Endometrial echo patterns of embryo transfer day affect pregnancy outcomes in frozen embryo transfer cycles: a retrospective clinical study

Abstract

Background

The echo pattern constitutes an ultrasonic index that reflects the characteristics of the endometrium across various phases of the menstrual cycle. However, consensus of endometrial echo pattern and pregnancy outcomes is lacking in clinical application.

Methods

The retrospective cohort study analyzed the data from the electronic records of women who underwent frozen embryo transfer (FET) with hormone replacement treatment (not only one cycle per patient) between July 2020 to August 2021 at Reproduction Medicine Center of Jbnminling Hospital of Medical College of Nanjing University. A total of 138 cleavage stage embryo transfer cycles and 561 blastocyst transfer cycles were analyzed in this study. Transvaginal ultrasound scans were performed on the embryo transfer day. The endometrial echo pattern was classified into four types: A, B, B-C and C. Pattern A was defined as trilinear type, Pattern B, B-C and C were defined as non-trilinear type. All scans were conducted by experienced clinicians, and the images were reviewed by the same two physicians. The outcomes measured included embryo implantation rate, clinical pregnancy rate, first trimester abortion rate, and live birth rate. For the data that conforms to a normal distribution, two independent sample t-tests are used for comparison between two groups, and one-way analysis of variance (ANOVA) is used for comparison among multiple groups. For the data that does not conform to a normal distribution, rank sum tests (Kruskal-Wallis test) are used for inter-group comparisons. Count data are analyzed using chi-square tests. Logistic regression models were applied for the analysis of clinical outcome.

Results

The embryo implantation rate (p = 0.066), clinical pregnancy rate (p = 0.140), early abortion rate (p = 0.515) and live birth rate (p = 0.578) were similar between the 4 type of endometrial pattern groups in cleavage-stage embryo FET cycles. In blastocyst cycles, the implantation rate (p = 0.201) and clinical pregnancy rate (p = 0.555) did not differ between the four endometrial patterns. Patients with a Pattern A endometrium on blastocyst transfer day experienced a decreased live birth rate (19.05%) compared with Pattern B, Pattern B-C and Pattern C (p = 0.006. p = 0.008, p = 0.031 for Pattern A vs. Pattern B, Pattern A vs. Pattern B-C, Pattern A vs. Pattern C). The first trimester abortion rate of Pattern A is up to 40.00%, although there was no statistical difference (p = 0.118). In the cycles of non-trilinear type group, the early miscarriage rate (0.248 [95% CI, 0.067–0.914]; p = 0.036) was lower and the live birth rate (0.269 [95% CI, 0.089–0.810]; p = 0.020) was higher than trilinear type group.

Conclusions

Our retrospective study suggests that a trilinear pattern endometrium on blastocyst transfer was associated with a higher first trimester abortion rate and lower live birth rate.

Introduction

The success of IVF-ET (In vitro fertilization and embryo transfer) relies on a high-quality embryo, a receptive endometrium, and a healthy maternal condition. The endometrium, undergoing a series of structural and biochemical changes during the reproductive cycle, must be in a receptive phase [1, 2]. For decades, reproductive scientists have worked diligently to identify a suitable method for assessing endometrial receptivity. Currently, transvaginal ultrasonography is the most widely used in clinical practice, which provides clinicians with a non-invasive approach for monitoring endometrial development throughout the menstrual cycle.

Endometrial thickness and echo pattern are the most commonly used markers for evaluating endometrial receptivity in transvaginal ultrasonography examine. Numerous studies suggest that endometrial thickness is an independent predictor of pregnancy occurrence [3, 4]. Insufficient endometrial thickness, diagnosed as thin endometrium with a maximum thickness ≤ 7 mm on an ultrasound scan accompanied by a normal uterine cavity, is closely associatedwith pregnancy failure [5, 6]. However, due to the absence of high-quality clinical evidence and guidelines, there remains controversy regarding the application of endometrial echo patterns in clinical practice.

