Effectiveness evaluation of endometrial receptivity analysis for recurrent implantation failure and analysis of clinical factors affecting implantation window.

No significant differences in age, BMI, infertility type, or other baseline indicators were observed between the ERA and non-ERA groups ( P  > 0.05, as detailed in Table  1 ). However, the incidence rate of endometriosis in the ERA group was significantly lower than that in the non-ERA group (3.1% vs. 9.8%, P  = 0.02). In addition, the level of AMH in the fresh cycle of the ERA group was slightly higher than that of the non-ERA group (3.75 ng/mL vs. 3.33 ng/mL, P  = 0.02), while the remaining indicators were balanced and comparable. Table 1 Comparison of general data of RIF patients Characteristic ERA group ( n  = 131) non-ERA group ( n  = 153) χ2/Z P Age (years) 32 (29, 36) 33 (30, 37) -1.677 0.09 BMI (kg/m 2 ) 21.68 (19.69, 23.86) 21.30 (19.72, 23.15) -1.28 0.20 Infertility years 2 (1, 4) 3 (2, 5) -1.618 0.11 Primary infertility, n (%) 66/131 (50.40) 70/153 (45.80) 0.606 0.44 AMH in the fresh cycle (ng/mL) 3.75 (2.7, 5.83) 3.33 (1.72, 5.22) -2.421 0.02 Number of embryos transferred 2 (1, 2) 2 (1, 2) -1.858 0.06 Number of optimal embryos transferred 2 (1, 2) 2 (1, 2) -1.046 0.29 Endometrial thickness on the day of embryo transfer (cm) 0.90 (0.80, 1) 0.90 (0.80, 1.10) -0.268 0.79 PCOS, n (%) 45 (34.4) 42 (27.5) 1.581 0.21 Endometriosis, n (%) 4 (3.1) 15 (9.8) 5.152 0.02 Ectopic pregnancy history, n (%) 24 (18.3) 31 (20.3) 0.17 0.68 Induced abortion history, n (%) 43 (32.8) 58 (37.9) 0.796 0.37 Comparison of general data of RIF patients ERA-guided personalized transfer was associated with significantly higher CPR (88.55% vs. 32.03%) and IR (63.72% vs. 27.42%) than non-ERA transfer, but also a higher early miscarriage rate (8.40% vs. 2.61%, P  < 0.05; Table  2 ). Table 2 Comparison of pregnancy outcomes in RIF patients between the two groups Clinical outcome ERA group ( n  = 131) non-ERA group ( n  = 153) χ 2 P Biochemical pregnancy rate, n (%) 117/131 (89.31) 71/153 (46.41) 58.06 <0.001 Clinical pregnancy rate, n (%) 116/131 (88.55) 49/153 (32.03) 92.62 <0.001 Embryo implantation rate, n (%) 144/226 (63.72) 68/248 (27.42) 63.01 <0.001 Ectopic pregnancy rate, n (%) 1/131 (0.76) 1/153 (0.65) 0.01 0.91 Clinical miscarriage rate, n (%) 11/131 (8.40) 4/153 (2.61) 4.71 0.03 Comparison of pregnancy outcomes in RIF patients between the two groups Stratified analysis by embryo stage showed significantly higher CPR and IR in the ERA group for both cleavage-stage and blastocyst transfers ( P  < 0.05). Within the ERA group, early miscarriage was increased in cleavage-stage (16.67% vs. 2.22%, P  = 0.03) but not blastocyst transfers (7.10% vs. 2.78%, P  = 0.14; Table  3 ). Table 3 Comparison of pregnancy outcomes after transfer of different embryo types Embryo types Outcomes ERA group ( n  = 131) non-ERA group ( n  = 153) χ 2 P Transfer of cleavage-stage embryos Clinical pregnancy rate, n (%) 16/18 (88.89) 14/45 (31.11) 17.21 <0.001 Implantation rate, n (%) 22/35 (62.86) 19/76 (25.00) 14.74 <0.001 Ectopic pregnancy rate, n (%) 0/18 (0) 1/45 (2.22) 0.41 0.52 Clinical miscarriage rate, n (%) 3/18 (16.67) 1/45 (2.22) 4.51 0.03 Transfer of blastocyst stage embryos Clinical pregnancy rate, n (%) 100/113 (88.50) 35/108 (32.41) 73.08 <0.001 Implantation rate, n (%) 122/191 (63.87) 49/172 (28.49) 45.48 <0.001 Ectopic pregnancy rate, n (%) 1/113 (0.88) 0/108 (0) 0.96 0.33 Clinical miscarriage rate, n (%) 8/113 (7.10) 3/108 (2.78) 2.16 0.14 Comparison