Impact of endometrial thickness before frozen-warmed blastocyst transfer on live birth rate in women with endometriosis

Out of 640 autologous artificial FETs, after PSM we included 95 single blastocyst transfers in women with endometriosis and 190 single blastocyst transfers in women without endometriosis. The characteristics of patients are presented in Table  1 . LBR was overall was 29.5% (84/285) and CPR was 57.2% (163/285). The EMT ranged from 6 to 15 mm. Live birth occurred for EMT between 6 to 13 mm. Table 1 Clinical characteristics of the included patients with endometriosis and without endometriosis. Mean age, body mass index (BMI) and level of Anti-Mullerian hormone (AMH) are shown with standard deviation in parentheses. In all other cases number of patients (with percentage according to the population in each category above) are represented. Chi-square was used for the comparison between categorical variables and Mann-Whitney U test was used for continuous variables. P <0.05 was considered as statistically significant Endometriosis ( n =  95) Women without endometriosis ( n =  190) p -value Age (years) 36.4 (3.3) 36.5 (3.8) 0.617 BMI (kg/m2) 21.8 (2.9) 21.9 (4.1) 0.482 AMH level (ng/ml) 2.1 (1.2) 3.7 (2.7) < 0.001 Double-vitrification 19 (20.0%) 39 (20.5%) 1 primary infertility 79 (83.2%) 144 (75.8%) 0.173 male infertility 36 (37.9%) 60 (31.6%) 0.291 PCOS 0 (0%) 31(16.3%) < 0.001 Unexplained infertility 0 (0%) 80 (42.1%) < 0.001 Method of fertilization 0.802  IVF 48 (50.5%) 100 (52.6%)  ICSI 47 (49.5%) 90 (47.4%) Protocol stimulation < 0.001  Antagonist 35 (36.8%) 125 (65.8%)  Short agonist 26 (27.4%) 48 (25.3%)  Long agonist 34 (35.8%) 17 (8.9%) Quality of embryos after vitrification 0.234  Low-quality blastocyst (BC/CB/CC) 0 (0%) 5 (2.6%)  Good-quality blastocyst 60 (63.2%) 110 (57.9%)  Top-quality blastocyst (AA) 35 (36.8%) 75 (39.5%) rASRM I&II 47 (49.5%) - rASRM III&IV 48 (50.5%) - Clinical characteristics of the included patients with endometriosis and without endometriosis. Mean age, body mass index (BMI) and level of Anti-Mullerian hormone (AMH) are shown with standard deviation in parentheses. In all other cases number of patients (with percentage according to the population in each category above) are represented. Chi-square was used for the comparison between categorical variables and Mann-Whitney U test was used for continuous variables. P <0.05 was considered as statistically significant The EMT (endometriosis group 8.7 (1.8) mm vs. control group 8.6 (1.46) mm, p =  0.932), CPR [endometriosis group 56.8% (54/95) vs. control group 57.4% (109/190), p =  1.000] and LBR [endometriosis group 27.4% (26/95) vs. control group 30.5% (58/190), p =  0.679] were comparable between the two groups (Table  2 ). The conditional density plots (CDP) for both women with endometriosis and women without endometriosis did not demonstrate either a linear relationship between the EMT, CPR and LBR or a threshold below which the CPR or LBR decreased perceivably. The peak LBR in women without endometriosis was observed at an EMT of 8 mm, while in women with endometriosis the peak was observed at 10 mm (Figs. 1 , 2 ). Table 2 Endometrial thickness is represented in mean millimeters (mm) and standard deviation in parentheses. Below, pregnancy and live birth rate are shown with number of patients and percentage in each group. P -value was calculated using Mann-Whitney U test for mm and Chi-square test for pregnancy and birth rate Endometriosis ( n =  95) Women without endometriosis ( n =  190) p -value Endometrial thickness (mm) 8.7 (1.8) 8.6 (1.46) 0.932 Pregnancy rate 54 (56.8%) 109 (57.4%) 1 Live birth rate 26 (27.4%) 58 (30.5%) 0.679 Fig. 1 Clinical pregnancy rates according to the endometrial thickness in women without endometriosis ( A ) and women with endometriosis ( B ). Women with 12 mm or more are grouped together. The range in women without endometriosis include a thickness of 12 mm to 15 mm. In women with endometriosis the maximal thickness was 12 mm Fig. 2 Live birth rate according to the endometrial thickness in women without endometriosis ( A ) and women with endometriosis ( B ). The range of thickness in the last group include a range of 12 mm to 14 mm in women without endometriosis and only one patient with endometriosis with 12 mm Endometrial thickness is represented in mean millimeters (mm) and standard deviation in parentheses. Below, pregnancy and live birth rate are shown with number of patients and percentage in each group. P -value was calculated using Mann-Whitney U test for mm and Chi-square test for pregnancy and birth rate Clinical pregnancy rates according