Prospective analysis of progesterone exposure in programmed single thawed euploid embryo transfer cycles and outcomes

Abstract

Purpose

In the era of personalized medicine and the increased use of frozen embryo transfer (FET), assay of the endometrium’s receptivity prior to transfer has gained popularity, especially among patients. However, the optimal timing for single thawed euploid embryo transfers (STEET) in a programmed FET has yet to be determined Mackens et al. (Hum Reprod. 32(11):2234–42, 2017). We sought to examine the outcomes of euploid FETs by length of progesterone (P4) exposure.

Methods

Prospective cohort study of programmed FETs of single euploid embryos between June 1, 2018, and December, 18, 2018, at our center. Subjects reported the exact start time for initiating progesterone. The transfer time was noted to calculate the primary independent variable, duration of progesterone exposure. Statistical analysis included ANOVA and Spearman’s rho correlation, with p < 0.05 considered significant.

Results

Inclusion criteria were met for 253 programmed STEET cycles in the analysis. There was no significant difference in P4 duration when comparing outcome groups (112.8 ± 3.1 ongoing pregnancy (OP), 112.4 ± 4.4 spontaneous abortion (SAB), 111.6 ± 1.7 biochemical pregnancy (BP), 113.9 ± 5.7 no pregnancy (NP), F 1.76, df 3, p = 0.16). An ROC curve assessing the ability of P4 duration to predict ongoing pregnancy (OP) had an area under the curve of 0.467 (p = 0.38).

Conclusion

Duration of P4 was not associated with outcome. Of the cycles, 65.6% resulted in ongoing pregnancy with our center’s instructions resulting in an average progesterone exposure of 112.8 h, with a range of 98.3–123.7 h. With growing popularity for individualized testing, these results provide evidence for patient counseling of the high likelihood of ongoing pregnancy without personalized testing.

Introduction

Endometrial receptivity is based upon the temporal window of implantation (WOI) when the endometrial epithelium is most prepared for blastocyst implantation [1, 2]. With the continued advancement of assisted reproductive technology (ART), studies have shown that the timing of progesterone exposure may impact the window of implantation and perhaps affect the endometrium [3]. However, the results of these studies vary, with data to support that both short [4] and long [5] durations of progesterone (P4) exposure are optimal to achieve the best WOI and implantation rates. Moreover, there is evidence that luteal phase support alone can also improve implantation [4, 6]. Regardless of P4 dosage and administration, the timing of progesterone should synchronize the endometrium and the embryo as a dyad, with, for example, blastocysts transferred 6 days following progesterone supplementation [4]. However, the exact duration in hours can vary substantially between institutions and even between providers. Therefore, the optimal timing for embryo transfer, and more specifically single thawed euploid embryo transfers (STEET), has yet to be determined [7].

Recently, the use of assays, such as microarray, has demonstrated the feasibility of a molecular classification of the endometrium and its receptivity [1]. This technology has shown that the genetic profile of the endometrium with respect to factors such as hormonal exposure during the luteal phase may differ from patient to patient. In the era of personalized medicine along with the simultaneously increasing use of frozen embryo transfer (FET), assays of the endometrium’s receptivity prior to transfer have gained popularity, especially among patients. There is data to support that assays of endometrial receptivity could be useful to determine a personalized WOI [8, 9] and therefore help doctors personalize duration of progesterone.

Physiologically, the WOI is considered a narrow window limited by time. Therefore, it is important to determine if progesterone exposure time impacts embryo transfer outcomes. As such, the objective of our study is to examine the outcomes of single euploid in programmed or hormone-replaced FETs by length of progesterone (P4) exposure at our clinic.

Materials and methods

We performed a prospective cohort study under IRB approval (NYU IRB #13-00389). All patients undergoing FET between June 1, 2018, and December 18, 2018, at the NYU Langone Fertility Center (NYULFC) were reviewed. At our center, outcomes are assessed at two specific times per year so that data is reported as part of “Series 1” or “Series 2.” The time points for this study were chosen to align with the end of Series 2 of 2018 for consistency.

