r e p r o d u c t i v e b i o l o g y x x x ( ) x x x x x x Available online at ScienceDirect

Transcription

1 Available online at ScienceDirect journal homepage: Review Article Segmented ART The new era in ART? Kemal Ozgur a, *, Peter Humaidan b, Kevin Coetzee a a Antalya IVF, Antalya, Turkey b The Fertility Clinic, Skive Regional Hospital and Faculty of Health, Aarhus University, Aarhus, Denmark a r t i c l e i n f o Article history: Received 21 January 2016 Received in revised form 29 March 2016 Accepted 3 April 2016 Available online xxx Keywords: Iatrogenesis ART GnRHa-trigger Segmented-ART Perinatal a b s t r a c t Currently up to 4% of infants born in developing countries are conceived through assisted reproductive technology (ART). Even though most of these conceptions occur and progress without complications, ART procedures and processes may increase iatrogenesis through complications in - and after conception. We herein review and discuss the clinically and scientific implications and evidence of iatrogenesis, and show how the evolution in ART technologies and procedures has led to the current presumption that frozen embryo transfer might be a more optimal strategy than fresh embryo transfer, in terms of not only reproduction, but also of maternal and fetal outcomes. There is increasing scientific evidence to support the notion that controlled ovarian stimulation could induce significant changes to the endocrine profile of a reproductive cycle, especially to the reproductively important early luteal phase. These changes may not only have a negative effect on implantation and early placentation, but also on the mother, the fetus, and the infant. The overt consequences of controlled ovarian stimulation include ovarian hyperstimulation syndrome, reduced embryo implantation, increased ectopic pregnancy, and altered placentation and fetal growth. The cumulative scientific evidence from this review suggests that GnRHa trigger in segmented ART might constitute the future routine treatment regimen for IVF patients, providing a safe, effective, and patient friendly treatment. # 2016 Society for Biology of Reproduction & the Institute of Animal Reproduction and Food Research of Polish Academy of Sciences in Olsztyn. Published by Elsevier Sp. z o.o. All rights reserved. 1. Introduction The use of ART has increased significantly since its inception, with up to 4.0% of infants currently conceived through ART. The aspirational mission of ART is unquestionably to use 'safe' technologies that deliver healthy infants to infertile couples. Although the vast majority of ART infants are believed to be healthy, some epidemiological studies have identified significant differences between the reproductive outcomes of ART and those of spontaneous conceptions [1,2]. Critically, these differences have the potential to increase perinatal morbidity and postnatal systemic and metabolic disease [3,4]. Definitive research, therefore, is crucially required to identify the molecular and/or cellular mechanisms underlying these adverse outcomes inherent to ART. The moral and ethical aversion to iatrogenic complications, such as, multiple pregnancy, luteal phase insufficiency, * Corresponding author at: Antalya IVF, Özel Antalya Tüp Bebek Merkezi, Halide Edip Cd. No. 7 Kanal Mh., Antalya 07080, Turkey. address: kemalozg@yahoo.com (K. Ozgur) X/# 2016 Society for Biology of Reproduction & the Institute of Animal Reproduction and Food Research of Polish Academy of Sciences in Olsztyn. Published by Elsevier Sp. z o.o. All rights reserved.

2 2 impaired embryo implantation, ovarian hyperstimulation syndrome (OHSS), and perinatal and long-term health issues have been the motivation behind many of the major changes that ART has undergone since its inception. While for most of these complications there were always apparent solutions, the implementation of the solutions have had to wait for evolutions in the medical, surgical, and laboratory technologies of ART to eliminate certain inadequacies. After nearly half-a-century of ART there may be more than a glimmer of hope that many of these solutions may now have become feasible to implement. Ultimately, only ART that is completely free of iatrogenesis has the potential to deliver truly healthy infants to infertile couples. Segmented-IVF has become a feasible treatment option to routine-ivf through major changes in controlled ovarian stimulation (COS), ovulation trigger, and embryo cryopreservation. Optimally implemented it may have the potential to limit a number of the current iatrogenic complications [5]. Currently, only the OHSS-free-clinic concept motivates the implementation of segmented-ivf. However, there are outcomes such as improved implantation, placentation, fetal growth, neonatal and long-term health, and lower ectopic pregnancy rates that are being found to be associated with frozen embryo transfer (FET), that in the future may become the most compelling motivation. If, IVF with fresh embryo transfer (ET) is to be replaced with segmented-ivf, however, it is imperative that there is no increase in treatment-related risks and stress, and that infant health remains paramount [5]. In this review, we discuss the scientific and clinical evidence of IVF treatment iatrogenesis, the evolutions that may limit this iatrogenesis, and whether segmented-ivf with FET is the ultimate solution. 