Endometrial compaction is associated with increased clinical and ongoing pregnancy rates in unstimulated natural cycle frozen embryo transfers: a prospective cohort study

Abstract

Purpose

To investigate the association between endometrial compaction and pregnancy rates in unstimulated natural cycle frozen embryo transfers.

Design

A single-center prospective cohort study. Endometrial thickness by transvaginal ultrasound and blood progesterone levels on the day of ovulation and the day of embryo transfer were evaluated in patients undergoing natural cycle frozen embryo transfer. Compaction was defined as > 5% decrease in endometrial thickness between ovulation day and day of transfer. Clinical and ongoing pregnancy rates in cycles with and without compaction were compared.

Results

Seventy-one women were included, of which 44% had endometrial compaction, with similar rates when subdividing the patients by day of transfer (day 3 or day 5). Clinical and ongoing pregnancy rates were higher in the compaction group compared to the non-compaction group (0.58 vs. 0.16, P < 0.001; 0.52 vs. 0.13, P < 0.001 respectively). Subdividing by degree of compaction > 10% and > 15% revealed similar pregnancy rates as > 5%, with no added benefit to higher degrees of compaction.

Conclusions

About half the patients in our study undergoing unstimulated natural cycle frozen embryo transfer experienced compaction of the endometrium, occurring as early as day 3 post-ovulation. This was significantly correlated with increased clinical and ongoing pregnancy rates.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10815-022-02544-7.

Background

Recent studies using chromosomal abnormality screening demonstrated ongoing pregnancy rates (OPR) of only 70-50% [20, 29] despite euploid embryo transfer, shifting the research focus from the embryo to its implantation environment [3, 4]. In attempts to personalize the window of implantation and optimize the chance for successful implantation and increased ongoing pregnancy and live birth rates, attention has been directed at endometrial receptivity and its management. Main research areas include genetic biomarkers and gene expression microarrays [1, 13, 25] and the endometrial microbiome [7, 9, 18, 19].

Despite these advances, ultrasound imaging still plays an important role in endometrial assessment, being a bedside, noninvasive, inexpensive modality. An endometrial width of < 8 mm in fresh cycles and < 7 mm in hormonally prepared frozen embryo transfer (FET) cycles has been associated with lower pregnancy and live birth rates [5, 15]. A recent addition to the sonographic endometrial assessment is the degree of endometrial compaction, defined as the change in endometrial thickness (EMT) between the end of the estrogen-only phase and the day of embryo transfer.

Studies measuring endometrial volume and thickness in normo-ovulatory women during spontaneous menstrual cycles demonstrate a rapid increase in volume during the follicular phase, reaching a plateau or a slight decrease around the time of ovulation, and remaining relatively stable throughout the luteal phase [14, 23]. These sonographic findings represent the endometrial physiologic changes. During the follicular phase, the endometrium is exposed to estrogen resulting in accelerated linear growth of endometrial glands and blood vessels with increased EMT and a typical trilaminar appearance. Rising progesterone levels following ovulation results in the cessation of endometrial proliferation, despite further growth of glands and blood vessels contributing to its density rather than volume [26]. Additional changes include the accumulation of glycogen in the gland lumens [6] and increased proliferation of T cells, macrophages, and lymphoid nodules [26]. These changes are manifested in the homogeneous hyperechoic appearance of the luteal phase endometrium [8]. The concept of endometrial compaction is based on these changes in endometrial density following progesterone exposure in the early luteal phase. The degree of compaction of the endometrium after progesterone exposure may indicate whether the endometrium is responsive to progesterone and could serve as a tool to evaluate endometrial receptivity.

Several studies have addressed the correlation between endometrial compaction and FET cycle outcomes, with contradictory results [18–23]. Two Canadian studies including 271 FET cycles in their first study and 225 single euploid FET cycles in the second study found a correlation between endometrial compaction and increased pregnancy rates in hormonal replacement therapy (HRT) cycles at all levels of compaction (5%, 10%, 15%, 20%) compared to cycles with no compaction [10, 30]. Several large Chinese studies including 2768–4465 patients, undergoing both stimulated NC and HRT cycles did not report a significant increase in pregnancy rates in correlation with endometrial compaction [12, 28]. All these studies are retrospective in nature, include HRT or modified natural cycle protocols, and use transabdominal ultrasound (TAUS) for the transfer day measurement as opposed to transvaginal ultrasound (TVUS) for the first measurement. To date, only one prospective study evaluating the correlation between endometrial compaction of > 5% compared to no compaction or expansion was performed, including HRT cycles only [24], with no significant difference in pregnancy rates or live birth rates.