Only programmed FET cycles with transfer of a single euploid embryo were included in this analysis. When undergoing IVF at our center, all patients are counseled on the risks and benefits of PGT-A and given the option to elect PGT. Counseling includes the potential PGT-A has for gained knowledge of future reproductive potential and possible stratification for embryo selection in future embryo transfers. Further discussion of the limitations includes cost and the possibility of no euploid embryos or a no result diagnosis as well as the low but not zero error rate. In order to control for the health and age of the embryo, only embryos which underwent PGT and yielded euploid diagnoses were included. These embryos all underwent vitrification using standard procedure and Irvine kit. All embryos are graded and the best quality embryo is chosen, unless another clinical factor (such as embryo sex) is requested. Programmed FET cycles were defined as treatment of daily oral estradiol followed by either 50–75-mg intramuscular P4 in oil or vaginal P4 suppository. Prior to starting progesterone support, transvaginal ultrasound monitoring was used to confirm ovarian suppression and progesterone levels were checked. The endometrium is monitored throughout the cycle by transvaginal ultrasound to assess for thickness and morphology. Once the endometrial lining is > 7 mm and trilaminar without fluid, ovaries suppressed, and serum estradiol > 150 pg/mL, the patient is informed to start progesterone and the transfer is scheduled for 6 days later. Patients are given a window of time to start their progesterone as per clinic protocol. Embryo transfers are scheduled in 15-min intervals between 11:00 am and 2:00 pm on the 6th day of progesterone administration. Thus, the duration of progesterone exposure with standard instructions varies per patient. Occasionally, patients start their progesterone outside the window of time instructed; both accidentally or per physician request, further leading to a variability in progesterone duration. For this study, subjects were asked to record the exact time of their first progesterone administration. The exact transfer time was noted to calculate the primary independent variable, the duration of progesterone exposure.

Exclusion criteria comprised programmed FETs with untested, mosaic, or double embryo transfers, as well as natural cycle FETs. Embryos in this study were tested by either array comparative genomic hybridization (aCGH) or next-generation sequencing (NGS). Furthermore, cycles where exact timing data could not be determined were excluded. Cycles that were adjusted by endometrial receptivity assays were noted and excluded, as these individuals had used a prior mock trial to perform an assay of the endometrium in order to adjust for optimal timing of transfer.

As mentioned, the primary independent variable was length of progesterone exposure, which was determined by the exact time of P4 initiation until the time embryo transfer was performed, prior to embryo transfer result. The primary outcome was pregnancy outcome. Statistical analysis included ANOVA and Spearman’s rho correlation where appropriate, with p < 0.05 considered significant.

Results

A total of 503 frozen, programmed embryo transfer cycles were reviewed. Of those, 253 programmed STEET cycles met criteria and were included in the analysis. Two hundred fifty were excluded for the following reason: 57 were untested embryos, 18 were mosaic embryos, 26 were excluded due to multiple embryo transfer, and 149 were excluded as timing of progesterone was not available. The majority of patients (248) were treated with the IM progesterone regimen; however, 2 patients had a combination of IM and vaginal and 3 patients had vaginal progesterone only. The average patient age at time of ET was 38.0 ± 4.5 years (Table 1). Three out of the 253 cycles utilized an assay for endometrial receptivity for adjusted timing. The mean duration of P4 exposure was 112.8 ± 2.9 h with a range from 98.25 to 124.5 h for all unadjusted cycles.

Table 1.

Patient demographics

Characteristics	Median	Range
Age (years) at transfer	38	29–54
Age (years) at retrieval	37	28–46
Endometrial thickness (mm)	8.5	6.2–13

Overall, 166 women had an ongoing pregnancy (OP), 25 had a spontaneous loss (SAB), 12 a biochemical pregnancy (BP), and 50 a negative pregnancy (NP) test, for a 65.6% ongoing pregnancy rate. When comparing outcome groups, there was no significant difference in P4 duration (112.8 ± 3.1 OP, 112.4 ± 4.4 SAB, 111.6 ± 1.7 BP, 113.9 ± 5.7 NP, F 1.76, df 3, p = 0.156, Fig. 1). An ROC curve was performed and showed that the duration of P4 did not predict ongoing pregnancy (OP), with an area under the curve of 0.467 (p = 0.38) (see Fig. 2). Furthermore, as shown in Fig. 3, there was no correlation in the day 28 serum hCG value based on the duration of P4 exposure (rs = 0.058, p = 0.399).

Fig. 1.

Progesterone (ng/mL) duration by outcome

Fig. 2.

Duration of progesterone exposure (hours) vs. cycle day 28 serum hCG

Fig. 3.

