Embryo implantation outside the uterine cavity represents an abnormal and dangerous form of pregnancy referred to as ectopic pregnancy (EP) (1), and according to reports, this condition is responsible for less than 1% of all maternal deaths in developing countries, with the figures rising to 5% for developed ones (2).

Assisted reproductive technology (ART) can theoretically decrease the incidence of EP, as both fertilization and embryo transfer (ET) do not involve the fallopian tubes. However, the rate of EP is around 1%–2% in spontaneous pregnancy as compared to 1.4%–5.4% in the case of ART (3, 4), and this can generally be attributed to risk factors such as low BMI (5), fresh ETs compared with freeze-thaw cycles (6–11), transfer of multiple embryos (12), and tubal factor infertility (TFI) (12–17). Similarly, the developmental stage of transferred embryos may be an important factor although its impact on EP incidence remains debatable (18–21).

In addition, few studies have also investigated the association between endometrial thickness (EMT) and EP occurrence after ART treatment, but there is no consensus on which EMT is applicable for EP prevention. Thus, identifying endometrial risk factors of EP and endometrial characteristics that could potentially predict EP after ART can be important.

Figure 1 shows the flowchart for cycle admission. This study included a total of 5,960 freeze-thaw cycles, including 5,777 IUP and 183 EP cycles.

The baseline characteristics were as shown in Table 1 . The women’s age, BMI, infertility type and duration, endometrial type, male factor infertility, polycystic ovary syndrome, diminished ovarian reserve, intrauterine adhesions, scarred uterus, and endometrial preparation regime were not significantly different between two groups. Significant differences were, however, found for previous history of EP, the developmental stage of the transferred embryos, the number of embryos transferred, gravidity, TFI, and endometriosis (P < 0.05).

EMTs of women with EP were significantly thinner in comparison with IUP patients (P <0.001). In addition, the EP rate in women with EMT < 8 mm (5.2%) was significantly higher than for those with an EMT of 8–9.9 mm (3.9%), 10–11.9 mm (1.5%), 12–13.9 mm (0.9%), and ≥ 14 mm (0%) (P < 0.001). In fact, none of the patients with an EMT of ≥ 14 mm developed EP in subsequent analysis and as such, those with EMT of 12–13.9 mm were selected as the reference group.

Table 2 shows the results of the univariate analysis. Compared with an EMT of 12–13.9 mm as the reference, the risk of EP in patients with an EMT of < 8 mm was fivefold (OR, 6.084; 95% CI 2.154–17.189; P = 0.001), with the risk decreasing to threefold in the case of those with an EMT of 8–9.9 mm (OR, 4.530; 95% CI, 1.663–12.337; P = 0.003). However, the EP risk for women with an EMT of 10–11.9 mm was not statistically different from that of the reference group (i.e., 12–13.9 mm) (OR, 1.738; 95% CI, 0.605–4.994; P = 0.304).

The risk factors for EP were a small BMI, an EP history, TFI, multiple embryos transfer, and transfer of embryos at the cleavage stage. In contrast, history of IUP and endometriosis infertility was protective against EP, whereas endometrial preparation regime was not identified as a risk factor for EP.

For multivariate stepwise regression analysis, all variables with a P-value of < 0.1 during univariate analysis were included, with the results shown in Table 3 .

A small BMI (aOR, 1.001; 95% CI, 1.000–1.002; P = 0.034) and blastocyst transfer (aOR, 0.451; 95% CI, 0.297–0.683; P < 0.001) were protective factors against EP after FET cycles. On the other hand, TFI (aOR, 2.221; 95% CI, 1.191–4.144; P = 0.012), a previous history of EP (aOR, 1.573; 95% CI, 1.102–2.243; P = 0.012) and an EMT of < 8 mm (aOR, 3.270; 95% CI, 1.113–9.605; P = 0.031) before P administration were found to independently predict EP. An EMT of 8–9.9 mm (aOR, 2.758; 95% CI, 0.987–7.707; P = 0.053) and 10–11.9 mm (aOR, 1.456; 95% CI, 0.502–4.225; P = 0.489) was not significantly linked with EP occurrence.

Figure 2 shows the prediction model for EP, with the embryo stage, BMI, previous history of EP, and TFI as the variables. The suitability of the model was assessed with an ROC curve for which the area under the curve was 0.651 (95% CI 0.612–0.689, P < 0.001). After including EMT as one of the variables of the model, the AUC increased to 0.686 (95% CI 0.650–0.722, P < 0.001) ( Figure 3 ).

This retrospective analysis of frozen ET cycles sought to investigate the possible link between EMT and EP after freeze-thaw transfer.

In this study, TFI increased the risk of EP by more than twofold, with the results supported by previous studies (12, 13, 15, 26, 28–31). Similarly, women with low BMI were more likely to develop EP after ART, with this outcome being consistent with a previous research (5). As for multiple embryos transfer, although some studies (26, 28, 32) reported that it was risk factor for EP, another one considered that it actually had no impact (29). In fact, previous history of EP (29, 33) is similarly debatable. In the present work, multiple embryos transfer was not identified as a risk factor.

