Expressions of candidate molecules in the human fallopian tube and chorionic villi of tubal pregnancy exposed to levonorgestrel emergency contraception

Ectopic pregnancy (EP) is the leading cause of pregnancy-related death in the first trimester of pregnancy, and it accounts for approximately 2% of all recognized pregnancies [1]. More than 98% of the EPs occur in the fallopian tube, particularly in the ampulla [2, 3]. The pathogenesis of tubal pregnancy remains unclear. One hypothesis is conditions that delay or prevent passage of the embryo through the tube and into the uterine cavity may lead to ectopic implantation [3]. It has been proposed that an elevated progesterone concentration could theoretically impair the fallopian tube cilia motility and thus increase the risk of EP in women using a progestin-only pill [4]. Levonorgestrel (LNG), a synthetic progestogen, is effective and widely used for emergency contraception (EC) either as a single dose of 1.5 mg or in two doses of 0.75 mg with 12 hours apart [5]. When administered within 72 h after an unprotected intercourse, it prevents about 57–93% of unwanted pregnancies resulting from unprotected intercourse [6, 7]. The LNG only EC regimen acts via inhibiting or delaying ovulation, but has no detectable effect on the endometrial receptivity or steroid hormone receptors expressed in the fallopian tube from non-pregnant women seeking sterilization [8‐10]. Although there is some epidemiological and pharmacological evidence, an eventual causative effect of LNG-EC in the development of EP has not been elucidated. Moreover, compared to intrauterine pregnancy, some changes in the endosalpinx environment and chorionic villi are specific to tubal pregnancy due to the anatomic structures involved. A variety of molecules expressed in the chorionic villi and fallopian tube, such as E2 receptors (ERs), P receptors (PRs), leukemia inhibitory factor (LIF), vascular endothelial growth factor (VEGF), inducible nitric oxide synthase (iNOS), and endocannabinoid receptor 1 (CB1), are potential regulators of the pathophysiological changes in tubal pregnancy [11‐25]. Cases of EP following LNG-EC failure were constantly reported as LNG-EC pills are more and more widely used, however, up to now the effect of LNG on the tubal microenvironment or chorionic villi in EP, which is the aim of this study, has not been documented.

Fifty-five tissue samples were collected from 27 women in the LNG-EC group and 28 women in the control group. Three women were diagnosed ruptured tubal pregnancy by pathologic evaluation, one of whom claimed having administered LNG-EC during the cycle of conception whereas the rest two had not. There were no significant differences in the age, gestational week, volume of abdominal hemorrhage, serum concentrations of beta-hCG, E2, P, or the occurrence of ruptured tubal pregnancy in women between the two groups (Table 3).

Levenorgesgtrel-EC acts via inhibiting or delaying ovulation [8, 9], but has no demonstrated effect on blastocyst viability or hatching, blastocyst attachment, early implantation, incidence of miscarriage, fetal malformation, pregnancy complications, or any other pregnancy outcome associated with intrauterine pregnancy [28, 29]. Up to now, there have been no studies specifically focusing on the molecular changes in either fallopian tube or chorionic villi in tubal pregnancy exposed to LNG-EC.

Same as in intrauterine pregnancy, P, E2 and hCG play crucial roles in ectopic pregnancy. The fallopian tube undergoes decidualization synchronized with the trophoblast development, and that process is regulated by the coordination of P and E2 acting via their receptors [15, 28]. Vascular endothelial growth factor and its receptors are known to be involved in the processes of vasculogenesis and angiogenesis in the decidua [17]. An up-regulation of LIF expression promotes decidualization in humans and mice [21]. The iNOS was reported to be present not only in the epithelium, but also in the decidua surrounding the implanted embryos in pregnant mice [22]. The anandamide present in the uterine environment may play an important role in regulating apoptosis through one of its receptors - CB1, and thereby modulate the stability and regression of the decidua during pregnancy [24]. Horne et al.[15] reported that ER-alpha was not expressed in the fallopian tube during tubal pregnancy. However, in our study, ER-alpha was expressed in both epithelial and stromal cells of the fallopian tube collected from both groups. This difference might be due to the source of the primary antibody. The presence of ER-alpha in the fallopian tube during tubal pregnancy indicates that E2 might also play a role in tubal pregnancy.

The placenta performs a wide array of metabolic and endocrine functions that are crucial for maintaining normal pregnancy and the fetal development, which are also affected by the balance between P and E2. The expressions of VEGF and its receptors have been found to be correlated with serum beta-hCG concentration, indicating that their expressions may be influenced by the signals sent from the embryo [18]. One interpretation is that increased expressions of VEGF and its receptors promote local angiogenesis, augmenting the supply of oxygen to the embryo [19]. Leukemia inhibitory factor was demonstrated to facilitate blastocyst formation and hatching of human embryos in vitro[22]. The iNOS mRNA was detected in both cytotrophoblast and syncytiotrophoblast cells in vitro, which suggests that nitric oxide is involved in the placental growth and function [23]. Although the effect of CB1 during placentation is undefined, the endocannabinoid has been shown to be important for successful pregnancy in humans [25]. Based on the findings of the present study, we are tempting to speculate that failure of LNG-EC itself seems to have no effect on the decidualization process or placental development in tubal pregnancy.

We have eventually noticed that the methodological weaknesses in this study could have resulted in serious biases. Another ongoing study implementing a larger sample size is to address the association between the tubal pregnancy exposed to LNG-EC and its clinico-pathological features and outcomes. Furthermore, the pharmacokinetics study of LNG shows that the serum level of hormone is above the threshold from 1 to 6 h after the intake of 1.5 mg LNG and with a half-life ranging from 20 to 60 hours [30]. When the patients were diagnosed tubal pregnancy, LNG-EC has already metabolized. The immediate effect of LNG-EC to the fallopian tube is still undefined. Since the cilia beat frequency of the fallopian tube significantly reduced after the treatment with high doses of progesterone, we hypothesize that the LNG-EC, a temporarily high dose of progestin, can disturb the synthetic balance of P and E2 right after the intake. Due to the ethical limits, it is of paramount importance to conduct LNG study using in vitro culture and animal models to cast more light on the association between the LNG-EC and tubal pregnancy.

In summary, we found that administration of LNG-EC during the cycle of conception had no effect on the expressions of a variety of molecules regulating the pathophysiological changes induced by tubal pregnancy in either chorionic villi or fallopian tube. Therefore, we speculate that the pathophysiological features of tubal pregnancy following LNG-EC failure did not differ from those unexposed to LNG-EC. We propose that a routine therapy for tubal pregnancy can thus be applied to women with tubal pregnancy exposed to LNG-EC and there was no need for additional interventions.