In our prospective study, we compared the clinical pregnancy rate (CPR) and OPR in relation to the compaction degree on day 3 and day 5, in non-stimulated natural cycle frozen embryo transfers (NC-FET), using TVUS for both measurements, with the initiation of the luteal support only the following transfer. Given the slight natural decrease in endometrial volume observed in normo-ovulatory women [14, 23], we hypothesized that endometrial compaction will be associated with an increase in pregnancy rates in non-stimulated NC-FET.

Materials and methods

This is a single-center prospective observational study including frozen embryo transfer cycles based on an unstimulated natural cycle protocol. Patients were recruited at the IVF unit, Shamir medical center, Israel, between August 2019 and July 2021. The study was approved by the Institutional review board.

Reproductive age women ≤ 40 years assigned to a non-stimulated NC-FET were included. The NC-FET is the default protocol in our unit for normo-ovulatory patients. As per our standard unit protocol, women arrived for their first evaluation 2–3 days prior to their expected ovulation based on their regular menstrual cycle (usually days 10–12 of the menstrual cycle) with repeat visits until ovulation was diagnosed. An experienced senior physician performed TVUS evaluating EMT, endometrial pattern (trilaminar versus iso/hyper-echoic) (Fig. 1A), and diameter of the leading follicle. Measurements were performed using Voluson P6. EMT was measured at the maximal distance from the anterior to the posterior stratum basalis-myometrial junction in the mid-sagittal plane [17]. All physicians involved underwent specific training for the purpose of the study. The blood was drawn for LH, estradiol, and progesterone levels. The day of ovulation was determined by the elevation of LH > 35 IU. Embryo transfer was scheduled on either LH + 4 for day 3 embryos or LH + 6 for day 5 embryos [16]. The number of embryos transferred was individualized by the treating physician and was determined prior to recruitment. Patients included had their embryos vitrified between 2016 and 2021. Embryos were assessed by experienced embryologists according to accepted criteria as detailed elsewhere [10, 22].

Fig. 1.

A Pre-ovulatory trilaminar endometrium. B Post-ovulatory hyperechoic/homogeneous endometrium

On the day of embryo transfer patients underwent transvaginal ultrasound (Fig. 1B) and a blood draw for progesterone and estradiol levels. The physician performing the measurement was blinded to the pre-ovulatory EMT. All patients were prescribed the same luteal support with vaginal Endometrin 100 mg (Ferring) twice daily, initiated only following embryo transfer either on day 3 or day 5 as per unit protocol. BhCG levels were drawn 10–14 days following the transfer, followed by an US exam 2 weeks later in case of a positive result. Luteal support was continued for 9 weeks. CPR was defined as the presence of fetal cardiac activity on ultrasound. OPR was defined as a fetal cardiac activity beyond 12-week gestation.

Exclusion criteria included irregular menstrual cycles or menstrual cycles < 21 or > 35 days, BMI > 35, infertility diagnoses of repeated implantation failure (≥ 3 good quality embryo transfers), recurrent pregnancy loss (≥ 2 missed abortions), poor ovarian response (based on the Bologna criteria), severe endometriosis, tubal factor with hydrosalpinx, uterine factor, PGT-A/PGT-M, egg donation, and more than 2 prior transfers (since the last pregnancy). Following recruitment, patients were excluded if the EMT on day 0 or − 1 was < 7 mm or other than trilaminar, had endometrial fluid on the day of ET, or complained of vaginal bleeding prior to ET.

Compaction percentage was defined as the difference in EMT between ovulation day and transfer day, divided by the thickness on ovulation day. We grouped patients according to the compaction percentage. To avoid minor measurement variations, the compaction group was defined as endometrial compaction of more than 5%, and non-compaction was defined as < 5% (including those that compacted 0–5% and those that expanded). Compaction percentage was defined based on previous studies using similar values [2, 10, 24]. Our primary outcome was CPR (the presence of fetal cardiac activity on ultrasound per embryo transfer) by the endometrial compaction group. The secondary outcome was OPR (pregnancy > 12-week gestation per embryo transfer). We further subdivided the study group by different degrees of compaction (> 5%, > 10%, and > 15%) and compared CPR and OPR.

Data analysis

Comparative statistical analyses were performed using the chi-square test for categorical variables, ANOVA for continuous variables, and Z test for two proportions. Sensitivity and specificity levels were assessed via the ROC procedure. The significance threshold was set at p < 0.05. All statistical analyses were conducted on the IBM Statistical Package for the Social Sciences (IBM SPSS v.25; IBM Corp.).