Consistent with some previous studies (34–37), transferring embryos at cleavage stage was a risk factor of EP in the current study. Embryos at this stage could be more prone to “traveling around”, unlike blastocysts that tend to immediately seek contact and attachment. Furthermore, a study has shown that decreased uterine contractions on the fifth day could also lower EP rate after blastocysts transfer (19). However, some studies (19, 38–41) do suggest that EP occurrence was not influenced by the stage of embryos. In fact, some reports even suggest that blastocysts transfer could actually heighten EP risks due to potentially higher implantation rates of each blastocyst (20, 25).

These inconsistencies could be the result of variations between studies, especially in terms of patients’ age, sample sizes, evaluation system for embryos, and differences in blastocysts culture techniques between reproductive centers.

Few studies have examined the suitability of EMT during ART therapy for predicting EP, but the currently applicable cutoff value that link EMT and EP is controversial. One study in which fresh or frozen embryo cycles were not separately analyzed suggested that, prior to ET, an EMT > 12 mm protected against EP (29). In the same vein, another study reported increased risks of EP when the EMT was < 12 mm in the frozen embryo cycle (28). The current study indicated that an EMT of < 8 mm significantly increased risks of EP. Because it is clinically impractical to perform ET only if the EMT is greater than 12mm, it would therefore be more clinically meaningful to select an EMT of < 8 mm rather than 12 mm as the threshold for being considered a risk factor for EP.

It remains unclear why a thin endometrium increases risks of EP rates, with the reason likely to be complex. Many researchers have suggested that there is a relationship between EMT and uterine receptivity (42–47), and, as such, it is generally believed that thinner EMTs may lead to a poorer endometrial receptivity.

Differences in oxygen concentrations could also explain the link between a thin endometrium and EP. Indeed, a thin EMT would bring embryos quite close to the spiral arteries in the basal endometrium layer, thereby exposing them to high oxygen concentrations which are known to inhibit embryonic development (48).

Uterine peristalsis could be another factor that links EP incidence with a thin endometrium. Previous studies have reported that, compared with IUP, women diagnosed with EP experienced higher uterine peristaltic wave frequencies, but these differences were not statistically significant due to uneven distribution of the sample size (49). A previous research further showed the EMT thickness was positively associated with risks of placenta previa (50). In this case, the authors hypothesized that high EMTs were an indication of the amplitude and/or frequency of uterine peristalsis waves, which can push the embryos downward, dislodging them from their transferred location. Another study (51) found that, compared with natural cycles, controlled ovarian hyperstimulation cycles showed increased uterine waves from the cervix to the fundus but reduced ones from the fundus to the cervix. Therefore, although the results of that study suggest that the direction of uterine peristalsis could influence EP occurrence, yet this link would need to be confirmed through additional studies. Altogether, this study’s findings point out that uterine peristalsis from the fundus to the cervix is more likely to occur when the endometrium is thicker as the embryos are more easily implanted in the lower segment of the uterine cavity, resulting in a higher incidence of placenta previa (50) and, correspondingly, a lower incidence of EP. Future studies will focus on how different EMT are associated with the frequency, amplitude, and direction of endometrial peristalsis waves, including the correlation between EP occurrence and the presence of different types of endometrial peristalsis at the time of ET.

As EP risks were higher for patients with an EMT of < 8 mm, especially when the transferred embryos were in the cleavage stage or when multiple embryos were transferred, such patients should be given clear advice. Furthermore, it should be carefully decided whether frozen ET should be performed in the current cycle while routinely examining the endometrial peristaltic wave (49, 52) and providing appropriate treatments such as phloroglucinol (53) and atosiban (54) when necessary as these would help in preventing EP or diagnosing the condition at an early stage.

This study was not without limitations, with the first one being that it was a retrospective study. Chromosomal abnormalities of embryos as a potential risk factor for EP were also not included as not all embryos were tested for chromosomes before transfer, but it was previously reported (55, 56) that chromosomal abnormalities in embryos were not actively involved in the etiology of EP. In addition, the study contained a high rate of multiple embryos transfer. Some researchers (26, 28) believe that multiple embryos transfer is an independent risk factor for EP as well as HP, but this study did not find this variable to be an independent risk factor after multivariate regression analysis, consistent with a previous study (32). This could have been due to the exclusion of HP in the present research.

In short, the results can be very meaningful in the absence of a consensus about the optimal EMT, which can predict or prevent EP after ART.

An EMT < 8 mm on the day of endometrial transformation was found to independently predict EP after freeze-thaw ET. Efforts to increase the EMT may reduce EP risks. It may also be necessary to perform endometrial peristaltic wave examination, endometrial receptivity testing as well as provide corresponding treatment in the peri-transfer period for patients with risk factors.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by YZ. The first draft of the manuscript was written by YZ, and all authors commented on previous versions of the manuscript. All authors contributed to the article and approved the submitted version.

This study was supported by the National Key Research and Developmental Program of China(2018 YFC1004800) and the National Natural Science Foundation of China (grant no. 8187061497).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.