Endometrial echo patterns observed sonographically are categorized as triple line (TL), appearing multilayered with hyperechogenic outer walls and a distinct central echogenic line; isoechogenic (IE), where the endometrium exhibits the same echogenicity as the myometrium with a poorly defined central echogenic line; and homogeneous hyperechogenic (HH), presenting as an entirely echodense endometrium, more echogenic than the myometrium without a visible central echogenic line [7]. The sonographic appearance of the endometrium changes throughout the menstrual cycle [8]. In the proliferative phase, the endometrium has a hypoechoic texture with a well-defined central echogenic line. This texture changes in the secretory phase, becoming hyperechogenic with no visualization of the central echogenic line. Currently, it is widely believed that echo patterns primarily correspond to different stages of the endometrium, there is uncertainty regarding their association with pregnancy outcomes.

Previous research predominantly focused on the endometrial morphology during the late proliferative phase. Some studies reported that endometrial pattern on the day of hCG injection or progesterone initiation was not associated with pregnancy outcomes in assisted reproductive treatment [9,10,11,12,13,14], while some studies reported higher pregnancy rates among women with trilinear endometrial pattern, compared to those with semitrilinear or unilinear endometrial patterns [15,16,17]. Within embryo transfer cycles, clinicians meticulously monitor endometrial thickness and morphology during the proliferative phase. Upon attainment of the desired endometrial standard, progesterone induction is initiated, followed by the determination of the appropriate time for embryo transfer. Unless there are special circumstances, most clinicians do not review the endometrial echo on the day of embryo transfer. However, in clinical practice, we observed that a small percentage of patients exhibited a triple line (TL) endometrial pattern on the day of embryo transfer, which is typically supposed to be isoechogenic (IE) or homogeneously hyperechogenic (HH) pattern during secretory phase.

Only a small fraction of previous studies have documented the correlation between endometrial morphology during the secretory phase and the outcomes of clinical pregnancies. J H Check et al. observed decreased pregnancy and implantation rates in cycles where the homogeneous hyperechogenic pattern did not occur three days after transfer [18, 19]. A non homogeneous endometrial echo pattern in the midluteal phase was identified as a potential factor in unexplained infertility [8]. Conversely, several studies have recorded no significant differences in the endometrial echo pattern during the secretory phase between conception and non-conception cycles in IVF-ET cycles [11, 20]. Whether the morphology of the endometrium during the secretory phase affects clinical pregnancy outcomes and whether an abnormal endometrial echo pattern on the day of embryo transfer can predict adverse clinical pregnancy outcomes remain inconclusive. Current studies are lacking in strict quality control and uniformity, particularly regarding ultrasound measurement timing and endometrial preparation strategies, indicating a need for more in-depth research.

In this manuscript, we conducted a retrospective analysis of 699 FET cycles at our reproductive medicine center. We selected FET cycles using an exogenous estrogen and progesterone protocol to minimize the potential impact of endogenous hormones, ovulation factors, and other variables on the endometrium in each cycle. The study’s objective was to thoroughly assess the relationship between the sonographic endometrial pattern on embryo transfer day and pregnancy outcomes in women undergoing hormonally prepared FET cycles, with the goal of enhancing clinical pregnancy outcomes and optimizing clinical diagnosis and treatment.

Methods and materials

Patients and study design

We conducted a retrospective cohort study, collecting data from the electronic records of women who underwent FET cycles (not only one cycle per patient) via hormone replacement treatment with or without GnRHa from July 2020 to August 2021 at the Reproduction Medicine Center of Jinling Hospital, Medical College of Nanjing University (Nanjing, China). The exclusion criteria were: (a) age over 44 years; (b) coexisting hydrosalpinx, endometrial lesions, or an abnormal uterine environment (uterine fibroids protruding into the uterine cavity, submucosal fibroids, uterine adhesions, thin endometrium, etc.); (c) uterine anomalies; (d) moderate or severe endometriosis or adenomyosis; (e) Recurrent spontaneous abortion or repeated planting implantation failure (failure to achieve a clinical pregnancy after transfer of at least four good-quality embryos in a minimum of three fresh or frozen cycles in a woman under the age of 40 years). Patients underwent no therapeutic interventions apart from routine procedures and received standardized medical treatment in the hospital. The embryos at the cleavage stage were evaluated according to the Istanbul consensus [21]. Cleavage-stage (D3) high-quality embryos were defined as having 6–10 cells, with fragments < 30% and no severe asymmetry. Blastocysts were assessed on D5-D7 based on the criteria of Gardner [22]. Blastocysts with a score higher than 3BB were categorized as high-quality blastocysts.

