Ectopic pregnancy: models and medical management

Abstract

Ectopic pregnancy, derived from the Greek “ektopos” meaning ‘out-of-place’, describes a pregnancy where the conceptus implants and develops at an abnormal location, most commonly the Fallopian tube. It complicates 1-2% of all pregnancies and is life-threatening due to aberrant placental trophoblast invasion of maternal blood vessels, which can result in catastrophic intra-abdominal haemorrhage. Until the 21st century, the only treatment option was surgical, in most instances involving removal of the tube containing the pregnancy. Following advances in ultrasound and the introduction of quantitative serum measurement of human chorionic gonadotrophin (hCG), it is now possible to select ‘lower risk’ cases where it is safe to offer expectant monitoring (until spontaneous resolution) or medical management. Medical treatment, with the chemotherapeutic methotrexate, can be offered to 40% of women presenting with an ectopic pregnancy. However, the efficacy of methotrexate is such that it is only offered to women presenting with low hCG levels; a group in which reported success rates, time-to-resolution and risk of haemorrhage are similar to that of expectant monitoring. There is consequently a clinical need to develop more efficacious and less toxic medical therapies for ectopic pregnancy. Further advances in therapy require development of experimental model systems to allow pre-clinical testing of drugs and to improve understanding of the pathophysiology of this condition; the latter traditionally limited by the fundamental difficulty in differentiating cause from effect in ex vivo specimens. Using in vitro and in vivo systems, this thesis investigates: 1) new potential experimental models of ectopic pregnancy; 2) methotrexate’s mechanism of action at the level of the trophoblast in the treatment of ectopic pregnancy; and 3) use of gonadotrophin releasing hormone (GnRH) receptor antagonists as a new medical treatment for ectopic pregnancy. To develop an in vivo model of ectopic pregnancy, mouse blastocysts, hatched from the zona pellucida, were transferred to oviducts of recipient mice. I found, due to murine embryos inherent ability to undergo embryonic diapause (a phenomenon not seen in humans) that this process did not result in ectopic implantation. I was able to generate an in vitro model of ectopic pregnancy, where mouse blastocysts attached to human Fallopian tube cells at low frequency (in comparison to the high numbers of blastocysts which attach to endometrial cells). This mirrors the in vivo situation where ectopic implantation occurs in a minority of pregnancies. Manipulation of this system could offer new insights into the pathogenesis of ectopic pregnancy. The extravillous trophoblast (EVT) cells of an ectopic embryo, which invade the maternal Fallopian tube, are a therapeutic target in ectopic pregnancy. EVT cell lines are routinely used in placental research but there is some disagreement in the literature as to how closely they represent primary EVT cells. In my routine assessment of EVT cell lines for use in this thesis, I unexpectedly identified co-expression of cytokeratin 7 (an epithelial marker, traditionally used as a trophoblast marker) with vimentin (a mesenchymal marker, traditionally used to demonstrate non-trophoblastic origin). This suggests these cells may have retained the ability to undergo partial epithelial mesenchymal transition, a characteristic of the primary cells from which they are derived. It is widely presumed that the mechanism of action of methotrexate in ectopic pregnancy is inhibition of DNA synthesis, as is the case where it is used at a 20-fold higher dose in the treatment of certain cancers. However, in vitro replication of maternal serum concentrations achieved following treatment with the low dose of methotrexate used in ectopic pregnancy, demonstrated only mild inhibition of cell cycle progression in just one of three EVT cell lines tested. In contrast, persistently high methotrexate concentrations, which more closely replicate cancer therapy, were able to inhibit DNA synthesis. Countertherapeutically, methotrexate was noted to increase trophoblast production of the pro-pregnancy hormones, hCG and progesterone. In rheumatoid arthritis, a different mechanism of action for methotrexate has been identified secondary to the extracellular accumulation of the anti-inflammatory nucleoside adenosine. I showed that EVT cell lines express adenosine receptors and secrete adenosine, both of which are stimulated by treatment with methotrexate. What role this may have in treatment of ectopic pregnancy requires future investigation. Gonadotrophin releasing hormone (GnRH) is a peptide that acts via its G-protein coupled receptor (GnRHR) as a central regulator of the hypothalamic-pituitary-gonadal axis. GnRHR antagonists are used routinely in artificial reproductive technologies. The GnRHR is also expressed in some reproductive cancers, where antagonism inhibits proliferation and invasion, and in some peripheral tissues including the EVT cells of ectopic pregnancies. I have shown that GnRHR antagonists inhibit trophoblast proliferation and hCG production in vitro. In mouse pregnancy, the effect of GnRHR antagonism was gestation dependent, with treatment at e9.5 causing fetal growth restriction and increased mid-gestation fetal loss, and

treatment at e7.5 inducing 100% litter loss. This pre-clinical study suggests that GnRHR antagonists could have utility as a future treatment for ectopic pregnancy. In conclusion, this thesis describes a novel in vitro model of ectopic pregnancy; identifies that the mechanism of action of methotrexate in ectopic pregnancy is not as presumed, perhaps explaining its limited efficacy; and that pre-clinical study of GnRHR antagonists suggest they may represent a promising new medical treatment for ectopic pregnancy.