of pregnancy outcomes after transfer of different embryo types A multivariable binary logistic regression analysis (adjusting for P  < 0.1 factors) was performed to further analyze the performance of ERA in improving clinical pregnancy, including factors of age, BMI, AMH level, endometriosis and other baseline indicators. Results showed that ERA-guided personalized embryo transfer remained an independent factor associated with significantly higher clinical pregnancy rates (OR = 3.78, P  < 0.001, Table  4 ). Table 4 Factors associated with clinical pregnancy in recurrent implantation failure patients: univariate and multivariate analyses incorporating ERA-guided transfer Risk factors Univariate Analysis Multivariate Analysis P OR (95% CI) P OR (95% CI) ERA test < 0.001 10.2 [55.72, 18.33] < 0.001 3.78 [2.08, 6.87] AMH (ng/mL) 0.002 1.06 [1.02, 1.09] 0.02 1.04 [1.01, 1.09] Number of embryos transferred < 0.001 2.192 [1.41, 3.39] 0.001 2.15 [1.33, 3.49] Age (years) 0.07 1.070 [0.99, 1.15] 0.19 0.95 [0.89, 1.02] Endometriosis 0.06 0.41 [0.16, 1.03] 0.58 0.74 [0.26, 2.11] BMI (kg/m²) 0.09 1.11 [0.98, 1.27] - - Secondary infertility 0.03 2.15 [1.07, 4.76] 0.87 1.09 [0.40, 3.01] Ectopic pregnancy history 0.65 1.25 [0.47, 3.28] - - Infertility years 0.54 0.96 [0.85, 1.08] - - PCOS 0.46 1.34 [0.61, 2.94] - - Factors associated with clinical pregnancy in recurrent implantation failure patients: univariate and multivariate analyses incorporating ERA-guided transfer Of 131 ERA patients, 87 (66.41%) had WOI displacement (85 pre-/early receptive, 2 post-receptive). No significant differences in age, PCOS, endometriosis, embryo parameters, or endometrial thickness were observed between the WOI displacement and normal WOI groups ( P  > 0.05). CPR was similar between groups (88.51% vs. 88.64%, P  = 0.98). However, the WOI displacement group had higher proportions of induced abortion history (40.23% vs. 18.18%, P  = 0.01) and secondary infertility (56.32% vs. 36.37%, P  = 0.03), with a trend toward higher BMI ( P  = 0.07; Table  5 ). Table 5 Analysis of general information and pregnancy outcomes of RIF patients with and without WOI displacement Index WOI normal group ( n  = 44) WOI displacement group ( n  = 87) χ 2 /Z P Age (years) 31.86 (29, 34.75) 32.79 (30, 36) -0.27 0.21 BMI (kg/m 2 ) 22.04 (19.96, 24.53) 20.81 (19.44, 23.05) -1.76 0.07 Infertility years 3.57 (1.25, 4.75) 3.29 (1, 4) -0.94 0.35 Primary infertility, n (%) 28/44 (63.63) 38/87 (43.68) 4.66 0.03 AMH in the fresh cycle (ng/mL) 4.73 (2.58, 5.68) 5.31 (2.85, 5.95) -1.57 0.70 PCOS, n (%) 13/44 (29.55) 32/87 (36.78) 0.68 0.41 Endometriosis, n (%) 2/44 (4.54) 2/87 (2.30) 0.03 0.89 Ectopic pregnancy history, n (%) 7/44 (15.90) 17/87 (19.54) 0.26 0.61 Induced abortion history, n (%) 8/44 (18.18) 35/87 (40.23) 6.44 0.01 Number of embryos transferred 1.75 (1.25, 2) 1.72(1, 2) -0.35 0.73 Endometrial thickness on transfer day (cm) 0.95 (0.86, 1) 0.94 (0.8, 1) -1.32 0.19 Clinical pregnancy rate 39/44 (88.64) 77/87 (88.51) 0.00 0.98 Analysis of general information and pregnancy outcomes of RIF patients with and without WOI displacement Univariate logistic regression showed induced abortion history (OR = 2.26, P  = 0.01), secondary infertility (OR = 2.15, P  = 0.03), and elevated BMI (OR = 1.11, P  = 0.09) were associated with increased WOI displacement risk. No associations were found with age, PCOS, or endometriosis ( P  > 0.1). Multivariate analysis (adjusting for P  < 0.1 factors) showed induced abortion history ( P  = 0.08, OR = 2.90) and