Transvaginal ultrasound scans (TVUS) were performed on the day of cleavage-stage embryo or blastocyst transfer. All scans were conducted by experienced clinicians, and the images were printed and reviewed by the same two physicians. The endometrial echo pattern was classified into three types [23, 24]: Pattern A, Pattern B, and Pattern C [8, 25, 26]. Pattern A represented a multilayered “triple-line” endometrium, consisting of a central hyperechogenic line surrounded by two hypoechoic layers; Pattern B was characterized as an isoechogenic pattern relative to the surrounding myometrium with an unclear central hyperechogenic line; Pattern C displayed a homogeneous, hyperechogenic endometrium with increased reflectivity compared to the myometrium. An intermediate pattern B-C featured a gradual increase in reflectivity from the peripheral region to the central line. Endometrium of Pattern A was defined as trilinear type, and endometrium of Pattern B, B-C and C were defined as not trilinear type. Throughout the entire assisted reproductive technology cycle, we consistently applied uniform criteria for the classification of endometrial echo patterns. Images of the four endometrial patterns are presented in Supplementary Fig. 1.

The outcomes measured included embryo implantation rate, biochemical pregnancy rate, clinical pregnancy rate, first trimester abortion rate, and live birth rate. Pregnancy outcomes were followed up by phone and recorded in our electronic medical record. All data were collected using an exclusive internal database at our reproductive medicine center, ensuring patient data security through advanced threat prevention and periodic password renewals for user access. The methodological pipeline is illustrated in Fig. 1.

Fig. 1

A flow chart of this study

Endometrial preparation and embryo transfer

In the early stages of menstruation (between the second and fourth day of the menstrual cycle), patients underwent a sex hormone serum test and transvaginal ultrasonography examination. In hormone replaced treatment (HRT) cycles, patients without abnormalities commenced oral intake of a fixed dose of exogenous estradiol for 10–16 days (Femoston, 4 mg b.i.d.). When the patients took the full course of estrogen treatment and returned to hospital, we firstly conducted the transvaginal ultrasonography examination. Patients with low endometrial thickness will receive additional vaginal medication (Femoston, 2 mg estradiol q.d.). Upon achieving the requisite endometrial thickness (≥ 8 mm) and ruling out abnormal endometrial echo, superior follicle development or other abnormal conditions, serum E2 and P levels were required to be examined. The endometrial thickness of every patient in this study reach the cut-off of 8 mm before introducing progesterone. Once the serum E2 and P showed no abnormalities, oral estradiol was combined with dydrogesterone compound tablets (Femoston, 4 mg estradiol and 20 mg dydrogesterone b.i.d.), supplemented with a vaginal progesterone soft capsule (Utrogestan, 400 mg b.i.d.) to induce endometrial transformation. If GnRHa-HRT protocol was applied, 3.75 mg of triptorelin acetate (Ipsen Pharma Biotech) was administered between the second and fourth day of the menstrual cycle. After 28–35 days of down-regulation, those without abnormalities conducted the HRT cycles which have been described above. The embryos were freezed at afternoon in oocyte retrieval (OR) cycles and were thawed before 8:00 on transfer day. Cleavage-stage embryos were thawed and transferred on the 5th day of progesterone supplementation and blastocysts were thawed and transferred on the 7th day of progesterone supplementation. All embryos are frozen and thawed following the standard procedures of the embryology laboratory. We have supplemented this part of the content in the manuscript according to the Reviewer’s suggestion.

TVUS assessed the endometrial thickness and echo pattern, and serum E2 and P levels were measured on the embryo transfer day. For luteal support, patients typically continued Femoston (4 mg estradiol and 20 mg dydrogesterone, b.i.d.) and vaginal progesterone soft capsule (Utrogestan, 400 mg b.i.d.). Serum β-human chorionic gonadotropin (β-hCG) was detected 2 weeks post-embryo transfer to determine biochemical pregnancies. Transvaginal ultrasound was performed 4 weeks post-embryo transfer in patients with positive β-hCG results to confirm clinical pregnancies and the number of implanted embryos. A clinical pregnancy was identified by the presence of a gestational sac. The implantation rate was calculated as the number of gestational sacs by ultrasound observation divided by the number of transferred embryos. Luteal support in pregnant patients was sustained until 2 months post-embryo transfer. Follow-up was conducted to detect abnormalities during pregnancy. Early miscarriage was defined as spontaneous abortion occurring before 12 weeks of pregnancy. Live birth was defined as the delivery of a living newborn after the 28th gestational week, with live birth rate calculated as the ratio of live birth cycle number to the number of embryo transfer cycles.