2. ART iatrogenesis 2.1. Controlled ovarian stimulation At this point of time in the history of ART, the collective evidence suggests that COS is responsible, directly and indirectly, for most of the significant iatrogenesis in IVF. Its ubiquitous use in IVF also means that all patients are more or less affected. Conventionally, COS involves the administration of serial doses of exogenous gonadotropins (i.e., follicle stimulating hormone FSH) to induce multi-follicular recruitment and sustain development and the trigger (i.e., human chorionic gonadotropin hcg) of final oocyte maturation at predetermined follicular developmental stages (i.e., follicular size by ultrasound measurement). While COS in most cases may achieve its goal in terms of oocyte number its follow-on consequences have been assumed and in some measure shown to include outcomes such as OHSS, reduced implantation, increased ectopic pregnancy, and increases in adverse perinatal and longterm developmental outcomes. The supraphysiological and irregulated endocrine conditions during the late follicular and early luteal phases are the main reasons for these adverse outcomes, because they result in altered endometrial development and function, and intrauterine conditions that may effect receptivity, implantation and placentation. In addition, the daily and total doses of gonadotropins used may have a significant negative impact, with increasing doses found to be associated with reducing live birth outcomes [6]. In the majority of IVF treatment cycles ovaries contain large numbers of developing follicles at the end of the follicular phase as the result of COS, with the supraphysiological levels of estrogen and progesterone on the day of trigger depending on the actual number of follicles [7,8]. Serum estrogen levels might reach levels 10 times greater than those found during a natural cycle. Cycles with exaggerated responses to ovarian stimulation were assumed to be at greater risk of iatrogenesis, however, recent studies have failed to show any significant independent adverse effect on reproductive outcomes [9 11]. Moreover, increasing and increased serum estrogen levels a function of gonadotropin dose and follicular number were found to be associated with increasing serum progesterone levels ( 1.5 ng/ ml) on the day of trigger [12 14]. Premature progesterone rises occur in 8 40% of cycles, despite the use of gonadotropin releasing hormone (GnRH) analogs to maintain pituitary suppression during COS. Generally, the lower the progesterone level on the day of hcg trigger, the higher the chance of pregnancy [12]. The clinical effect of premature progesterone rise on the day of trigger was confirmed in a meta-analysis on more than cycles, which showed significant reduction in pregnancy rates [14 16]. Progesterone levels, in the presence of estrogen, play a pivotal and determining role in endometrial (i.e., induction of maturation, morphology, activity and ultimately the timing of receptivity) and corpus luteal function, which have a direct impact on pregnancy outcomes [17,18]. In addition to the supraphysiological estrogen and progesterone levels during the late-follicular-early-luteal phase and as the result of the conventional use of a bolus hcg trigger, the early luteal phase is characterized by significantly reduced endogenous LH levels, caused by aberrant hypothalamic pituitary gonadal feedback control [19]. Luteal endogenous LH plays a crucial role in the induction and maintenance of the corpus luteum, the stimulation of implantation promoting factors (i.e., cytokines and growth factors), and, to a lesser extent, in the maturation and function of the endometrium [17 21]. Because of the biological and structural similarities between hcg and LH, a bolus of hcg can be used as a surrogate for the mid-cycle endogenous LH surge. However, the half-life of hcg is significantly longer than that of LH, resulting in sustained luteogenesis (>5 days) [20,21]. In the last decade, the introduction of GnRH antagonist cotreatment has provided the opportunity to use GnRH agonist rather than hcg for final oocyte maturation. The displacement of GnRH antagonist from the pituitary GnRH receptor elicits a surge of LH and FSH similar to that observed in natural cycles, inducing the required oocyte maturation and corpus lutea development [20]. Although, the GnRHa-induced surge of LH was found to be sufficient to secure optimal oocyte maturation, its shorter duration (24 vs h, respectively) [20] as compared to that of natural cycles, resulted in early luteolysis and consequently in high early pregnancy loss rates in fresh ET cycles with standard luteal phase support [19,20]. Using the combination, GnRH antagonist co-treatment, GnRH agonist trigger, and fresh ET cycles, therefore, requires effective modified luteal phase support, including LH activity supplementation to ensure ongoing pregnancies [19].