to the endometrial thickness in women without endometriosis ( A ) and women with endometriosis ( B ). Women with 12 mm or more are grouped together. The range in women without endometriosis include a thickness of 12 mm to 15 mm. In women with endometriosis the maximal thickness was 12 mm Live birth rate according to the endometrial thickness in women without endometriosis ( A ) and women with endometriosis ( B ). The range of thickness in the last group include a range of 12 mm to 14 mm in women without endometriosis and only one patient with endometriosis with 12 mm ROC curve analyses did not suggest a predictive value of EMT for live birth for both women with and without endometriosis and the area under the curve values were 0.532 (95% CI 0.392–0.672) and 0.456 (95% CI 0.372–0.540). For clinical pregnancy the ROC curve analyses showed area under the curve values for women with endometriosis 0.494 (95% CI 0.376–0.612) and without endometriosis 0.474 (95% CI 0.390–0.559) (Fig. 3 ). Fig. 3 Receiver Operating Curve (ROC) for parametric prediction of pregnancy outcome. The ROC curve was plotted with pregnancy outcome = 1 and no pregnancy = 0. Area under the curve (AUC) Receiver Operating Curve (ROC) for parametric prediction of pregnancy outcome. The ROC curve was plotted with pregnancy outcome = 1 and no pregnancy = 0. Area under the curve (AUC) The subgroup analysis according to EMT showed that women with endometriosis have significantly lower LBR for EMT between 8 and 10 mm (9/47 (19.1%) in women with endometriosis and 34/83 (41.0%) in women without endometriosis, p  = 0.012) (Table 3 ). Even after using the ideal EMT in both groups, i.e. setting a cutoff of 8 mm for women without endometriosis and 10 mm for women with endometriosis we did not find a difference between both groups (Table 4 ). The subgroup analysis according to rASRM stage showed that women with endometriosis rASRM I-II and rASRM III-IV with clinical pregnancies and live births had comparable endometrial thickness (8.21 ± 1.22 mm vs. 8.96 ± 1.51 mm, p = 0.084 and 8.14 ± 1.41 mm vs. 9.33 ± 1.67 mm, p = 0.070). The ROC analyses in these subgroups did not suggest a predictive value of EMT for clinical pregnancy and live birth for both subgroups rASRM I-II (AUC = 0.503 (95% CI 0.328–0.677) and AUC = 0.482 (95% CI 0.294–0.669), respectively) and rASRM III-IV (AUC = 0.517 (95% CI 0.352–0.683) and AUC = 0.517 (95% CI 0.352–0.683), respectively) (Fig. 4 ). Table 3 Pregnancy and live birth rates according to different cut-off of endometrial thickness in women with and without endometriosis. Significant levels ( p <0.05) are marked in bold Endometrial thickness 6–8 mm Endometriosis ( n =  25) Women without endometriosis ( n =  54) p -value Pregnancy rate 15 (60.0%) 29 (53.7%) 0.635 Live birth rate 8 (32.0%) 14 (25.9%) 0.598 Endometrial thickness 8–10 mm Endometriosis ( n =  47) Women without endometriosis ( n =  83) p -value Pregnancy rate 25 (53.2%) 53 (63.9%) 0.266 Live birth rate 9 (19.1%) 34 (41.0%) 0.012* Endometrial thickness 10–12 mm Endometriosis ( n =  20) Women without endometriosis ( n =  35) p -value Pregnancy rate 13 (65.0%) 18 (51.4%) 0.403 Live birth rate 8 (40.0%) 6 (17.1%) 0.106 Endometrial thickness > 12 mm Endometriosis ( n =  3) Women without endometriosis ( n =  18) p -value Pregnancy rate 1 (33.3%) 9 (50.0%) 0.593 Live birth rate 1 (33.3%) 4 (22.2%) 0.676 Table 4 Subgroup analysis comparing women without endometriosis and endometrial thickness of <8mm vs. ≥8mm. And for women with endometriosis a comparison between less vs. 10mm or more was conducted using Chi-square test Subgroup analysis women without endometriosis and cut-off for endometrial thickness 8 mm Endometrial thickness  = 8 mm ( n =  136) p -value Pregnancy rate 29 (53.7%) 80 (58.8%) 0.521 Live birth rate 14 (25.9%) 44 (32.4%) 0.485 Subgroup analysis women with endometriosis and cut-off for endometrial thickness 10 mm Endometrial thickness  = 10 mm ( n =  23) p -value Pregnancy rate 40 (55.60%) 14 (60.9%) 0.810 Live birth rate 17 (23.6%) 9 (39.1%) 0.181 Fig. 4 ROC curve analysis for prediction of pregnancy and live birth rate in a subgroup of rASRM I&II and III&IV Pregnancy and live birth rates according to different cut-off of endometrial thickness in women with and without endometriosis. Significant levels ( p <0.05) are marked in bold Women without endometriosis ( n =  54) Women without endometriosis ( n =  83) Women without endometriosis ( n =  35) Women without endometriosis ( n =  18) Subgroup analysis comparing women without endometriosis and endometrial thickness of <8mm vs. ≥8mm. And for women with endometriosis a comparison between less vs. 10mm or more was conducted using Chi-square test ROC curve analysis for prediction of pregnancy and live birth rate in a subgroup of rASRM I&II and III&IV