BMI ( P  = 0.09, OR = 1.12) remained marginally significant; the association with secondary infertility was attenuated ( P  = 0.87; Table  6 ). Table 6 Analysis of influencing factors of WOI displacement in RIF patients Risk factors Univariate Analysis Multivariate Analysis P OR (95% CI) P OR (95% CI) Induced abortion history 0.01 2.26 [1.26, 7.29] 0.08 2.90 [0.89, 9.47] Secondary infertility 0.03 2.15 [1.07, 4.76] 0.87 1.09 [0.40, 3.01] BMI(kg/m²) 0.09 1.11 [0.98, 1.27] 0.09 1.12 [0.98, 1.29] Age (years) 0.17 1.07 [0.97, 1.17] PCOS 0.46 1.34 [0.61, 2.94] Endometriosis 0.47 0.48 [0.06, 3.54] Infertility years 0.54 0.96 [0.85, 1.08] Ectopic pregnancy history 0.65 1.25 [0.47, 3.28] Analysis of influencing factors of WOI displacement in RIF patients

A retrospective analysis was conducted on 284 ART patients with RIF who underwent frozen embryo transfer (FET) at Jinjiang Maternal and Child Health Hospital (Chengdu, China) from January 2022 to December 2023. Patients were stratified into (1) ERA group ( n  = 131): Underwent ERA, further divided into WOI displacement ( n  = 87, pre- or post-receptive endometrium) and normal WOI ( n  = 44, receptive endometrium) subgroups; (2) non-ERA group ( n  = 153): Declined ERA. The protocol of this study has been reviewed and approved by the Ethics Review Committee of Chengdu Xi’nan Gynecology Hospital (approval no. 2024-001). Signed informed consent was obtained from all participants. Inclusion: (1) Oocyte retrieval age < 43 years; (2) RIF (≥ 3 failed transfers of high-quality embryos); (3) FET performed after RIF diagnosis. Exclusion criteria: (1) abnormal karyotype in either or both male and female partners; (2) patients with endometrial lesions (such as adhesions in the uterus, endometrial hyperplasia, polyps of the endometrium, endometritis, endometrial tuberculosis, etc.); (3) patients with hydrosalpinx and without proximal tubal ligation; (4) submucosal fibroids; (5) combined with medical or surgical diseases, such as severe liver and kidney insufficiency, coagulation dysfunction, autoimmune diseases, endocrine or metabolic diseases, etc. Before the FET cycle, patients were tested for ERA. All included patients were treated with artificial cycles of hormone replacement: Starting from the third day of menstruation, patients were administered oral estradiol valerate (Progynova, Bayer, Germany) at a dosage of 3 mg twice daily (bid). Vaginal Doppler ultrasound was then performed to monitor endometrial thickness and determine if any adjustments to the drug dosage were necessary. Hormone was used for 10–14 days, and dydrogesterone tablets (Dapton, Abbott, the Netherlands) 10 mg tid was given when the endometrium was ≥ 8 mm. At the same time, Progesterone Sustained-release vaginal gel (Merck Serono, Switzerland) 90 mg qd were administered vaginally, and endometrium was harvested 120 h after transformation. Endometrial sampling: Before sampling, the cervix of the patient was cleaned with normal saline, and the vaginal discharge and blood routine examination were performed on the day of surgery. The upper and lower lip of the cervix were clamped by cervical forceps to fix the cervix. The probe was inserted into the uterus to its full depth. Once the sampling device reached the uterine fundus, the left hand held and fixed the tube sleeve, the right hand held the handle, and quickly pulled the tube core out to form negative pressure to absorb endometrial tissue. The endometrial tissue in the sampling tube was drained onto sterile gauze, and 5 to 