3 3 Unquestionably, it is the changes to the endocrine regulatory conditions of the pre-ovulatory, post ovulatory, and early luteal phases that may most significantly impact the reproductive outcomes of fresh cycle ET [20 23]. These conditions are critically important, because of their regulatory role in the complex bi-directional communication between embryo and endometrium, i.e., processes and conditions that regulate and promote trophoblast apposition, adhesion, and invasion [3,8]. Sibling donor oocyte (i.e., donor ET versus recipient ET) and fresh ET versus FET comparative studies have served to confirm that COS adversely effects the early luteal phase. While, oocytes and subsequent embryos were similarly produced, embryos were transferred to different intrauterine conditions, with better outcomes in cycles not associated with COS [14,24]. Moreover, in a study comparing the peri-implantation outcomes of fresh ET and FET, with the transfer of matched blastocysts, not only was the implantation rate significantly higher in FET (51.5% versus 40.6%), but so too was the day 14 serum bhcg concentrations (176.2 IU versus IU), per fetal heart implantation [25]. Ultimately, in a recent donor oocyte study no differences were found in the birth weights and gestational ages of consecutive singleton siblings delivered to the same mother after fresh ET and FET, in either order. Importantly, both the consecutive transfers were performed in cycles free of the effects of COS [23]. In addition, recent studies investigating the effects of COS on reproductive outcomes have identified a number changes in the luteal phase that may explain the adverse reproductive outcomes observed in fresh ET. These changes include, altered endometrial gene expression and secretion patterns [26 31], and in particular vascular endothelial growth factor (VEGF) secretion [22], accentuated and advanced endometrial development and function [17,27,31], decreased immune capacity [32], altered intrauterine peristalsis [33], and altered corpus luteum development and function [18]. In summary, the large variety of relatively unpredictable COS-related changes to the pre- and post-trigger endocrine conditions, with follow-on consequences of altered regulation and control of endometrial and placental structures and function, indicate the extent to which the intrauterine conditions of fresh ET might be considered non-physiological. In future with FET as the preferred transfer strategy more deliberate COS protocols can be followed, focusing solely on oocyte number, as all other factors will be of lesser concern. This will maximize the pregnancy potential of an IVF cycle, in terms of the number of viable embryo available for transfer [12] and more physiological conditions at ET Embryo transfer Fresh embryo transfer Multiple pregnancy has been a well publicized iatrogenic complication of IVF from its inception, the result of treatment strategies chosen to overcome low embryo implantation. Multiple pregnancies are associated with significant maternal and perinatal risks (i.e. maternal morbidity and fetal preterm birth, intrauterine growth retardation, and morbidity) and health-related economic costs (i.e. long admission to neonatal intensive care units, long-term health issues). Importantly, only 35% of twin pregnancies are delivered full term, whereas, more than 80% of singleton pregnancies are delivered full term [25]. In the first decades of IVF, more or less 30% of all pregnancies were multiple pregnancies, in contrast to the only 1% in spontaneous conceptions. The low implantation rates of IVF, necessitated the use of multi-embryo transfers ( 2 embryos per transfer) to maximize pregnancy rates. Single embryo transfer (SET), the only natural foil to the multipregnancy problem in IVF [34], however, remained an unrealistic option for at least the first two decades of IVF [35]. In a questionnaire, potential participants of the III World Congress of in vitro Fertilization and Embryo Transfer reported on the outcomes of 7993 transfers performed for the period ending January These transfers resulted in 523 deliveries and 572 ongoing pregnancies, which translates to a reported ongoing pregnancy rate of only 13.7% [36], with majority presumed to be multi-embryo ET. Embryo quality has universally been regarded as the single most important factor in IVF success. Over the last three decades, we have seen a major increase in the understanding of embryo physiology, which resulted in major improvements in the in vitro culture technologies used (i.e., culture media and incubator technologies). These changes better satisfied the metabolic (i.e., sequential or balanced culture media) and physiological (i.e. improved incubator controls and technology) needs of gametes, zygotes and embryos [3], paving the way for extended (i.e., blastocyst) culture [37 39]. In a study by Gardner et al. [40], it was concluded that the transfer of one high scoring blastocyst should lead to a >60% pregnancy rate without the complication of twins, having obtained an implantation rate of 70% and a pregnancy rate of 87% with the transfer of two top-scoring blastocysts. Even though there was increasing evidence of improved implantation rates for blastocyst transfer over cleavage stage embryo transfer, many ART programs were reticent to change [41,42]. The controversy in blastocyst culture and transfer focussed on the concerns of adverse reproductive outcomes (i.e. increases in monozygotic twinning, preterm birth, congenital abnormalities, birth weight) and the high cycle cancelation rate and setup costs [3,43,44]. However, the clinical pregnancy rate for blastocyst transfer determined from the meta-analysis of Papanikolaou et al. [41] was 39.3%, which equates to a 3-fold increase compared to the rate reported in 1984 [36]. Notwithstanding the significant improvement in implantation rates, the pregnancy rates of SET will always remain inferior to those of double embryo transfer (DET) - based simply on the laws of probability. In response, cumulative pregnancy outcomes, combining fresh ET and FET outcomes from a single cycle were proposed to be more equitable measures of comparison and success [45]. In a recent review, the cumulative pregnancy and live birth outcomes of SET were found to be non-significantly different to those of DET [46]. SET as the prime transfer strategy has now been implemented autonomously at many ART centers and under conditional legislation in some countries [45,47]. While, fresh SET may significantly reduce iatrogenesis by eliminating multiple pregnancies and the vanishing-twin phenomenon, some adverse perinatal outcomes still remain in terms of preterm birth, low birth weight, small for gestational age, congenital malformations and cerebral palsy, as compared to the outcomes of spontaneous singleton conceptions [1].