Specimen collection

Endometrial samples of the fifth day post-ovulation were collected from the Reproduction Medicine Center of Jinling Hospital, Medical College of Nanjing University (Nanjing, China). The patients were infertile with diagnosis of the fallopian tube blockage, male factor or unexplained infertility, excluding endometrial disease such as endometrioma, adenomyosis, endometrial hyperplasia, endometrial polyps and intrauterine adhesion. TVUS was performed on the same day prior to the endometrial biopsy to examine the endometrial thickness and morphology. Immunohistochemistry of four endometrial samples were presented in this study, including two cases where ultrasound imaging showed Pattern A, and two cases displaying Pattern C. The four women were 25–34 years old and the body mass index (BMI) were 19.3–22.5 kg/m². The endometrium thickness was 9–13 mm. The study procedures were approved by the Clinical Ethics Review Committee of Nanjing Jinling Hospital.

Immunohistochemistry staining

The endometrial samples were fixed in neutral buffered formalin (NBF). Following deparaffinization and rehydration, the sections were subjected to antigen retrieval. Afterward, the sections were incubated with primary antibody of antibodies against Estrogen receptor α (ready-to-use, DAKO), Estrogen receptor β (5ug/ml, Abcam) or Progesterone receptor (ready-to-use, DAKO) overnight at 4 °C. Subsequently, the slides were incubated with secondary antibody at room temperature. Haematoxylin was used to counterstain the sections.

Statistical analysis

SPSS 26.0 (SPSS Inc., Chicago, IL, USA) software was used for statistical analysis. Chi-square testing was introduced for analysis, supported by Fisher’s exact test with Monte Carlo simulation for expected cell frequencies below five. Measurement data are assessed for normality using histograms. Data fitting the normal distribution are depicted as mean ± standard deviation and analyzed with one-way ANOVA to compare group differences. For non-normal distributions, data are shown as medians (interquartile ranges) and analyzed by the Kruskal-Wallis H test. Univariate analysis was used to preliminarily evaluate variables related to the clinical pregnancy outcome, and a multivariable logistic regression model was further employed to analyze the effect of endometrial echo patterns of embryo transfer day in HRT cycles on the clinical pregnancy outcome. Odds ratio (OR) and 95% confidence intervals (95% CI) were estimated using logistic regression. P < 0.05 was considered statistically significant.

Results

Characteristics of all FET cycles

This study involved 699 FET cycles, including 138 cleavage-stage embryo transfer cycles and 561 blastocyst transfer cycles. Transvaginal ultrasound scans were conducted on embryo transfer day, 1 to 3 h before transfer operation. As shown in Fig. 2, in cleavage-stage embryo cycles, 9 (6.50%) cycles were pattern A, 41 (29.7%) cycles were pattern B, 30 (21.7%) cycles were pattern B-C and 58 (42.0%) cycles were pattern C. In blastocyst cycles, 21 (3.74%) cycles were pattern A, 120 (21.39%) cycles were pattern B, 134 (23.89%) cycles were pattern B-C and 286 (50.98%) cycles were pattern C.

Fig. 2

The number and proportion of hormone replacement therapy FET cycles of 4 type of endometrial patterns on embryo transfer day

The demographic characteristics, the numbers and scores of embryos transferred were demonstrated in Tables 1 and 2. No significant differences were noted between the groups regarding the female age, the body mass index (BMI), infertility type, main infertility cause, infertility duration, the endometrial thickness on progesterone administration day, as well as the number, grade and fertilization method of transferred embryos both in cleavage-stage embryo and blastocyst transfer cycles. Cycles with Pattern C endometrium had a relatively thinner endometrial thickness on blastocyst transfer day compared with Pattern B group (10.37 ± 2.56 mm vs. 11.33 ± 2.84 mm; p = 0.006). Cycles with Pattern C endometrium had a higher serum estrogen level on the day of progesterone initiation compared with Pattern B-C group (1560.5 [1069.0-2838.0] pmol/L vs. 2067.0 [1168.0-5606.0] pmol/L; p = 0.037).