This is a retrospective cohort study, including autologous single blastocyst transfers from frozen-thawed cycles of women with and without endometriosis who underwent embryo transfer between January 2017 and June 2023 at the University Hospital Zurich in Switzerland. All embryo transfers were performed following artificial endometrial preparation for single blastocyst transfer. All participants provided written informed consent for the use of their data in research, and the study was approved by the local ethics committee of the Canton of Zurich (BASEC No. 2018–00796). All women who underwent a single autologous FET were eligible for inclusion in the study. Exclusion criteria were single embryo transfers from preimplantation genetic testing (PGT) cycles, vitrified-warmed oocyte cycles and cycles with surgically retrieved sperm samples. Women with untreated intrauterine pathologies, congenital uterine anomalies, EMT thinner than 6 mm as well as hydrosalpinx were excluded from the study. For artificial endometrial priming, patients received oral estradiol (6 mg/day) starting on the first day of menstruation. EMT was monitored by ultrasound between days 10 and 14. EMT was assessed on a midsagittal image with transvaginal ultrasound measuring the maximum anterior–posterior thickness of the endometrial echo. When the EMT was 6 mm or more, patients began vaginal progesterone (1000 mg/day), and the blastocyst transfer was performed six days after the initiation of progesterone. All cryopreservation was performed using vitrification. For single embryo transfers, blastocysts were cultured in either Global Total or G2 (Vitrolife) and transferred intrauterinely using the Guardia™ Access Embryo Transfer Catheter (COOK Medical, Bloomington, IN, USA) within one minute of loading. Prior to vitrification of zygotes, the presence of distinct two-pronuclear (2PN) zygotes was confirmed to ensure that only those not yet undergoing syngamy were cryopreserved. All 2PN zygotes and blastocysts were vitrified using the Cryotop method, which utilizes an open carrier system consisting of a polypropylene strip covered by a protective sheath. Excess vitrification medium was removed from the strip, leaving only a minimal film surrounding the cells, thereby facilitating ultra-rapid cooling and warming rates—approximately 2300 °C/min and 4210 °C/min, respectively [ 13 ]. This approach is recognized for its high efficiency and minimal cryoinjury. For all procedures in this study, vitrification (VT601) and warming (VT602) solutions from Kitazato (Kitazato Corporation, Shizuoka, Japan) were used in combination with the Cryotop® Open System for oocyte/embryo vitrification. Between one and a maximum of three 2PN-stage zygotes were loaded per Cryotop device and subsequently thawed for embryo transfer. Zygotes were vitrified between 17 and 21 h post-insemination. For blastocyst cryopreservation, one blastocyst was allocated to each Cryotop. Embryos were assessed according to the Gardner grading system (Alpha Scientists in Reproductive Medicine and ESHRE SIG Embryology, 2011). Only blastocysts with a grade of at least 3BB on day 5 or 4BB on day 6 were considered suitable for freezing; those rated as grade C for inner cell mass (ICM) or trophectoderm (TE) were cryopreserved only under exceptional circumstances, for example when it was the only blastocyst available from a cycle of couples with multiple failed cycles and advanced maternal age. Post-warming, blastocysts were allowed to recover for a minimum of two hours prior to transfer. LBR defined as the delivery of a live fetus after 22 weeks of gestation was the primary outcome. CPR characterized as a pregnancy which was confirmed by ultrasound was the secondary outcome. Continuous variables were described using means and standard deviation, while categorical variables were described using frequencies and percentages. A Propensity score matching (PSM) 2:1 using the nearest neighbor matching to adjust for male infertility, fertilization method, quality of blastocyst, double vitrification, maternal age and BMI was performed [ 14 ]. The presence of a linear relationship between the EMT, LBR and CPR were assessed visually by conditional density plots (CDPs) of the LBR and CPR across EMT. Receiver operating characteristic (ROC) curve analyses and visual inspection of CDPs were used to identify whether a threshold of the EMT existed to predict the occurrence of live birth and clinical pregnancy [ 15 ].