4 Frozen embryo transfer The greater acceptance and implementation of SET as the preferred transfer strategy, however, was dependent on a second wave of technology innovation in the laboratory the development of improved cryopreservation technologies. For much of the first two decades of ART practice the reproductive outcomes of FET were inferior to those of fresh ET, which had relegated FET to a lesser, supportive-only role in most IVF programs [48]. During the last decade significant improvements were made to cryopreservation technologies motivated by the increased need to cryopreserve oocytes/embryos to prevent OHSS, to allow time to perform genetic analyses, to preserve fertility, as well as, to cryopreserve embryos not transferred as the result of SET [48,49]. These improvements resulted in significant improvements in post-thaw embryo survival and developmental competence [24,49]. The increased use of FET is graphically illustrated in the Society for Reproductive Technology (SART) IVF reports for 2003 and The pregnancy rates of FET and fresh ET performed with non-donor oocytes in the year female age group increased from 28.3% (2.7 embryos per transfer) to 35.7% (1.8 embryos per transfer), and 36.9% (2.9 embryos per transfer) to 38.4% (2.1 embryos per transfer), respectively, during the period 2003 to 2011 ( Two embryo cryopreservation methods have been used in ART, slow freeze thaw and vitrification [48]. Initially, the slow freeze thaw method was the method of choice in ART. Recently, with significant improvements in technology, the use of vitrification has gained in popularity [48], demonstrating superior post-warming embryo survival and reproductive competence with similar perinatal outcomes [49 52]. The >90% post-warming survival rate now achievable with vitrification means that the risk of FET cycle cancelation would not be significant [49]. Moreover, similar pregnancy rates can be achieved, using any of the FET cycle protocols (i.e. natural cycle, modified-natural cycle, artificial cycle, and artificial cycle supplemented with GnRH agonist) [53] and more physiological conditions during the early luteal phase of FET may have a profound effect on reproductive outcome [22]. All these factors have challenged the notion that fresh ET should be regarded as the primary transfer strategy in IVF [54]. Four retrospective cohort studies in which cycles were matched according to blastocyst quality to compare the pregnancy outcomes of fresh ET and FET are presented in Table 1. In all four studies, the implantation outcomes of vitrification-warmed blastocyst FET were found to match those of fresh ET [54 57]. Importantly, these studies confirm the positive effect of improvements in cryopreservation technology on the reproductive outcomes of FET, with an average implantation rate of 48% and a high of 59% reported in the study of Roy et al. [56]. Although, there have been concerns regarding the potential toxicity of the high concentrations of cryoprotectants used in vitrification, studies examining FET with vitrified-warmed blastocyst transfers found no increased risk of adverse perinatal and neonatal complications as compared to fresh ET [25,52,58,59]. Fig. 1 is a smoothed graphical depiction of pregnancy outcomes during periods of specific embryo transfer strategies at our ART center. In the period January 2011 to May 2013 >80% of transfers were cleavage stage fresh embryo transfers with Table 1 Implantation outcomes in fresh blastocyst ET compared to vitrified-warmed blastocyst FET. Authors Zhu et al., 2011 [54] Feng et al., 2012 [55] Roy et al., 2014 [56] Ozgur et al., 2015 [57] Study Retrospective cohort Retrospective cohort Retrospective cohort Retrospective cohort Patients Good prognosis Good prognosis, FET cycles previous Good prognosis Good prognosis; 8 28 oocytes collected failed fresh ET 2BB 3BB Grade 1: cavitating blastocyst that contained a discrete compact inner cell mass (ICM) with many trophectoderm cells forming a cohesive layer Blastocysts Good quality blastocysts: clearly observable trophectoderm and inner cell mass. Cryopreservation Vitrification: Cryotop Vitrification: in-house Vitrification: Cryotop Vitrification: Cryotop FET cycle type Natural and artificial cycles Artificial cycles Natural and artificial cycles Artificial cycles Procedure Fresh Frozen Fresh Frozen Fresh Frozen Fresh Frozen N Patient age (yrs) blastocysts/transfer (n) Implantation rate (%) 25.2 (56/222) 37.0 (111/298) S (315/604) (179/384) NS 59.5 (388/652) 58.8 (94/160) NS 40.5 (385/950) 51.5 (306/594) S Survival (%) 85.7 (379/442) 94.8 (384/405) 93.9 (200/213) 89.1% (2096/2353) S, significantly different; NS, non-significant different; ET, embryo transfer; FET, frozen embryo transfer; 3BB, a blastocyst with expansion grade 3, an inner cell grade of B, and a trophectoderm grade of B.

