Background
Over the past decades, advances in clinical and laboratory techniques have substantially improved pregnancy rates and live-birth rates for assisted reproductive technology (ART). However, as reported by the Centers for Disease Control in the United States by the end of 2012, only 38.1% of all transfer cycles could result in live births [
1]. Maternal age has been proved as the most predominant predictor for successful ART outcome [
2]. During the process of ART treatment, the endometrial receptivity, and the number and quality of embryos transferred are also regarded as the key determinants for the outcome of in vitro fertilization (IVF)/intracytoplasmic sperm injection (ICSI) cycles [
3‐
5].
As an indicator of endometrial receptivity, endometrial thickness (EMT) on the day of human chorionic gonadotropin (hCG) administration has been reported and reaffirmed as a potential prognostic tool for ART outcomes in multiple studies [
6‐
8], despite significant advancements of ultrasonic [
9], immunologic [
10] and molecular [
11] markers for endometrial receptivity. Although it remains a controversial issue, it has been widely suggested that a thin endometrium is associated with lower chance to conceive after IVF/ICSI, with cut-off values of EMT varying 7–9 mm in earlier studies [
6,
12]. Also, several investigators have addressed the question whether there is a threshold value for thickness above which implantation is unlikely to occur, but the conclusion is contradictory [
13,
14]. In a one-center study of 3350 IVF cycles, the non-linear association of EMT with live-birth was shown and required verification [
7].
The number and quality of embryos transferred are the most important factors for IVF/ICSI outcome [
15,
16]. Embryo morphological features assessed by optical microscopy are routine criteria applicable for embryo quality evaluation [
17]. Several classification systems for embryo quality have been developed, and embryos have been graded into three categories: Good-quality, Fair-quality and Poor-quality [
18,
19]. Evidences support the positive role of good-quality embryo for a better outcome of IVF [
20,
21]. In addition, although two embryos transferred in one cycle have higher probability of live-birth than one embryo, multiple pregnancy may lead to the elevation of maternal and neonatal risks [
22]. Therefore, to determine the optimal number and quality of embryos to transfer, cumulative embryo score (CES) and other embryo scoring methods based on the number of good-quality embryos transferred have been proposed [
15,
23,
24] to predict pregnancy outcomes. However, the efficacy of these embryo scoring systems needs more optimization, because the association between quality of different embryo combinations and live-birth was not taken into consideration.
Successful pregnancy depends upon implantation, a complex process involving reciprocal interactions between the receptive endometrium and functional embryo in ART cycles. However, little is known about this embryo-endometrial interaction on live-birth. Thus, we hypothesized that the combination of EMT and embryo quality may improve the prognostic value in clinical practice for live-birth. Therefore, this multicenter study was designed to assess the association of EMT, embryo quality with live-birth independently and their interaction on live-birth based on a large scale of IVF/ICSI cycles.
Discussion
In the current study, we confirmed the nonlinearity in EMT-LBR association and a plateau of LBR per cycle when EMT was 11 mm or thicker. The embryo quality graded by cumulative score was significantly associated with LBR for cleavage-stage and for blastocyst ETs, and these associations were independent of the number of embryo transferred and EMT. To the best of our knowledge, this is the first multicenter retrospective cohort study to evaluate the independent and interaction effects of EMT on hCG day and embryo quality on the outcome of fresh IVF/ICSI cycles.
Our results are consistent with the reported association of thin EMT with poor pregnancy outcomes in fresh ET cycles [
4,
7,
12], while a consensus is still lacking on what the precise definition of thin endometrium. The present study showed EMT above 11 mm is related to a higher chance of delivery. This finding indicates that adequate endometrial development is favorable for improving chance of live-birth, which concurs to some extent with one recent study [
12]. As discussed in the meta-analysis, a thickness threshold of 7 mm was frequently reported below which pregnancy rates decreased rapidly, but the case of EMT below 7 mm occurred infrequently [
6] and the threshold was not available for live-birth evaluation. Therefore, there may be of more clinical significance for the EMT cut-off value of 11 mm, which could be more helpful for decision making on fresh ET or freezing of embryos for potentially better chances to conceive with an endometrium developed under natural conditions in subsequent cryo-cycles. However, some others could not establish a significant correlation between EMT and the chance to conceive in the study population with euploid ETs [
26]. Debate on the predictive value EMT in clinical outcomes is ongoing [
6]. In order to investigate the real endometrium’s effect, it is interesting to investigate the value of EMT in patients undergoing frozen embryo transfer cycles in the further study because in this population the possible negative effect of ovarian hyperstimulation on endometrial development is absent.
The mechanism of the association between thin endometrium and difficulty for implantation and development remains elusive. One speculation relates to that the implanting embryo would be much closer to the basal layer endometrium with higher oxygen concentrations in patients with thin functional layer EMT [
27]. It is well known that high oxygen tensions could be detrimental for embryo implantation and development due to the production of reactive oxygen species [
28]. In addition, it has been speculated that low late-follicular estradiol levels could result in inadequate endometrial proliferation [
7,
29]. Follicular estrogen production is a reflection of follicular maturity and impacts both endometrial receptivity and follicular growth, oocyte maturation, sperm transport, embryo survival [
30]. In other words, it is possible that the observed reduction in pregnancy rates associated with a thin endometrium may be a consequence of another unmeasured aspect of oocyte/embryo quality besides their morphologic grading. Furthermore, it is plausible that thin endometrium may be related to iatrogenic event like Asherman’s syndrome [
31] or use of clomiphene citrate, leading to an E receptor abnormality [
27] and the subsequent lack of a normal proliferative response to the rising E level.
