1. Introduction
The placenta must guarantee the maintenance of the pregnancy and the fetal well-being through correct exchange of gases, growth factors, endocrine signals, cytokines, and nutrients. Placenta development starts at approximately 6–10 days post-conception, when trophoblast cells of blastocyst adhere to the decidua
[1]. In early gestation, the human placenta is constituted by two layers: an inner one of proliferating cytotrophoblasts, that ensures the exchange of nutrients and oxygen from maternal blood and an outer which assures a correct amount of blood during the pregnancy by invading the endometrial stroma and remodelling the uterine spiral arteries
[2]. The placentation process represents a complex and not fully understood process of immunotolerance: during the adhesion and invasion of the myometrium by the blastocyst, an immunomodulation release of pro-angiogenic and endothelial factors happens, which leads to adaptive changes of the uterine spiral arteries
[3]. New studies are focusing their attention on the origins of placental mesenchymal cells as they appear to have a pivotal role in establishing and sustaining the development of placental vasculature
[1][2][3][4]. Despite its importance in the success of reproduction, the development of the human placenta is yet to be fully understood despite an altered placentation could lead to miscarriage, unexplained stillbirth, preterm labor, placental abruption, PE, and fetal growth anomalies
[2][3][4][5][6][7].
Despite the improvement occurring in laboratory technology and clinical management of infertile women requiring IVF, this procedure is still associated with an high rate of adverse perinatal outcomes and overall placenta-related pregnancy complications
[8][9][10].
Recent meta-analysis studies have confirmed how pregnancies obtained from IVF techniques are associated with an increased risk of poor obstetric outcome including: miscarriage, chromosomal abnormalities, PE, PTB, FGR), placenta previa, abruptio placentae, post-partum haemorrhages, as well as peri and postnatal complications, such as neonatal death, low birth weight infants, congenital malformations, musculoskeletal abnormalities and childhood cancers
[11][12].
The risk of obstetric complications can be largely increased by many factors: presence of twin pregnancies after multiple embryo transfers, as well as an older pregnant population and gametes quality, previous history of recurrent abortions (RPL) and causes of infertility itself (polycystic ovarian syndrome)
[13][14]. Unfortunately, only little data are available on the explanations of such augmented risk. Different mechanisms have been assumed to play a role in the defective early placentation including genetic and epigenetic mechanisms of implantation, alterations in endometrial receptivity, invasion, and growth of the trophoblast, genetic and/or epigenetic alterations of oocyte and/or embryos due to biological manipulations, and immunotolerance in case of egg donor pregnancies
[15].
Recent meta-analyses have proved that in singleton IVF pregnancies there is an increased risk of placental abruption (RR 1.83, 95% CI 1.49 to 2.24), placenta previa (RR 3.71, 95% CI 2.67 to 5.16), antepartum (RR 2.11, 95% CI 1.86 to 2.38), and postpartum hemorrhage (RR 1.29, 95% CI 1.06 to 1.57)
[12][16]. A higher incidence of gestational hypertension and diabetes, cesarean deliveries, PTB, SGA, and perinatal mortality was also described
[16]. Nevertheless, relevant biases were present due to the inclusion in the natural conception group of women who obtained the pregnancy with ovulation induction or intrauterine insemination, leading to an underestimation of the association between ART and adverse outcomes
[16]. Consequently, the risk of developing gestational diabetes, placental abruption, PTB, fetal growth defect, and perinatal mortality may be further increased when the control group is constructed excluding these women from the spontaneous conception definition
[16] and limit the recalculation of the odds ratio or relative risk of the maternal and perinatal complications occurring in IVF women.
Of interest is the lack of IVF specific pathologies and they resemble the same characteristics of when these diseases are present in a naturally conceived population. In other words, there are no different phenotypes of PE, FGR or placenta accrete spectrum between the 2 groups of women despite the higher prevalence in the IVF group.
2. The Role of Ovarian Stimulation
Concerning safety-evaluation in IVF it is necessary to highlight the difficulties in discerning the influence on the outcomes that the underlying causes of infertility might bring versus potential risks related to IVF procedures themselves. IVF are characterized by the ovarian hormonal stimulation followed by the pick-up of the oocytes and their subsequent fertilization. This procedure implies the transfer of a single or fresher or eFET embryos. eFet embryos after thawing may be transferred in the uterus during natural or hormonally artificial induced cycles. In recent years, the number of eFET has increased and so have pregnancy rates, which are now better than those following fresh IVF embryos transfer
[17].
It has been suggested that controlled ovarian stimulation (COS) (e.g., subcutaneous gonadotropins) lead inevitably to a change in the maternal hormonal structure, determining changes in the woman’s reproductive system, as modifications of the endometrium. Different hormonal treatment strategies used in controlled ovarian stimulation and laboratory IVF techniques can negatively impact endometrial receptivity and gamete status. Hence, it was suggested that performing eFET was better than fresh embryo transfer being associated with decreased ovarian hyperstimulation incidence with improved reproductive outcomes
[18]. However, it is still unclear how the different preparation methods of the endometrium can affect the outcomes of eFET pregnancies and the selection of the treatment of choice
[19].