5 5 Fig. 1 Pregnancy rates achieved at a private IVF center from 2011 to illustrating transfer strategy changes leading to segmented-ivf. January 2011 to May 2013 >80% cleavage stage fresh ET; May 2013 to March 2014 >80% blastocyst fresh ET; March 2014 to September 2015 >80% vitrified-warmed blastocyst ET. 4 embryos per transfer. May 2013 to March 2014 >80% of transfers were fresh blastocyst transfers; initially 80% were DET and at the end of the period 65% were SET. March 2014 to September 2015 >80% of transfers were from blastocyst freezeall IVF cycles; initially 80% were DET and at the end of the period 60% were SET. In early 2012 cabinet incubators (HeraCell 150, Heraeus, Thermo Fisher Scientific Inc, USA) were replaced with mini-incubation-systems (K-Systems, Kivex Biotec Ltd, Denmark), with embryo culture conditions changed from 20 to 5% O 2. This figure illustrates the improvements in pregnancy going from a strategy of cleavage stage fresh ET to a strategy of blastocysts freeze-all with FET. In summary, while research and innovation have resulted in significant improvement in embryo competence (i.e., improved metabolic and genetic competence), continued research is essential. Whether blastocyst culture becomes universally accepted may not rely only on reported reproductive outcome improvements, but also on the economics of blastocyst culture set-up and the expertise in the management of extended culture. However, transferring embryos at the blastocyst stage, because of their post-embryonic genome activation status and increased euploidy may currently be the most effective strategy for SET. The increasing reports of higher pregnancy rates in FET, as the result of better embryo endometrium synchrony and the more physiological intrauterine conditions at transfer, may change the way IVF will be practiced in the future. However, whether FET becomes the universally accepted way of ET in IVF will depend on epidemiological studies continuing to show superior perinatal, neonatal, and long-term developmental outcomes for FET and epigenetic studies showing no increase in genetic modification as the result of gamete and embryo manipulations Peri-implantation conditions Reduced embryo implantation, a disappointing iatrogenesis of IVF, was in the past the consequence of both poor embryo competence and the luteal phase defect of fresh ET. Currently, with the significant improvements in in vitro culture and consequently in in vitro embryo competence, the luteal phase defect and reduced endometrial receptivity bear the greater responsibility for reduced implantation rates in IVF with fresh ET. Moreover, during the first 20 years, IVF embryo transfers were performed on days 1 to 3 (i.e., cleavage stage embryos), although, in synchrony with the endometrium the premature transfer of cleavage stage embryos to the uterus resulted in embryos being exposed to alien physio-metabolic conditions. This strategy was imposed upon IVF by the poor in vitro culture conditions of the time, which did not allow for optimal blastocyst development. Implantation, therefore, might have been affected by embryonic stress and the length of time embryos had to remain within the intrauterine cavity before initiating hatching and implantation. In addition, factors underlying the luteal phase defect of IVF such as altered uterine contractility and shifts in time of optimal endometrial receptivity, may have significantly contributed to reduced implantation rates. The human endometrium is a highly specialized, hormonally regulated organ which is only receptive to embryo interaction for a self-limiting period known as the window of implantation (WOI) [31,47]. During this period, the endometrial epithelium acquires optimal receptivity by progressing through specific structural, functional, and morphological changes induced by pre-ovulatory estrogen and progesterone and post-ovulatory progesterone, LH, and bhcg. While the WOI may vary in timing and duration from patient to patient [60], it is generally believed to occur on LH + 7 to 11 days, with most clinical implantations occuring on LH + 9 days. In natural cycles, synchrony between the developing embryo and endometrium is achieved through complex bi-directional communications [61]. In IVF, however, the supraphysiological estrogen and progesterone levels prematurely prepare endometria for maturation [17] and the final oocyte maturation trigger induces the rapid and premature maturation of endometria, with the majority (>90%) reaching an advanced stage of maturation (i.e., by 2 to 4 days) by the time of oocyte collection [27,31]. Endometria that are advanced by >3 days at oocyte collection may essentially have non-receptive

6 6 endometria at the time of embryo transfer [62]. Importantly, animal-model studies have demonstrated that embryos implanting beyond the normal WOI results in defective placentation and fetal growth [8]. The large array of coincidental changes (i.e. to endocrine, metabolite, electrolyte, and endometrial systems) that are brought about by COS and ovulation induction have the potential to affect the reproductive potential of all IVF cycles to some degree, are not present in FET, especially natural cycle FET. Critically, however, a competent blastocyst and a receptive endometrium do not guarantee successful implantation, as developmental synchrony may also significantly affect reproductive outcome. The regulation of synchrony between embryo and endometrium has not been fully elucidated in the human, however, up to 30% of fresh ET [60] are affected by endometrial recepetivity and synchrony. If a fresh ET strategy is preferentially followed the reproductive outcomes will remain affected, as currently there is no noninvasive way to determine the time of optimal receptivity in an IVF cycle. Implantation is a complex sequential process that requires specific molecular mediators during each of the three stages