Results

Ninety-four women were initially recruited, of them 71 (mean age 32.8) were included in the final data analysis. Four patients refused to participate on transfer day, 3 had a change of protocol to HRT-FET, 2 had a change in luteal support following transfer (Ovitrelle, Duphastone, IM progesterone), one patient had vaginal bleeding, one had an EMT of 5.5, 2 had a day 2 embryo transfer, 3 had a non-transferable embryo, and 7 did not participate due to technical reasons. Fifty-three patients underwent day 5 embryo transfer, and 18 patients underwent day 3 embryo transfer. Of 71 included participants, one patient had an extrauterine pregnancy. There were overall 25 clinical pregnancies (35.2%), with two ending in missed abortion, and a total of 23 ongoing pregnancies (32.4%). Mean ovulation EMT was 10.04 mm (± 2.2, range 7–15) and mean transfer EMT was 9.94 mm (± 2.35, range 6.1–16). Mean compaction was − 0.128% (± 19.16, range + 33.3 to − 48.75). Mean compaction rate of 18 cases with a day 3 transfer was 0.89 (± 16.43) and of 53 cases with a day 5 transfer − 0.47 (± 20.13) (p = 0.795).

A total of 33 participants (46%) had compaction > 5%. Of those, the mean compaction was 14.85% (± 7.67, range 5.5–33.3). When looking at participants by day of transfer, 44% (8/18) of patients undergoing embryo transfer on day 3 had compaction of > 5%, with a mean compaction of 14.68%, similar to those undergoing embryo transfer on day 5 (47%, 25/53, 14.9%) (Fig. 2).

Fig. 2.

Percentage of cycles with compaction (> 5%) by day of transfer

Comparing baseline characteristics between groups (compaction vs. no compaction), there was no difference in age at ovum pick up and transfer, BMI, gravity, parity, number of previous embryo transfers, basal FSH, and infertility diagnosis. Cycle characteristics were also similar including ovulation EMT, serum levels of progesterone on ovulation day and transfer day, estrogen/progesterone ratio, cycle day of ovulation, number of embryos transferred, day of embryo transfer, and embryo quality (Table 1). EMT on transfer day was lower in the compaction group (8.95 mm vs. 10.81 mm, p < 0.01).

Table 1.

Baseline and cycle characteristics with and without compaction

	Compaction (> 5%)	No compaction (< 5%)	P
N (cycles)	33 (46.5%)	38 (53.5%)
	Mean ± SD	Mean ± SD
Age at OPU (y)	31.80 ± 5.01	31.63 ± 5.03	.890
Age at transfer (y)	32.52 ± 4.86	33.03 ± 4.93	.665
BMI	24.40 ± 4.25	22.75 ± 5.80	.193
Gravidity	1.09 ± 1.09	1.61 ± 1.69	.144
Parity	0.63 ± 0.75	1.05 ± 1.01	.052
Prior ET	1.64 ± 0.70	1.76 ± 0.80	.508
Basal FSH (IU/mL)	6.40 ± 1.71	7.24 ± 1.92	.065
Peak follicular EMT (mm)	10.57 ± 2.57	9.58 ± 1.70	.058
EMT at transfer (mm)	8.95 ± 2.13	10.81 ± 2.22	.001
Peak follicular progesterone (nmol/L)	3.81 ± 2.53	3.77 ± 2.24	.941
Progesterone at transfer (nmol/L) Estradiol at Ovulation (pmol/L) Estradiol at transfer (pmol/L)	47.71 ± 25.19 518.45 ± 285.83 528.29 ± 238.50	44.30 ± 19.43 506.87 ± 257.09 558.39 ± 216.60	.545 .886 .598
Estradiol/progesterone at transfer Cycle day of ovulation	14.27 ± 10.44 16.67 ± 4.46	17.16 ± 14.33 15.84 ± 3.83	.362 .405
	N (%)	N (%)
Infertility diagnosis   Male factor   Unexplained   Tubal   Fertility preservation (non-medical)   Endometriosis—mild   Age	17 (51%) 12 (36%) 3 (10%) 0 1 (3%) 0	20 (53%) 8 (21%) 7 (18%) 1 (3%) 0 2 (5%)	.276
Embryo quality   Moderate   High	12 (36%) 21 (64%)	16 (42%) 22 (58%)	.621
No. of embryos transferred   1   2	28 (85%) 5 (15%)	31 (82%) 7 (18%)	.714
Day of transfer   Day 3   Day 5	8 10	10 28	.841
	Rate	Rate
Clinical pregnancy	0.58	0.16	< .001
Ongoing pregnancy	0.52	0.13	< .001