3. Differences between Fresh and Freeze and Thawed Embryo Transfer
Two recent systematic reviews and meta-analyses demonstrated that singleton pregnancies obtained from eFET show a more favorable maternal and perinatal outcomes than those reached after fresh oocyte transfer including a lower risk of PTB (<37 weeks) (RR 0.84, 95% CI 0.78 to 0.90), SGA (RR 0.45, 95% CI 0.30 to 0.66) and birthweight <2500 g) (RR 0.69, 95%CI 0.62 to 0.76). The incidence of perinatal mortality, antepartum hemorrhage, congenital anomalies, and admission to neonatal units resulted similarly between the two procedures. Conversely, a large, randomized trial demonstrated a higher risk of delivering LGA newborns and the development of hypertensive disorders of pregnancy in the eFET group
[17]. Probably, these discoveries suggested that the hyperestrogenism following controlled ovarian stimulation in fresh ET, immediately before embryo implantation, might lead to abnormal endometrial angiogenesis resulting in a reduced implantation and altered placentation. Conversely, hormonal levels in eFET cycle could recreate a more natural uterine environment
[20]. However, the underlying mechanisms suggesting a greater incidence of LGA babies in eFET are still to be clarified. Possible explanations should be a better implantation potential, better placentation, and subsequent fetal overgrowth or epigenetic modifications in the early embryonic stages due to freezing and thawing procedures
[21].
Unfortunately, there is a heterogeneity among studies, which made their comparability difficult in terms of population sampled, design of the studies, freezing methods (slow freezing or vitrification), embryo stage, natural cycles, or hormone replacement used
[22][23]. Furthermore, the results of these meta-analyses were based on observational studies, making them subject to bias.
In contrast with previous Roque M et al.
[9] in a recent meta-analyses analyzing 11 randomized controlled studies including 5379 patients showed no difference in rates preterm birth between fresh ET and eFE a result different from that previously reported
[22][23]. It also showed a significant increase in live birth rates (LBTs) with eFET solely in hyper-responders patients and in pregnancies undergoing PGT-A. Further, this study confirmed the risk of pre-eclampsia was higher with eFET compared to fresh ET, probably due to endometrial priming with supraphysiological concentrations of estrogen during artificial FET cycles. This conclusion is in agreement with the result of a recent Cochrane review, showing a lower prevalence of ovarian hyperstimulation syndrome in eFET cycle despite no difference in the cumulative life birth ratio between the two strategies
[24]. This explains why the transfer of frozen-thawed embryos has become the standard procedure in most fertility clinics. Although this procedure does not seem to reduce IVF success rates, an increased prevalence of PE after eFET technique has been reported
[24][25].
4. Endometrial Preparation
Endometrial preparation for an embryo transfer (e.g., oral estradiol and luteal phase support) may influence the endocrine uterine environment during the embryo transfer, playing an essential role in vascular adaptation of the mother to pregnancy, increasing the risk of placental development and weight of the offspring.
Endometrial preparation before eFET can occur how ovulatory or programmed cycles. To date, in frozen embryo transfer there is no consensus on the best endometrial preparation method or the duration of hormonal replacement
[26]. Emerging data suggests that these differences could have a detrimental impact on adverse obstetrical outcomes in pregnancies from artificial cycles, above all in hypertensive disorders
[19].
In 2019 Saito et al.
[27] evaluated the pregnancy outcomes of 100,000 patients undergoing FET during natural or hormonal replacement cycles. Pregnancies conceived in a hormone replacement cycle had higher odds of hypertensive disorders of pregnancy (4% vs. 3%, aOR 1.43; 95% CI, 1.14–1.80), placenta accreta (0.9% vs. 0.1%, aOR 6.91; 95% CI, 2.87–16.66) cesarean section (44.5% vs. 33.7%, aOR 1.69; 95% CI, 1.55–1.84) and post term delivery associated with a decreased risk to develop gestational diabetes mellitus (1.5% vs. 3.3%, aOR 0.52; 95% CI, 0.40–0.68) in comparison to natural cycle FET.
In agreement with these results, Ginström Ernstad et al.
[28] in a large retrospective study, found an increased risk of hypertensive disorders in pregnancy (10.5 vs. 6.1%, aOR 1.78; 95% CI 1.43–2.21) and postpartum hemorrhage (19.4% vs. 7.9%, aOR 2.63; 95% CI, 2.20–3.13) in hormone replacement cycles when compared to natural cycles. Moreover, higher risks for post-term birth, macrosomia, and cesarean delivery were detected
[28].
Given that endometrial preparation is a less physiological condition than a natural cycle, the increased risk of hypertensive disorders may be due to changes in endometrial receptivity modulating placental development. Moreover, it was hypothesized that in patients who have programmed cycles have a decrease of substances produced by the corpus luteum in early pregnancy, particularly the potent vasodilator relaxin and vascular endothelial growth factor levels, lower angiogenic and nonangiogenic circulatory endothelial progenitor cells and a lack of drop in mean arterial pressure during pregnancy
[29][30]. It was demonstrated that the CL is implicated in the adaptation of the maternal cardiovascular system in early gestation and its absence in eFET may be associated with reduced aortic compliance and increased risk of PE
[29][30]. Anyway, the association between endometrium preparation and adverse obstetric outcomes must be clarified with further studies that include other possible confounders.
Indeed, every single step or procedure carried out during IVF can play an independent and essential role in determining obstetric risks: cryopreservation methods, different hormonal treatment, and laboratory techniques. Vitrification showed higher pregnancy rates than slow-freezing, however perinatal outcomes are similar between the two methods
[31][32]. Potential impact of gamete manipulation, as intracytoplasmic sperm injection (ICSI) and in vitro embryo culture were investigated in recent literature. Specific laboratory procedures, such as incubation systems, types of embryos culture used, the duration of the culture, and ICSI could constitute a source of “stress” for the developing embryo. At last, further large studies are required to identify the contribution of each single confounder on pregnancy and obstetrical outcomes after ART.