of implantation, i.e. recognition, adhesion, and invasion. These mediators are secreted in a coordinated fashion by the blastocyst, the endometrium and other cells (i.e. leucocytes, natural killer cells) and include cell adhesion molecules, cytokines, growth factors, and prostaglandins. The contents of the viscous fluid secreted into the uterine lumen reflect the hormone regulated endometrial state and function of a cycle. Proteomic analyses have shown that the secretomes in COS cycles are significantly different to those of natural cycles when aspirated at comparative times (LH + 7 and trigger + 7) [64,65]. In addition, new microarray technology has shown that endometrial gene expression is also different, with >1000 genes differentially expressed in natural and IVF endometria [27,28,30]. The gene expression profiles in IVF cycles (trigger + 7) were found to be more akin to those associated with nonreceptive endometria of natural cycles [27]. While compensatory or counter-acting measures have been explored [19], the luteal phase and more specifically the early luteal phase of IVF cycles remains non-physiological, requiring effective exogenous luteal phase support to ensure ongoing pregnancy. However, there are promising reports that the administration of low dose metformin between first monitoring and ovulation induction might prevent premature luteinization and advanced endometrial maturation [66]. On the other hand, recent developments in microarray technology have allowed this technology to be used to determine the optimal day of endometrial receptivity. The Endometrial Receptivity Array (ERA), using the transcriptomic signature of 238 genes to identify the timing of endometrial receptivity, has been purported to be superior in accuracy to any histological assessment [30]. In the future, ERA might be routinely applied to increase implantation rates in human IVF [60]. In summary, improvements in ART technologies and procedures have led to significant improvement in embryo competence, for both fresh in vitro cultured embryos and cryopreserved embryos. The implementation of successful blastocyst culture and cryopreservation programs, maximize, not only the reproductive outcomes of a single embryo transfer procedure, but also the reproductive potential of a single treatment cycle. However, with no effective means to pharmacologically correct the defective luteal phase and no non-invasive means to determine the day of optimal receptivity of fresh ET, a subsequent FET offers the opportunity to effectively achieve both, i.e., a timed transfer in a more physiological environment. ERA in combination with FET may in the future provide an individualized treatment approach with optimized reproductive outcomes Ectopic pregnancy Ectopic pregnancy is another acknowledged iatrogenic complication of IVF, with rates observed in women undergoing IVF with fresh ET ranging between 2% and 11% [67,68]. The ectopic rates of IVF are generally considered to be double those of spontaneous conceptions. The reproductive health characteristics (i.e. increased rates of tubal disease) of IVF patients might contribute significantly to these higher rates [69]. Theoretically, however, the outcomes should be lower in IVF, as the IVF process circumvents the role of the fallopian tubes in the reproductive process by transferring embryos directly into the uterine cavity. Therefore, the increased risk of ectopic pregnancy must be the consequence of the intrauterine conditions encountered by the embryo at the time of transfer. In this respect, intrauterine peristalsis, which plays an important role in the success of implantation, is significantly affected by COS. Thus, the intrauterine peristaltic wave frequency and the wave types at the time of fresh ET differ significantly from those observed in post-ovulation natural cycles [33]. In particular, the wave direction at the time of fresh ET is predominantly from the cervix to the fundus, which may facilitate the migration of transferred embryos toward the tubal ostia. Recently, the increased levels of ectopic pregnancy in IVF with fresh ET have been demonstrated in studies that compared reproductive outcomes of fresh ET with FET [67,68]. In a large population-based study, the ectopic pregnancy rate declined according to the following transfer strategies, from 1.9% in fresh cleavage embryo ET, to 1.7% in cleavage embryo FET, 1.3% in fresh blastocyst ET, and 0.8% in blastocyst FET [68]. In a large single-center study, the ectopic pregnancy rate of fresh blastocyst stage ET was 1.8%, with blastocyst FET again being lower at 0.32% [67]. Moreover, in a single-center study the ectopic pregnancy rate was 1.55% for fresh blastocyst ET and 0% for blastocyst FET in cycles with good grade blastocysts (grade 5) transferred [25,40]. These outcomes indicate that at least two factors might increase the risk of ectopic pregnancy in IVF; the time taken for the embryo transferred to hatch and the COS-induced non-physiological peri-implantation conditions. In summary, recent studies suggest that ectopic pregnancy rates are reduced in accordance with the following embryo transfer strategies: SET, blastocyst transfer, and FET [68]. The lower risk for ectopic pregnancy in FET cycles significantly adds to the evidence of the adverse peri-implantation conditions of fresh ET. The lower risk seen in blastocyst transfers provides evidence to support the use of extended culture. Ultimately, combining the strengths of both strategies