In conclusion, our findings suggest that in EMT above 6 mm no specific threshold is associated with higher LBRs in women with or without endometriosis undergoing artificial FET cycles. Further studies are needed to clarify the role of endometriosis—and the commonly associated adenomyosis—in the eutopic endometrium and its impact on endometrial receptivity and related pregnancy chances.

To the best of our knowledge, this is the first study to examine the role of EMT in women with endometriosis undergoing FET in artificial cycle. The results of our study did not demonstrate a linear relationship between EMT on the day of progesterone administration and either LBR or CPR in both the endometriosis and non-endometriosis groups. Furthermore, no specific EMT cutoff for women with EMT more than 6 mm could be identified below which LBR or CPR significantly decreased in artificial FET cycles, regardless of endometriosis status. Additionally, the predictive values of EMT for clinical pregnancy and live births were low in both groups. Although a difference was observed in the EMT associated with peak LBR between groups—10 mm in the endometriosis group versus 8 mm in the control group—this finding may be incidental. A subgroup analysis revealed significantly lower LBRs in women with endometriosis with an EMT of 8–10 mm compared to the control group with the same EMT range. Although Mahutte et al., after examining 43,383 fresh IVF-ET cycles and 53,377 FET cycles, showed that pregnancy chances are significantly reduced with EMT less than 6 mm [ 16 ], previous studies have similarly failed to find a linear relationship between EMT above 6 mm and LBR or to define a threshold below which LBR significantly declines, however these studies showed that an EMT about 10 to 11 mm correlates with better live birth rates in FET cycles [ 3 , 17 ]. Similarly, Mahutte et al., found peak LBRs with an EMT of 10–12 mm in fresh cycles and 7–10 mm in FET cycles [ 16 ]. As EMT more than 6 mm, does not seem to play a significant role in FET-related pregnancy chances, qualitative features of the endometrium may explain differences in CPRs and LBRs between women with and without endometriosis. A substantial body of evidence suggests that eutopic endometrium and endometrial receptivity differ in women with endometriosis compared to those without the condition [ 18 ]. Several pathogenic mechanisms have been proposed, including immune dysregulation, chronic inflammation, and hormonal imbalances—specifically, disordered estrogen and progesterone signaling characterized by progesterone resistance and estrogen dominance [ 19 ]. Moreover, alterations in the expression of molecular biomarkers affecting gene regulation and receptivity have been described [ 9 , 12 , 20 ]. Despite these findings, a pilot study which examined the endometrial receptivity transcriptomics with the endometrial receptivity assay (ERA) test did not find differences in the expression of the 238 genes included in ERA test between women with endometriosis and controls [ 21 ]. Last but not least, clinical outcomes from oocyte donation cycles and preimplantation genetic testing for aneuploidy (PGT-A) cycles have shown comparable implantation rates between women with and without endometriosis [ 22 , 23 ]. Our study focuses on differences between women with and without endometriosis who underwent medical endometrial preparation before FET. Artificial-cycle FET has long been the standard approach for endometrial preparation, largely due to the ease of synchronizing embryo transfer with the schedules of the IVF laboratory, clinical team, and the patient [ 24 ]. However, the last years, several clinicians support a shift to natural-cycle FET, as recent evidence increasingly suggests that natural-cycle FET, because of the presence of the corpus luteum is associated with better obstetric and neonatal outcomes compared to artificial-cycle FET, such as lower rates of early pregnancy loss, preterm birth, very preterm birth, hypertensive disorders of pregnancy, pre-eclampsia, placenta previa, and postpartum hemorrhage [ 25 ]. Ata et al. found that both artificial- and natural-cycle FETs achieved peak LBRs with an endometrial thickness of 11 mm, however, the predictive values of EMT in both the above endometrial preparations were low [ 3 ]. A previous retrospective study comparing pregnancy and prenatal outcomes across different endometrial preparation protocols in women with endometriosis did not find significant differences between the artificial- and natural-cycle FET groups. However, it is important