ROC curve assessing the ability of P4 duration to predict ongoing pregnancy (OP)

Discussion

In our study, the duration of P4 did not predict the outcome of embryo transfer. With our center’s standard instructions, patients were exposed to progesterone over a range of 98.3–123.7 h prior to transfer and 65.6% of patients who underwent a euploid embryo transfer had an ongoing pregnancy, despite the variability in progesterone exposure. Transfers that resulted in negative pregnancy tests as well as biochemical and spontaneous pregnancies did not have significantly different progesterone exposures. The results of our study support that the ongoing pregnancy rate after embryo transfer may not be affected by utilizing standard instructions for all-comers to frozen embryo transfer. Moreover, we recognize that our center’s instructions may be different from those of other centers with similarly excellent clinical pregnancy rates which provides further support that personalization may be unnecessary for the majority of patients.

Similarly, the length of progesterone exposure did not correlate with serum hCG level on cycle day 28. This illustrates that the length of progesterone may not impact receptivity as robustly as hypothesized. This is consistent with prior studies which have suggested that changing the timing of progesterone supplementation by a day did not impact pregnancy rates in fresh donor egg cycles [10]. Our present study continues to suggest that the personalization of frozen embryo transfer protocol with regard to progesterone supplementation may not be necessary for everyone. Large studies on the average patient population undergoing FET have shown that receptivity testing in a mock cycle prior to FET does not improve the pregnancy outcome [11]. In fact, this testing adds to treatment cost and creates a delay in treatment which may be unnecessary for the majority of patients.

Our study suggests that personalized progesterone exposure is not necessary to achieve a high ongoing pregnancy rate. There may be a subset of patients who benefit from this approach, but this is largely unnecessary in patients undergoing their first frozen embryo transfer cycle. Further studies are needed to understand exactly which patient populations may be best served by a personalized progesterone exposure protocol. Patients with a history of recurrent implantation failure may benefit from personalization of FET [3, 12, 13] or history of early loss or biochemical pregnancies; however, it may not be generalizable to the whole population pursuing embryo transfer. Therefore, we would recommend that patients start progesterone based on set clinic protocols, such as day 6 of progesterone, rather than taking an individualized approach.

The main strengths of this study are its prospective design and inclusion of only single euploid embryo transfers. The standard protocol at our clinic allows for variability in progesterone exposure and therefore allowed us to examine outcomes across a range of exposure times at a single institution. Our study has several limitations. First, the sample size in this study was small and the differences in progesterone duration were modest. Larger studies are needed to confirm our results but keeping consistency with the series timeline was important for consistency with our centers’ data evaluation timing. Second was inclusion of cycles utilizing both vaginal and injectable progesterone. Several studies have evaluated potential differences in clinical outcomes by progesterone modality, with data showing that the modalities are equivalent [14] and data showing that one modality is superior [15, 16] and we did not control for this potential confounder. Third, there is increasing utilization of a “natural” or non-hormone-replaced frozen embryo transfer cycle which was not represented in this study. There may be differences between these protocols with respect to personalization that are not addressed here. Additionally, our patient population is diverse, which could be a potential confounder. Lastly, multiple providers did the transfers and while the differences between providers are small given similar pregnancy rates, these small effects were also not controlled for.

In the era of personalized medicine, there is encouragement from all angles to provide effective and precise treatments. This pressure can be felt by research which bridges the gap from conventional medicine to a more patient-personalized approach [17]. The availability of receptivity testing to adjust the timing of FET may lead some to believe that timing of FET must be personalized. Further investigation, perhaps with non-selection design, would be helpful to delineate the efficacy and need for personalized receptivity testing prior to first transfer.

Conclusion

Our study shows that personalized FET based on progesterone timing is not necessary for those undergoing programmed STEET at our institution. While receptivity testing may be helpful in the appropriate patient population, the window of implantation in the general STEET population may not be as narrow as previously suspected.

Author contribution

All authors contributed to study conception and design. Data collection and analysis were performed by Carly Hirschberg and Jennifer Blakemore. The first draft of the manuscript was written by Carly Hirschberg and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Declarations

Ethics approval and consent to participate

All procedures performed in this study were in accordance with the ethical standards of the institution (NYU School of Medicine Institutional Review Board, #s18-00698). Informed consent was obtained from all individual participants included in the study.

Conflict of interest

The authors declare no conflict of interest.