7 7 in IVF, i.e., FET and extended culture, means that ectopic pregnancy rates would essentially be the same as those after spontaneous conceptions Ovarian hyperstimulation syndrome OHSS is the major iatrogenic complication of COS. Severe OHSS holds significant health-related economical costs, emotional and physical costs, and ultimately may lead to patient death. Moderate to severe OHSS occurs in 0.2% to 2% of IVF cycles, however, in patients with OHSS risk factors the condition may occur in up to 30% of IVF cycles [70]. OHSS is symptomatic of increased intravascular permeability, leading to possible organ dysfunction, respiratory impairment, hemorrhaging, thromboembolism, and oliguria [71,72]. In general, exogenous hcg (i.e. trigger for ovulation induction) induces early onset OHSS, while endogenous hcg (i.e. produced by embryogenesis of pregnancy) induces late onset OHSS. The prolonged half-life of hcg not only increases, but also prolongs the luteotropic activity of the multiple corpora lutea, significantly increasing VEGF levels and consequently intravascular permeability and abdominal fluid accumulation [71 73]. The prophylactic prevention of OHSS in IVF is complicated by its unpredictability, as at present there are no markers that can predict OHSS with certainty. The prediction of OHSS is further complicated by unknown patient predispositions and whether or not pregnancy occurs. In case of a presumed high OHSS risk the cycle can be canceled prior to triggering ovulation induction with an hcg trigger [72]. While this strategy may prevent OHSS, such a strategy has immense psychological and economic implications for the patient. The alternate strategy is oocyte or embryo cohort cryopreservation. This strategy, however, still requires a final oocyte maturation trigger, which means that OHSS might not be completely eliminated [5,73]. The introduction of GnRH antagonist into IVF treatment offered new hope in the battle against OHSS, because of its different pharmacological mode of pituitary suppression. In this respect, GnRH antagonist co-treatment during COS allows multiple follicular development to occur with less side-effects (i.e. hypo-estrogenesis), shorter stimulation durations, and lower total gonadotropin usage [72,74,75], therefore, ideally suited for high responder patients. Most importantly, GnRH antagonist co-treatment provides the option to use GnRH agonist as final oocyte maturation trigger. GnRH agonist triggers a more physiological hormonal surge, including a surge of both LH and FSH, which results in lower luteotropic activity and, therefore, a lower risk of OHSS, even for the high risk patient [76,77]. Successful GnRH antagonist co-treatment, GnRH agonist trigger, and fresh ET IVF cycles, however, require concerted luteal phase support strategies to ensure ongoing pregnancies [70,78,79]. Low dose (1500 IU) hcg supplementation, given on the day of oocyte collection, was shown to overcome the induced luteal phase defect of GnRHa trigger [78,80,81]. While the combination of GnRH agonist plus low dose hcg may significantly reduce chances for severe OHSS, this complication may still occur in high-risk patients that have a fresh ET [73,82]. If, however, the fresh embryo transfer is deferred (i.e., cohort cryopreservation), luteal hcg is not required, reducing the chances of OHSS to zero even in high-risk patients. In summary, it is clear that using conventional COS protocols in IVF with fresh ET, many patients have to receive proactive clinical management to prevent or reduce OHSS risks and complications. The provision of safe IVF treatment means no OHSS and, therefore, certainly paradigm shifts in pituitary suppression and ovulation induction, and possibly in embryo transfer policy. Whether GnRH antagonist co-treatment COS, GnRH agonist trigger, and FET becomes the gold standard in IVF treatment depends on the technologies surrounding embryo culture and cryopreservation. If all fresh embryo transfers were to be deferred to FET, the focus of clinical management in COS would be solely on obtaining sufficient oocyte numbers Perinatal and long-term health outcome With IVF treatment gaining a high degree of respectability over the last two decades, the focus of attention has now to be turned toward the iatrogenic complications of perinatal, postnatal and birth defect outcomes that are also associated with IVF [1,85]. Importantly, perinatal studies have found that poor outcomes occur both in multi-gestational and in singleton IVF pregnancies. In this respect, perinatal studies have reported higher rates of low birth weight (LBW), preterm birth, perinatal mortality, placental complications (i.e., preeclampsia, placenta previa, and placenta abruptio), and birth defects in pregnancies from IVF with fresh ET compared to fertile controls with spontaneous conceptions, with outcomes persisting even after adjusting for maternal confounders [3,22,83 85]. Importantly, the low birth weight prevalence from IVF pregnancies may be a significant marker of intrauterine stress, which has been linked to cardio-metabolic illness in adult life [2,3]. The precise molecular and cellular mechanisms involved in the observed adverse perinatal outcomes of IVF have not been fully resolved, neither have the most significant contributing factors (i.e. procedural and/or maternal) been identified [1,2,69]. The contribution of patient characteristics to reproductive outcomes in IVF has often been of unknown significance in the comparison made with reproductive outcomes of spontaneous conceptions. However, in a study comparing birth weight between fresh IVF, ICSI, GIFT (gamete intrafallopian tube transfer), and FET, it was found that GIFT results were similar to those of fresh ET from IVF and ICSI, suggesting that altered birthweight outcomes may be less related to in vitro embryo culture factors and more to factors associated with ovarian stimulation and oocyte collection [86]. Similarly, a donor oocyte study, investigating birth weight of consecutive fresh ET and FET singleton siblings deliveries to the same recipient mother, reported that neither the freezing method used (i.e., slow freezing or vitrification), the genetics of the biologic mother, the stage of embryo development, nor the cryo-storage period influenced the outcomes [23]. The absence of COS in the recipient mothers indicate that COS, i.e., high estrogen levels, might be a significant determining factor in infant outcomes [1,3,8,23]. In animal reproduction models, the endocrine conditions induced by COS were found to directly affect implantation, placentation, and fetal growth through significant morphological changes to endometria and the placentae, and their