10 mm 3 of clean, blood-free endometrial tissue was selected and immediately placed into 1.5 ml RNA-later buffer (batch No. AM7021, Thermo Fisher Scientific, USA), sealed and frozen at -20℃ until use. Cryopreserved tissue was sent to Hangzhou Yizhen Medical Laboratory. Total RNA was extracted using a QIAGEN QIAcube workstation and spin-column kits (Qiagen, Germany), with RNA integrity number (RIN) ≥ 7 required for analysis [ 6 ]. Within 7 days, RNA extraction, library construction, high-throughput sequencing, and analysis were performed using validated prediction software (CLIA/CAP-accredited laboratory). The endometrial receptivity status was assessed by the ERA computational predictor as described previously. The ERA results of the endometrial tissue classification included three phases: receptive, pre-receptive, and post-receptive [ 6 , 7 ]. According to the recommended WOI predicted by the artificial intelligence model, when the endometrium is in an ideal state for embryo transfer, individualized embryo transfer is performed: that is, the exact time for embryo transfer after the use of progesterone is determined according to the “guiding recommendations” in the test results. According to the results of ERA, the endometrium was prepared for cryo-embryo transfer again. According to the results of ERA, the individual time of embryo transfer was arranged. All the patients in the WOI displacement group used the same hormone replacement artificial cycle protocol as the simulated cycle of this patient for endometrial preparation and underwent cleavage stage and blastocyst stage embryo transfer according to the WOI determined by the ERA results. In the WOI normal group and the control group, cleavage stage embryo transfer was performed 72 ± 3 h after endometrial transformation, or blastocyst stage embryo transfer was performed 120 ± 3 h after endometrial transformation. Serum β-human chorionic gonadotropin (hCG) level was measured 2 weeks after embryo transfers. If β-hCG ≥ 25 mIU/mL, biochemical pregnancy was defined. Transvaginal ultrasound was performed 28–35 days after embryo transfers to determine whether there was an intrauterine gestational sac. Clinical pregnancy was determined if a gestational sac was present. Miscarriage was defined as pregnancy termination after less than 28 weeks of gestation and fetal weight less than 1000 g. Baseline data: Age, BMI, infertility type/duration, polycystic ovary syndrome (PCOS, per Rotterdam criteria), endometriosis (surgically proven or imaging-confirmed within 5 years), induced abortion/ectopic pregnancy history, number/quality/type of transferred embryos, AMH, endometrial thickness. Outcomes: CPR = (clinical pregnancy cycles/transfer cycles) ×100%; Implantation rate (IR) = (ultrasound-detected gestational sacs/transferred embryos) ×100% (multiple sacs from one embryo = 1); Miscarriage rate = (miscarriage cycles/transfer cycles) ×100%. Data was analyzed using SPSS 27.0. Categorical variables (rates, proportions) were compared via chi-square test; normally distributed continuous variables (mean ± SD) via independent samples t-test; non-normal variables [M(P25, P75)] via Mann-Whitney U test. Univariate and multivariate binary logistic regression identified factors associated with WOI displacement. For the modeling population, univariate and multivariate binary Logistic model was employed to identify the factors affecting implant window displacement. All statistical tests were two-sided, and P  < 0.05 was considered statistically significant.