to note the potential for selection bias, as the artificial-cycle FET group was considerably larger (950 versus 74 embryo transfers) [ 26 ]. This is, to our knowledge, the first study to evaluate the effect of EMT in artificial single FET cycles among women with endometriosis. Although the single-center, retrospective, observational design has inherent limitations, for example, patients were not systematically assessed using the MUSA criteria for adenomyosis, the use of propensity-score matching reduced selection bias and resulted in two groups with comparable baseline characteristics. In our study—as in most studies assessing the effect of EMT in FET—ultrasound measurement was performed before the initiation of progesterone; however, some studies measured EMT on the day of FET [ 27 ]. Recent studies have examined changes in EMT between the start of progesterone and the day of embryo transfer, but there is no clear evidence that such changes affect clinical outcomes [ 28 ]. Further prospective studies are needed on the EMT differences between women with and without endometriosis on natural-cycle FET which, as noted above, is increasingly relevant in clinical practice. In addition future studies on EMT and endometriosis should also account for coexisting uterine conditions, including adenomyosis and other pathologies such as chronic endometritis, uterine malformations, isthmoceles, and endometrial polyps [ 9 – 12 , 29 – 32 ]. Altogether, larger study groups, especially for the sub-analysis and including endometrial thickness below 6 mm would be beneficial.

The endometrium, which consists of epithelial, stromal, immune, and vascular cells, undergoes cyclical changes during the menstrual cycle under the influence of hormones [ 1 ]. These changes transform it into an environment optimized for embryo attachment. The ability of the endometrium to provide a supportive setting for embryo development and placentation is referred to as endometrial receptivity [ 1 ]. It is estimated that impaired endometrial receptivity and disrupted embryo–endometrial dialogue are responsible for approximately two-thirds of implantation failures [ 2 ]. Various markers of endometrial receptivity—such as endometrial thickness (EMT), pattern, blood flow, and wave-like activity—have been associated with clinical pregnancy, however, their predictive value remains poor [ 2 ]. The independent effect of EMT on live birth rates (LBR) after embryo transfer, and whether there is a threshold below which the chances of live birth decrease, remains unclear. EMT is associated with endometrial receptivity and should be evaluated using transvaginal ultrasound before embryo transfer, however although numerous previous studies using different cutoff EMT values have been conducted, the results are contradictory and there is still no consensus on the optimal EMT for fresh or frozen embryo transfer (FET) [ 3 ]. An EMT of less than 6 mm warrants investigation for potential causes, such as intrauterine adhesions or Asherman syndrome [ 4 ]. A thin endometrium appears to negatively affect the outcome both of fresh and frozen embryo transfer with lower clinical pregnancy rate (CPR) and LBR, nevertheless there is no EMT at which pregnancy is impossible [ 5 ]. A possible mechanism proposed for the reason why the thin endometrium may decrease the embryo implantation and the live birth rates is that it brings the embryo closer to the spiral arteries, exposing it to higher oxygen concentrations, which can be harmful to its development [ 3 , 6 ]. Although the eutopic endometrium in women with endometriosis exhibits various molecular and cellular alterations across multiple pathways, whether endometrial receptivity is impaired in these patients remains controversial [ 7 , 8 ]. Furthermore, endometriosis often coexists with other uterine pathologies that may negatively affect endometrial receptivity, such as uterine anomalies, chronic endometritis, endometrial polyps, and adenomyosis [ 9 – 12 ]. Although the first study by Check et al. in 1995, which compared EMT between women with and without endometriosis undergoing laparoscopy, found no significant difference between the two groups, to the best of our knowledge, no study has explicitly assessed the prognostic role of EMT in fresh or FET among women with endometriosis. This study aims to evaluate the role of EMT in women with and without endometriosis on the LBR in single artificial FET cycles, examining if there is a linear relationship between the EMT and if there is an optimal EMT which is associated with higher LBR.