Table 1 Characteristics of FET cycles comparing endometrial patterns on cleavage-stage embryo transfer day

Table 2 Characteristics of FET cycles comparing endometrial patterns on blastocyst transfer day

Pregnancy outcomes

The embryo implantation rate (p = 0.066), biochemical pregnancy rate (p = 0.127), clinical pregnancy rate (p = 0.140), early miscarriage rate (p = 0.515) and live birth rate (p = 0.578) were similar between the 4 type of endometrial pattern groups in cleavage-stage embryo cycles (Table 3). In blastocyst cycles (Table 4), the implantation rate (p = 0.201), biochemical pregnancy rate (p = 0.734), clinical pregnancy rate (p = 0.555) did not differ between the four endometrial patterns. Patients with Pattern A endometrium on blastocyst transfer day experienced a decreased live birth rate (19.05%) compared with Pattern B, Pattern B-C and Pattern C (p = 0.006. p = 0.008, p = 0.031 for Pattern A vs. Pattern B, Pattern A vs. Pattern B-C, Pattern A vs. Pattern C). Specially, the first trimester miscarriage rate for Pattern A was 40.00%, markedly exceeding that of the general population [27,28,29,30], although there is no statistically significant difference.

Table 3 Pregnancy outcomes of FET cycles comparing endometrial patterns on cleavage-stage embryo transfer day

Table 4 Pregnancy outcomes of treatment FET cycles comparing endometrial patterns on blastocyst transfer day

Endometrial echo of Pattern A was defined as trilinear type and Pattern B, B-C, C were gathered as non-trilinear type. The pregnant outcomes of blastocyst cycles were demonstrated in Fig. 3. In univariate analysis for blastocyst cycles (Table 5), the non-trilinear group had a similar proportion of biochemical pregnancy (0.695 [95% CI, 0.290–0.414]; p = 0.414) and clinical pregnancy (0.640 [95% CI, 0.267–1.532]; p = 0.316) compared with trilinear type endometrium group. In the cycles of non-trilinear type group, the early miscarriage rate (0.248 [95% CI, 0.067–0.914]; p = 0.036) was lower and the live birth rate (0.269 [95% CI, 0.089–0.810]; p = 0.020) was higher than trilinear type group.

Fig. 3

The clinical outcomes of trilinear groups and non-trilinear groups in blastocyst transfer cycles

Table 5 Univariate analysis of pregnancy outcomes in blastocyst FET cycles

A univariate analysis of 540 blastocyst FET cycles was conducted to identify key factors influencing live birth rates (Supplementary Table 1). Consistent with previous studies, female age, BMI, and the number and grades of transferred embryos were associated with live birth rates. Meanwhile, endometrium with a non-triple line (0.269 [95% CI, 0.089–0.810]; p = 0.020) on embryo transfer day was a critical factor affecting live birth rates.

Potential factors for the endometrial morphological patterns on blastocyst transfer day

Univariate analysis showed that that female age, BMI, duration of infertility, infertility type, endometrial preparation protocol were not related to the endometrial morphological patterns. Higher serum progesterone level on the progesterone initiation day (1.506 [95% CI, 0.089–0.810]; p = 0.020) and lower serum estrogen level on the day of embryo transfer (0.999 [95% CI, 0.999–1.000]; p = 0.024) were related to the occurrence of triple line endometrium on blastocyst transfer day (Table 6).

Table 6 Univariate analysis of triple line endometrium on blastocyst transfer day in FET cycles

Four endometrial samples from the fifth day post-ovulation, in which two cases had ultrasound indicated an endometrium of Pattern A and the other two were Pattern C, were collected. Estrogen receptor and progesterone receptor were detected by immunohistochemistry staining (Supplementary Fig. 2). In one case of Pattern A, we observed an intriguing phenomenon where ERα was predominantly expressed in the cytoplasm rather than in the nucleus, indicating that an aberrant estrogen pathway might be present in the mid-luteal phase endometrium of this patient.