This study evaluated ERA’s utility in RIF patients and identified potential WOI displacement factors. The key findings revealed that ERA-guided personalized embryo transfer significantly improved clinical pregnancy and embryo implantation rates in RIF patients, though it was accompanied by a higher early miscarriage rate; most WOI displacements were in the pre-receptive phase, and a history of induced abortion were likely associated with WOI displacement. Successful embryo implantation relies on the precise synchronization of embryo development and endometrial receptivity, and the WOI (a transient period of maximal endometrial receptivity) is critical to this process [ 8 ]. Traditional FET protocols adopt a “one-size-fits-all” approach (e.g., 72 ± 3 h for cleavage stage embryos, 120 ± 3 h for blastocysts), which fails to account for individual variations in WOI. For RIF patients, WOI displacement is a common endometrial factor [ 3 ]. Traditional endometrial receptivity assessment methods such as ultrasound, hormone indicators such as estrogen and LH, endometrial morphological markers such as peristaltic waves and pinopods, and molecular biological markers such as integrin, which can evaluate endometrial receptivity to a certain extent, but they are controversial due to their limitations and poor repeatability [ 9 , 10 ]. ERA resolves this limitation by detecting endometrial receptivity at the transcriptomic level: it identifies differentially expressed genes related to cell adhesion and hormone response, thereby accurately classifying the endometrium into receptive, pre-receptive, or post-receptive phases and guiding personalized transfer timing [ 6 ]. In this study, 66.4% of patients had a displaced WOI and the ERA group presented a significantly higher CPR (88.55% vs. 32.03%) than the control group. These findings indicated that ERA testing was associated with improved pregnancy outcomes. This result aligns with the study by Jia et al. [ 11 ], which reported that ERA-guided transfer increased CPR by 25% in RIF patients. Notably, our stratified analysis further confirmed that ERA improved CPR and IR regardless of whether cleavage stage or blastocyst stage embryos were transferred. This suggests that endometrial receptivity (rather than embryo stage) is the primary limiting factor for implantation in RIF patients, and ERA effectively addresses this bottleneck. In addition, the results of this study showed that early miscarriage rate of cleavage-stage embryo transfer in the ERA group was higher than that in the control group. This may be attributed to the embryo quality differences between cleavage and blastocyst stages. Blastocysts have undergone natural selection, reducing the risk of aneuploidy-related miscarriage. For blastocyst stage embryos, the combination of ERA (optimizing endometrial conditions) and natural selection during blastocyst formation (improving embryo quality) may achieve favorable outcomes. While future randomized controlled studies with larger sample sizes are needed to validate these findings. Among 87 patients with WOI displacement, 85 (97.7%) were in the pre-receptive/early receptive phase, and only 2 (2.3%) in the post-receptive phase. This distribution is consistent with previous studies [ 11 , 12 ], which reported that ~ 90% of WOI displacements in RIF patients are pre-receptive. A potential physiological mechanism is that artificial hormone replacement cycles may delay the transition from the proliferative to secretory phase. This finding has clinical implications: for RIF patients undergoing ERA, most will require a slight delay in embryo transfer (to align with the delayed WOI), rather than an advance. Clinicians can prioritize adjusting progesterone administration duration (e.g., extending progesterone exposure by 12–24 h) based on ERA results, which simplifies personalized transfer protocols. Clinical pregnancy outcomes between the WOI displacement group and the normal WOI group were similar, which is consistent with the reported conclusions [ 13 , 14 ].These findings confirm that ERA-guided pET can effectively match the timing of embryo transfer for different receptivity phases and eliminate the negative impact of WOI displacement on pregnancy outcomes. In addition, exploring the influencing factor of WOI displacement helps to identify the specific benefited subgroups of