EMT, endometrial thickness; OPU, ovum pick up; ET, embryo transfer

CPR and OPR were significantly higher in the compaction group compared to the non-compaction group (0.58 vs. 0.16, P < 0.001; 0.52 vs. 0.13, P < 0.001 respectively) (Table 1). We redefined our groups at varying levels of compaction (10%, 15%) (Table 2). CPR was higher at all levels of compaction cutoffs. OPR was higher at > 10% compaction compared to < 10% but did not reach statistical significance at the 15% cutoff. Of note, only 13 patients had endometrial compaction > 15%. Comparison of outcomes in patients with 5–10% compaction (n = 11, mean compaction 7.32 mm) and above 10% compaction (n = 22, mean compaction 18.61%) revealed no difference in CPR (63% vs. 54%, p = 0.719) or OPR (63% vs. 50%, p = 0.712). ROC analysis of the probability to achieve clinical pregnancy as a function of the compaction rate was performed with a calculated area under the curve of 0.716 (Fig. 3).

Table 2.

Clinical and ongoing pregnancy rates at different compaction cutoffs

Compaction	Pregnancy	p-value 2-proportion z	Sensitivity	Specificity
Clinical
> 5% (n = 33)	19 (57.6%)	p < .001	.760	.696
< 5% (n = 38)	6 (15.8%)
> 10% (n = 22)	12 (54.5%)	p = .022	.480	.783
< 10% (n = 49)	13 (26.5%)
> 15% (n = 13)	8 (61.5%)	p = .028	.320	.891
< 15% (n = 58)	17 (29.3%)
Ongoing
> 5% (n = 33)	18 (54.5%)	p < .001	.773	.673
< 5% (n = 38)	5 (13.2%)
> 10% (n = 22)	11 (50.0%)	p = .034	.445	.755
< 10% (n = 49)	12 (24.5%)
> 15% (n = 13)	7 (53.8%)	p = .068	.273	.857
< 15% (n = 58)	16 (27.6%)

Fig. 3.

ROC analysis of the probability of clinical pregnancy as a function of compaction percentage. Vertical lines from left to right represent 15%, 10%, and 5% compaction. Area under the curve = 0.716

A comparison of baseline and cycle characteristics between patients with ongoing pregnancy and patients who did not become pregnant revealed no differences other than the degree of endometrial compaction (8.44% vs − 4.24%, p = 0.008) (supplementary table 1).

Discussion

We demonstrated a significant increase in pregnancy rates with an endometrial compaction of 5% and above between day of ovulation and embryo transfer in unstimulated natural FET cycles.

Our study followed natural cycles of normo-ovulatory patients with a pre-ovulatory EMT of at least 7 mm. In line with the abovementioned studies [14, 23], about half the participants demonstrated a decrease in EMT around ovulation. This was associated with a significant increase in implantation rates and OPR. No association was found between pre-ovulation EMT or transfer EMT and compaction rates or ongoing pregnancy rates. This might indicate that the compaction percentage, and not the absolute EMT, contributes to implantation and pregnancy. As only cycles with 7 mm and above were included, this finding is not surprising. We found similar compaction rates on day 3 and day 5, suggesting that when the endometrium does compact, it happens relatively early, around day 3 and possibly earlier. This suggests that clinicians may use the compaction rate to aid decision-making regardless of the day of transfer. The day of a clinically efficient compaction should be further explored. Prospective studies with measurements on both day 3 and day 5 of the same patient may provide more information on whether decision-making can be done as early as day 3, regarding a day 5 transfer.

We speculate, in accordance with previous studies [3, 10, 30], that non-compaction of the endometrium may reflect an inadequate progesterone effect. Of note, the degree of compaction or OPR was unrelated to the progesterone levels measured on the day of ovulation or day of transfer. This is in line with previous reports showing that secretory endometrial development is unrelated to circulatory progesterone levels [27]. We did not administer any luteal support until the day of transfer, thus avoiding any exogenous effect on circulating levels. It has previously been proposed that an inadequate estrogen to progesterone ratio results in the continuous growth of the endometrium under the influence of too much estrogen [10]. Not surprisingly, as our study only included natural cycle protocols, estrogen levels did not reach supraphysiologic levels and estrogen-progesterone ratios did not differ between patients with compacted or non-compacted endometrium. Another possible explanation for the lack of endometrial compaction may be the presence of progesterone receptor deficiency or resistance in the endometrium of some infertile women [10]. Possible etiologies include overexpression of BCL-6 and SIRT-1 [3], progesterone receptor gene polymorphisms, altered microRNA expression, epigenetic modifications to progesterone receptors, and chronic endometrial inflammation [11, 21]. Future studies should focus on genetics and microbiome studies in patients whose endometrium does not compact or continues to expand following ovulation.