Discussion

Appropriate clinical interventions can optimize endometrial receptivity, achieving an ideal conception state and a successful pregnancy outcome. Endometrial receptivity evaluation involves various methods, including endometrial biopsy, magnetic resonance imaging (MRI), and ultrasound. An endometrial biopsy permits examination of the endometrium at the histological level [31], yet it is an invasive technique with a notable incidence of complications, leading to limited acceptance in clinical practice. MRI offers advantages in delineating endometrial thickness and its relation to the myometrium [32]; however, this method is costly and time-intensive. Ultrasound, prevalent in clinical practice due to its convenience, non-invasiveness, and affordability, is frequently utilized. Research indicates that endometrial thickness might serve as a proxy for endometrial receptivity [33]. Nevertheless, the utility of the endometrial echo pattern remains a topic of debate as a predictor for endometrial receptivity.

Endometrial echo patterns are identified by comparing the echogenicity of the endometrium with that of the myometrium and by noting a central echogenic line within the endometrium [24, 34, 35]. The echo pattern serves as an ultrasonic index reflecting endometrial proliferation and/or stromal decidualization. In 2019, Laurentiu Craciunas et al. conducted systematic meta-analyses to evaluate the evidence from observational studies regarding the use of endometrial receptivity markers as prognostic factors for pregnancy outcomes [1]. The study concluded that triple line patterns, assessed on the day of hCG injection in IUI, were linked to higher clinical pregnancy rates. However, the endometrial echo pattern on the hCG injection day/embryo transfer day in women undergoing IVF with fresh embryo transfer, or on the day of progesterone supplementation in FET, showed no significant impact on clinical outcomes. However, this conclusion was analyzed from studies with various ovarian stimulation protocols, different embryos transfer numbers and strategies [11, 20, 36, 37].

Regarding whether the morphology of the endometrium on embryo transfer day (middle luteal phase) impacts clinical outcomes, we posit that the fresh embryo transfer cycle is not an ideal subject of study. The endometrium, a dynamic tissue, undergoes complex physiological changes in response to ovarian steroid hormones [38]. Various ovulation-inducing treatments lead to significant hormonal fluctuations, which may influence endometrial morphology, complicating classification and discussion. In this study, we selected hormone replaced treatmenr FET cycles with or without GnRHa. From a mechanistic perspective, both the HRT protocol and the GnRHa-HRT protocol utilize exogenous hormones to induce proliferation and transformation of the endometrium. We also analyze the characteristics and clinical outcomes of FET cycles with or without GnRHa pre-treatment to eliminate the effects of exogenous or endogenous gonadotrophins. Between 2020 and 2021, HRT and GnRHa-HRT were our primary endometrial preparation protocols, with clinicians not selectively assigning these two protocols to patients. Moreover, our study excluded patients with moderate or severe endometriosis or adenomyosis. The proportion of these 2 protocols in the 4 types of the endometrium were similar (Tables 1 and 2), and did not affect the live birth rate (Supplementary Table 1).

In cleavage-stage embryo FET cycles, the embryo implantation rate, clinical pregnancy rate, early miscarriage rate, and live birth rate were similar in 4 groups and showed no association with endometrial morphology. Regarding the blastocyst cycles, a significant decrease in the live birth rate was observed in the Pattern A group (19.05%), compared with Pattern B (51.67%), Pattern B-C (50.00%), and Pattern C (43.01%). Univariate analysis indicated that a triple line endometrium on the blastocyst transfer day was associated with a higher first trimester miscarriage rate and a lower live birth rate. Within the 10 clinically pregnant cases in the Pattern A group, 6 women did not achieve a live birth. 4 women experienced first trimester abortions, with 3 undergoing villus tissue analysis, all yielding euploid results. 1 woman had an ectopic pregnancy, which was the her second ectopic pregnancy post-embryo transfer. One woman experienced an ectopic pregnancy, her second following an embryo transfer. Another woman underwent induced labor due to abnormal fetal development. Among the six patients, none exhibited severe male factor infertility, and all the transferred blastocysts originated from in vitro fertilization (IVF) cycles characterized by high quality. Through the analysis of the total data and individual abnormal pregnancy cases, we conclude that a trilaminar endometrium on the day of blastocyst transfer is associated with early pregnancy loss. This may be attributed to inadequate decidualization or “retarded” endometrium in the luteal phase, resulting in trophoblast cell failure to continue invasion and development.

