryos were transferred on Day 3, whereas suboptimal embryos were cultured to blastocyst stage for transfer. This contrasts with the work of Shi et al. [31] in which only good-quality embryos were transferred on Day 3. Additionally, it was decided to terminate the current study early because of increasing doses of PG being administered for LPS. This may have reflected the fact that study clinicians were more familiar with using the combined VPG + OPG LPS regimen, so were concerned that VPG alone may result in a drop in pregnancy rates. Therefore, it is possible that increased PG dosages were administered to compensate for the perceived differences between VPG + OPG and VPG monotherapy. However, the early termination reduced the sample size of the study, substantially hindering the statistical power and ability to detect definitive differences between treatment groups. A final limitation of our study was the open-label design, which may have increased the risk of bias for subjective endpoints, such as AE reporting, as noted above.

In conclusion, a statistical conclusion cannot be reached considering the small sample size in this study, so it should be considered a pilot study. However, a numerically higher ongoing pregnancy rate with VPG + OPG was observed. These results suggest that a minimal dose of PG, such as the VPG 90 mg daily and OPG 20 mg daily combination therapy reported here, may be needed in HRT-FET cycles to support good outcomes in terms of implantation and clinical pregnancy. Additional studies with larger, well-powered, and multicentre trials are needed to elucidate the benefits and risks of further increasing the PG dose for LPS during HRT-FET cycles.

This study protocol was reviewed and approved by the Peking University Third Hospital Ethics Committee (Approval No. D2018096). The study was conducted in accordance with the protocol and the ethical principles derived from international guidelines including the Declaration of Helsinki, Council for International Organizations of Medical Sciences, International Ethical Guidelines, International Council for Harmonisation Good Clinical Practice (ICH GCP) guidelines, and other applicable laws and regulations. All study participants provided written informed consent before being enrolled.

All authors have no conflicts of interest to declare.

This research was financially supported by Merck Serono Co., Ltd., Beijing, China, an affiliate of Merck KGaA, Darmstadt, Germany. Merck Serono Co., Ltd., Beijing, China, was involved in the development of the study design and the collection, analysis, and interpretation of data; supported the development of the manuscript; and was involved in the decision to submit the manuscript for publication.

Ningning Pan: investigation, writing – original draft, and writing – review and editing. Xiumei Zhen, Yanhong Fan, and Jianhuai Zheng: investigation, resources, and supervision. Yuanyuan Wang: methodology and supervision. Qiao Liu and Xun Liao: data curation, resources, and supervision. Rui Yang: conceptualization, data curation, project administration, and writing – review and editing.

Medical Affairs, Merck Serono Co. Ltd, Beijing, China is an affiliate of Merck KGaA, Darmstadt, Germany.

The data that support the findings of this study are not publicly available because of information that could compromise the privacy of research participants but are available from the corresponding author upon reasonable request.