Previous studies have investigated the correlation between endometrial compaction and OPR and live birth rates. The Canadian studies support our study results of increased OPR with increased endometrial compaction. The study by Hass et al. in patients that underwent HRT-FET cycles compared the pregnancy rates in patients whose endometrium compacted after starting progesterone with patients that had no change or increased EMT. Significantly increased OPR was demonstrated at all levels of compaction (45.2% vs. 23/1%, 51.8% vs. 23.9%, 58.9% vs. 25.6% respectively) [10]. A follow-up study of the same group included euploid embryos only, reaching similar results with OPR after compaction of > 15% of 51.5%, compared with 30.2% in cycles where the endometrial lining did not compact [30]. The authors concluded that these findings may partially explain why some euploid embryos fail to implant. These studies only included HRT-FET cycles and were retrospective in nature, comparing re-measurements from previously recorded images. The transfer measurement was transabdominal as opposed to the first transvaginal approach making the comparison suboptimal.

In contrast, several large retrospective cohort studies from China (2768–4465 patients) including HRT and stimulated natural FET cycles found no significant effect of endometrial compaction on CPR and LBR. They calculated the compaction rate based on recorded measurement [2, 12, 28]. Despite the strength of the large cohorts of these studies, they relied on the retrospective review of EMT values. Endometrial measurements were not performed as part of the study, with no documented images to remeasure and confirm the values recorded. Furthermore, differences in endometrial preparation protocols and the mode and timing of the post-progesterone ultrasound assessment may have contributed to the conflicting results in these studies. A recent prospective study, using TVUS for both measurements, including 259 single euploid HRT-FETs found a relatively low rate of compaction (16.6%), with no difference in CPR, LBR, and spontaneous abortion rate in patients with a compacted vs. non-compacted endometrium [24]. This was the first published prospective study but is not comparable to our study in the endometrial preparation method and had several limitations including a very small proportion of cycles undergoing compaction and lack of infertility diagnosis, possibly including patients with a diagnosis affecting implantation rates.

The main strength of our study is its prospective nature and the inclusion of a homogenous population of normo-ovulatory patients undergoing non-stimulated natural cycles. Patients with potential infertility diagnosis affecting implantation and pregnancy rates were excluded. Experienced senior physicians, undergoing specific training for the purpose of the study, measured the endometrial width around ovulation and on the day of transfer using transvaginal ultrasound for both measurements, being blinded to the first result during the second measurement. Moreover, 5% compaction was used in previous studies evaluating compaction. Variability in measurement, particularly when investigating differences of sub-millimeters, is unavoidable, but as this was a prospective study, special attention was given to the accuracy of the endometrial measurement, making our results more accurate and reliable for interpretation. Our study has several limitations, mainly our limited sample size. Another limitation is that PGT-A is not performed routinely at our institution, especially in young patients with good prognosis, like those included in our study; thus, PGT-A cycles were not included, and embryonic factor could not be excluded. Nevertheless, there was no difference in embryo quality or transfer day between cycles with and without compaction, minimizing the embryonic influence on PR. Finally, our study included NC-FET only; thus, our results are not generalizable to other embryo transfer protocols.

In conclusion, about half the patients in our study undergoing unstimulated NC-FET experienced compaction of the endometrium following ovulation, occurring as early as day 3 post-ovulation. This was significantly correlated with increased CPR and OPR. We suggest that in selected cases, such as repeated implantation failure or the last available embryo for transfer, the compaction degree may serve as a tool in decision-making about further workup (chronic endometritis workup and treatment), postponement of transfer to allow optimization of transfer conditions or change of protocols. Larger prospective studies and meta-analyses are needed to validate these results in NC and other ET protocols.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

A. H., M. Y., and S. A. designed the study. All authors contributed to data collection, M. Y., A. H., and S.A. drafted the first version of the manuscript. M. Y., A. H., and S. A. contributed to data analysis and interpretation. All authors revised the manuscript and approved the final submitted version.

Declarations

Conflict of interest

The authors declare no competing interests.