The echo pattern serves as an ultrasonic indicator of different stages of the endometrium, however, the underlying molecular mechanism remains elusive. During the proliferative phase, the endometrial glands and blood vessels grow rapidly due to exposure to estrogen, resulting in a trilaminar structure and increased endometrial thickness. The process of endometrial transformation from proliferative phase to secretory phase under the steroids hormonal milieu is called endometrial decidualization. Upon entering the secretory phase, the endometrium ceases proliferation with rising progesterone levels, while further vascular and glandular development contribute to increased density instead of thickness or volume. Fleischer et al. speculated that the hyperechoic texture of secretory endometrium was related to the increased storage of echogenic mucin and glycogen in the distended and tortuous endometrial glands [39], whereas Grunfeld et al. hold the view that the homogeneous hyperechoic endometrium in the late secretory phase might indicate the stromal edema after comparing the endometrial chronological date with glandular histology and stromal histology respectively [40]. A fully developed luteal phase endometrium is more receptive to the implantation of a blastocyst, allowing for better adherence. In contrast, a retarded endometrium during the luteal phase has been reported to have a higher miscarriage rate and a lower pregnancy rate than one with normal development [41]. The persistence of a proliferative-like appearance likely indicates delayed endometrial development and insufficient decidualization [42]. ‘.

Endometrium is a highly dynamic tissue undergoing striking physiological changes in response to ovarian steroid hormones [43]. Stimulation by progesterone (P) during the luteal phase results in the growth and activation of endometrial glands, which release secretory products into their lumens [44]. Progesterone is the decisive factor in the endometrium’s transition to the secretory phase, and a deficiency in progesterone secretion can lead to abnormal gland development, thus resulting in an abnormal transformation of the endometrial morphology [45]. However, previous research suggested that serum P cannot always accurately predict endometrial development, and no relationship between echo patterns and serum P levels was observed [46, 47]. In our studies, the serum progesterone levels on blastocyst transfer day was slightly lower in Pattern A group, with no statistical difference. Univariate analysis indicated that higher serum progesterone levels on the the day of progesterone initiation were associated with the occurrence of a triple line endometrium on the day of blastocyst transfer. It seems challenging to clarify the significance of the minute progesterone changes at the molecular/clinical level in terms of endometrial morphology transformation. Progesterone receptor presents periodic dynamic changes in the endometrium, and high progesterone receptors in the pre-ovulatory phase promoted endometrial responsiveness to progesterone stimulation after ovulation [48]. Abnormal PR-B expression may lead to the persistence of a proliferative endometrium [42]. Therefore, we should not study serum progesterone levels at a single time point in isolation. Instead, the progesterone receptor pathway should be integrated to explain the effects of progesterone on endometrial decidualization, ultrasound morphology, and clinical outcomes from both molecular and clinical perspectives.

It is important to note that a satisfactory secretory phase of endometrial transformation could occur based on the full development of the proliferative phase, which requires sufficient estrogenic activity [49]. In our study, serum estrogen levels on the progesterone initiation day in the Pattern C group were higher than those in the other three groups during blastocyst transfer cycles, and lower serum estrogen levels on blastocyst transfer day were associated with the occurrence of a triple line endometrium. From 2020 to the present, we have collected two samples from the fifth day post-ovulation, where ultrasound indicated an endometrium of Pattern A. We conducted ER and PR testing on these two samples alongside other Pattern C endometrial samples (Supplementary Fig. 2). Due to the limited number of samples, quantitative analysis was not feasible; however, we observed an intriguing case with trilinear pattern ultrasonically, where ERa predominantly expressed in the cytoplasm, rather than in the nucleus. The purpose of estrogen priming and the achievement of endometrial proliferation is to induce progesterone receptors, which enable subsequent progesterone stimulation to promote endometrial receptivity. Estrogenic stimulation may significantly affect the subsequent luteal phase, and the luteal progression of the endometrium depends not only on the duration and intensity of progesterone stimulation but also on prior estrogen priming [50]. We therefore consider the morphologically untransformed endometrium of mid-luteal phase may be closely related to aberrant estrogen pathway. Due to the limited availability of secretory phase endometrial samples exhibiting the triple line morphology, we need to further increase the sample size for more in-depth research.