The thickened endometrium provides a site for attachment, as well as the source of nourishment for an implanting embryo during its first few weeks, until the placenta starts to develop. Whether a thick endometrium has a detrimental effect on clinical outcome is controversial [
12,
13]. Our findings indicated that thick EMT (>17 mm) neither played an adverse role nor conferred additional benefit, similar to results reported in recent studies [
12,
32], which suggested a ceiling effect of EMT after excluding patients with intrauterine pathologies such as polyps or fibroids.
Owing to the adverse perinatal risks and multiple pregnancy related costs, increased use of elective single-embryo transfer (eSET) in couples attempting assisted conception has been proposed and gradually adopted by some countries, such as Sweden and Belgium [
22]. In other countries, however, the adoption of eSET in practice is hampered due to the desire of both clinicians and patients to ensure pregnancy and to reduce the costs of multiple ART cycles, especially in countries where ART is not publicly funded [
33]. Opinion should, therefore, shift towards the view that the decision to transfer one or two embryos should be made according to not only the number but quality of available embryos. With regard to single embryo quality, there was a robust correlation between morphologic parameters and live-birth in previous studies transferring one cleavage-stage embryo [
34] or one blastocyst [
35]. In general, our findings support the validity of morphologic grading to be used to guide the embryo selection.
In addition, we also have realized that the association of live birth with embryo quality of different embryo combinations is poorly understood, while it is imperative for decision-making of optimal embryos number and quality to transfer. CES and mean embryo score (MES) based on 4-point embryo score have been proposed to predict pregnancy outcomes restricted to cleavage-stage embryos transfer [
17,
36]. Our data provided adequate support for the efficiency of CES (Poor =1 point, Fair =2 points, Good =3 points) in predicting live-birth when restricted to cleavage-stage ETs. For blastocyst transfer, it is also appropriate to determine the group of blastocyst embryo quality by cumulative score although we observed that a good-quality blastocyst transfer had similar LBR compared to that of a fair-quality blastocyst and combinations of blastocyst which was graded ‘Poor’ or ‘Fair’, which may be potentially caused by the limited cycles. To reduce the risk of multiple gestations, eSET was recommended to women with both one fair or good embryo and embryo pairs with the same cumulative score. Our findings doubted about other embryo scoring methods based on the number of good-quality embryos transferred [
15] and grading of best embryo transferred [
37] applied to distinguish different embryo combinations and to predict live-birth. Consequently, our findings filled the gaps of current embryo scoring systems and could help clinicians and infertile couples make a better choice.
It is estimated that embryos account for one-third of implantation failures, while suboptimal endometrial receptivity and altered embryo-endometrial dialogue are responsible for the remaining two-thirds [
38]. A two-way communication between the embryo and receptive endometrium leads to implantation and the course of pregnancy, but ethical restrictions and the lack of mechanistic studies have delayed studies on embryo-endometrial interactions in humans [
39]. Endometrial receptivity and selectivity are two complementary concepts introduced to describe the endometrium, the emerging concept as a biosensor of embryo quality [
40]. Either impaired embryo development potential or impaired endometrial selectivity/receptivity has a negative effect on the embryo-endometrial cross-talk. It is gradually acknowledged that the thickness of the endometrium can be used as a noninvasive parameter to infer on endometrial receptivity. Therefore, the investigation of interaction between embryo quality and EMT would provide some evidence. Our research showed that LBR could not be improved by thicker EMT when only one poor cleavage-stage embryo is available for transfer. In this circumstance, it is recommendable to restart the treatment for better embryo quality. What’s more, the LBR did not increase significantly with thicker EMT with transfer of two good-quality cleavage-stage embryos and blastocysts in groups of embryo quality except in Q1 group. Our findings suggested that the requirement for optimal endometrial environment is less stringent when high quality embryos are available for transfer. It coincides with a previous study with a smaller sample size [
29]. It seemed that the additional 2–3 days delay may provide an inadequate endometrium with additional time to develop. Taken together, the significant embryo-endometrial interaction on live-birth suggests that the combination of EMT and embryos quality may improve the prognostic value in clinical practice for live-birth in patients undergoing transfer of 1–2 fresh cleavage-stage embryos, and for blastocyst transfers, embryo quality may be more important for treatment outcome compared with EMT.
Our findings were strengthened by 2-mm EMT-LBR relationship evaluation in consideration of clinical value in practice [
7], in addition to establish a cutoff value of 11 mm for endometrial thickness above which LBR per cycle reached a plateau. Relying on a large sample size, transfers of different combinations of embryos were first compared with a good-quality ET to predict live-birth in detail, which confirmed the efficiency of CES separately for both cleavage-stage embryos and blastocysts and might compensate for the deficiency of some embryo scoring methods and improve embryo selection. It is noteworthy that the independent effects were confirmed in a prediction model with analysis of 100 prospectively recorded variables [
4], while this was the first time to estimate the interaction of endometrial characteristics and embryo quality on LBR of fresh IVF/ICSI cycles. Still, the weaknesses of this study included its retrospective nature of the data, the possibility of unidentified confounding variables which could not be excluded, and the reproducibility of EMT and embryo grade assessment of different observers in different centers. Moreover, due to only a few blastocyst transfer cycles were conducted before 2016 (only 5.2% included in this study), these findings of blastocyst transfer required verification in future studies based on a larger sample. The observations in our study were based on fresh cycles, so the conclusion may not apply to frozen cycles.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.