ERA. According to the existing studies, the displacement of endometrial WOI in patients with RIF may be related to advanced age [ 15 , 16 ], obesity [ 17 ], severe previous failed cycles [ 12 ], endometriosis [ 18 ], PCOS [ 19 ], ectopic pregnancy history [ 20 ] etc. Such patients need ERA testing to clarify the WOI timing more, to avoid blind transfer leading to repeated failure. In this study, logistic regression analysis showed that the history of induced abortion was a marginally associated factors for WOI displacement, which has not been reported before. Uterine manipulation during abortion can directly damage the endometrial basal layer, which maybe influence the endometrial receptivity. The lack of significant factors in multivariate analysis in this study may be due to two limitations: (1) small sample size: the normal WOI group included only 44 patients, reducing the statistical power to detect weak associations; (2) confounder collinearity: secondary infertility and induced abortion history are highly correlated (many secondary infertility patients have a history of abortion), leading to overlapping variance in multivariate models and masking individual effects. Despite the study’s limitations, it is speculated that subgroups with a history of induced abortion might have a potentially higher risk of WOI displacement. Given the limited statistical power and potential confounding, further validation studies are needed. In addition, elevated BMI also presented a marginal association with an increased risk of WOI displacement, indicating weight management may have a protective effect on WOI—a finding that aligns with earlier research [ 17 ]. This study has several limitations that should be considered when interpreting the findings. First, single-center retrospective design: the study was conducted at a single hospital with a limited sample size. This may introduce selection bias (e.g., socioeconomic status, treatment compliance, unrecorded lifestyle or medical details) and limit the external validity of the results. Second, while multivariate analyses adjusted for key covariates (e.g., endometriosis, AMH levels), residual confounding cannot be fully excluded. Third, lack of live birth rate data: though the CPR and IR get improved by ERA-guided transfer, the long-term benefits should also be evaluated. Multicenter prospective randomized controlled trials with larger samples and with longer follow-up period are required to confirm the efficacy of ERA. In conclusion, ERA combined with personalized embryo transfers program has positive significance for improving the outcome of RIF patients.

With advancements in assisted reproductive technology (ART), the clinical pregnancy rate of in vitro fertilization-embryo transfer (IVF-ET) has gradually increased from 6% to 57% [ 1 ]. The development of preimplantation genetic diagnosis (PGD) and preimplantation genetic screening (PGS) has further enhanced clinical pregnancy rates through the transfer of euploid blastocysts. However, recurrent implantation failure (RIF) remains a major challenge [ 2 ]. At present, RIF is mostly defined as failure to implant after ≥ 3 transfers of high-quality embryos (≥ 4–6 cleavage-stage or ≥ 3 blastocyst-stage embryos) [ 3 ], with complex etiologies involving both embryo quality and endometrial receptivity. Endometrial receptivity—the endometrium’s ability to accept embryos—is critical for embryo localization, adhesion, and implantation, peaking during the window of implantation (WOI), a brief period when synchronization between embryo and endometrium is essential. If the synchronization deviates by more than 1.5 days, the success rate of embryo transfer significantly decreases [ 4 ]. WOI displacement is a key endometrial cause of RIF, making accurate WOI prediction vital. Endometrial receptivity analysis (ERA), a molecular diagnostic assay based on high-throughput sequencing and transcriptomics, detects endometrial gene expression to identify WOI and guide personalized transfer timing, improving outcomes [ 5 ]. In this study, we first analyzed pregnancy outcomes following ERA-guided embryo transfer in RIF patients to assess the technique’s efficacy, and second, examined factors linked to WOI displacement by analyzing the clinical profiles of patients with this condition—ultimately seeking to provide a reference for optimizing RIF management.