The underdeveloped secretory endometrium associated with luteal phase defect was first described by Noyes in 1959 [51]. Subsequently, Moszkowski et al. [52] and Wentz [53] demonstrated that this condition may reflect one of four abnormalities: (a) deficient corpus luteum production of progesterone, (b) an abnormally high estrogen-progesterone ratio, (c) inadequate estrogenic stimulation of the endometrium, or (d) the endometrium’s inability to respond to normal hormonal stimulation. A previously reported study did not find any relationship between the pre-ovulatory echo pattern and endometrial steroid hormone receptor concentration; however, the ratio of progesterone to estradiol concentration was slightly lower when a pre-ovulatory triple line echo pattern was not observed [45]. Future studies may determine whether the ratio of serum estrogen and progesterone or the ratio of endometrial estrogen and progesterone receptors affect luteal echo pattern. Additionally, future research could ascertain whether administering extra estrogen or progesterone during the luteal phase can alter reproductive outcomes.

Endometrial compaction, characterized by the reduction of endometrial thickness from the late proliferative to the secretory phase, has been suggested as a predictor of endometrial responsiveness to progesterone and has been proposed as a novel marker for embryo implantation receptivity [54, 55]. Compaction of the endometrial lining during FET cycles resulted in a significant increase in the ongoing pregnancy rate compared with cycles in which the lining did not compact [54]. In this study, we found that during blastocyst transfer cycles, when the endometrium transformed from a trilaminar pattern to a homogeneous hyperechogenic pattern, endometrial compaction increased. This was consistent with the viewpoint that endometrial compaction is associated with sustained pregnancy. Future research could encompass the comprehensive analysis of endometrial morphology alongside the endometrial compaction.

Unlike the measurement of endometrial thickness, the assessment of endometrial echo possesses a degree of subjectivity. In Julian A’s 2015 study, within 180 FET cycles, the endometrial types on blastocyst transfer day were classified as 14 type 2 and 166 type 3, with no type 1 endometrium observed. This could be related to medication methods or ethnic demographics, but it is most likely associated with the subjectivity of the ultrasound physicians. Endometrial pattern is the most crucial observational indicator in our article. Each image was evaluated by the same two doctors, ensuring more stable interpretation. However, this also suggests that if we wish to use endometrial echo pattern as a primary clinical judgment criterion, we should establish a more detailed evaluation system. In the future, artificial intelligence can be utilized to accurately identify endometrial echo features through deep learning, thereby quantifying enhanced accuracy and reproducibility in clinical practice. Simultaneously, endometrial thickness, blood flow, and other parameters can be combined to predict endometrial receptivity and clinical pregnancy outcomes.

In this study, we found that a triple-line endometrium on the day of blastocyst transfer during FET cycles with hormone replacement therapy is associated with a higher rate of first trimester abortions and a lower rate of live births. Our research presented a potentially effective clinical evaluation dimension to improve the live birth rate, particularly for patients experiencing repeated pregnancy loss. It may be reasonable to cancel the embryo transfer plan when a trilinear pattern endometrium occurs on the day of blastocyst transfer in FET cycles, especially if the embryo is precious or the patient has had an adverse pregnancy history. If patients have previously experienced mid-luteal phase endometrial non-transformation, we can increase the dosage and method of estrogen administration, or switch to alternative endometrial preparation protocols in subsequent treatment cycles. There were also limitations to our study. We conducted a retrospective study and the sample size was limited. Among the 540 blastocyst transfer cycles, only 19 cases exhibited untransformed endometrium, with a detection rate as low as 3.74%. Therefore, we need to further increase the patient sample size and consider the design of prospective experiments as well as multicenter clinical studies to enhance the accuracy of the results for their effectiveness in clinical application.

Conclusions

In this manuscript, we discussed the endometrial morphology on cleavage-stage and blastocyst in hormone replacement therapy FET cycle. Our retrospective study suggested that a trilinear pattern endometrium on blastocyst transfer day was associated with a higher first trimester abortion rate and lower live birth rate. Estrogen and estrogen receptor may play a significant role in the transformation failure of the endometrium. It is reasonable to cancel the embryo transfer plan when a trilinear pattern endometrium occurs on the day of blastocyst transfer in hormone replacement therapy FET cycles, especially if the embryo is precious or the patient has had an adverse pregnancy history.