8 8 function [8,22]. In the human model, a covariate (covariates: gestation, sex, parity, socio-economic status, perinatal mortality, birth defects, and Cesarean section) analysis showed that transferring embryos in fresh cycles was the primary factor affecting birth weight [86]. Conditions that prevent sufficient invasion of the uterine epithelium and vascular development will affect fetal growth and development even with late gestation compensatory placental developments observed in in vitro conceptions [3,63]. These alterations to placental/fetal development increase the risk for fetal demise or cardio-metabolic disease in adult life of the infant. Evidence suggests that IVF procedures alter normal fetal growth patterns overtly seen as fetal growth restriction in early to mid-pregnancy, followed by significant increases in placental weight and accelerated fetal growth toward the end of gestation [3,63]. These placental and fetal growth changes are believed to be compensatory to poor placentation, as the result of the non-physiologic early luteal phase conditions generated by COS [22,25,26]. Notwithstanding this late accelerated fetal growth, the birth weights of IVF infants are lower in comparison to FET and spontaneous conceptions. In addition, COS induced epigenetic alterations in oocyte maturation, embryogenesis, and endometrial maturation may significantly impact fetal growth, development, and health [22,87]. The renewed interest in the use of FET in ART and its introduction to mainstream treatment has spurred numerous investigations into the perinatal, neonatal, and postnatal health of FET infants [88], mainly to establish the safety of cryopreservation. In general, FET singletons have lower relative risks for adverse perinatal and birth defect outcomes, compared to fresh ET singletons [1,83,85 87], however, some still remain higher in comparison to spontaneous singletons [1,84,87,89]. Nonetheless, it is important to recognize that there have been inconsistencies in the reporting of relative perinatal outcomes. These inconsistencies may be attributable to differences in embryo stage at cryopreservation, embryo stage at ET, methods used for cryopreservation, modes of endometrial preparation, and patient characteristics (age, parity, infertility duration) [1]. The perinatal outcomes that have consistently been reported to be at lower risk in FET singletons compared to fresh ET singletons are: preterm birth [1,52,87,89,90], LBW [1,52,84,86 90], small for gestational age (SGA) [52,86,90]. In contrast, the perinatal outcomes that have been reported to be of higher risk in FET singletons are, perinatal mortality [84,86], large for gestational age (LGA), and macrosomia [1,52,84,86,87,90]. What is immediately evident from the perinatal studies of fresh ET and FET is that adverse outcomes associated with fetal growth (i.e. LBW, SGA, LGA, macrosomia) predominate. The consequences of these adverse growth dependent outcomes warrant further investigation, especially as the LGA and macrosomia rates observed in FET singletons are higher even than those observed in spontaneous singletons [1,86,87]. Controversially, LBW outcomes were found not to be different between sibling oocyte donation fresh ET and FET [91], while the LGA outcomes were significantly different between sibling fresh ET and FET infants to the same parents [87,92]. Underlying this contradiction may be the unknown effect of endometrial preparation protocols in FET, requiring further investigation [84,86]. In Table 2, four recent perinatal studies are presented in which blastocysts were transferred either in fresh ET or FET, illustrating not only between study similarities in live birth outcomes, but also in birth weight and very preterm delivery outcomes. Birth weight and preterm delivery are important outcomes in determining the success of treatment, due to their strong relationship with perinatal mortality and morbidity. Importantly, all four studies reported approximately g higher birth weights for FET infants. Importantly, only 10% of pregnancies conceived from both fresh ET and from FET resulted in deliveries with birth weights lower than 2500 g (Table 2). In summary, treating infertility with IVF may predispose pregnancies to adverse perinatal outcomes and increased risks for birth defects. On the other hand, singletons from FET treatment have better perinatal outcomes, and no increased risk for birth defects compared with singletons born after fresh ET. This is of particular importance, because perinatal abnormalities may have a direct bearing on adult cardiometabolic and respiratory health. However, epigenetic disturbances are generated in procedures leading to both transfer strategies, which requires further investigation to help identify the underlying mechanisms. 3. Segmented-IVF the future? At this point-in-time in the evolution of ART, the cumulative scientific evidence is starting to support the notion that the best chance of delivering a singleton healthy baby at term with the lowest chance of long-term health issues may be best achieved through the use of FET. Technology innovations in cryopreservation have seen the reproductive outcomes of FET improve significantly with live births of FET now equal to those of fresh ET (Table 1). The historical arguments for the preferred transfer of embryos in fresh ET, therefore, may no longer be valid (Table 3). Evidence suggests that currently the cumulative iatrogenesis of fresh ET outweighs that of FET. Moreover, by combining ovarian stimulation with GnRH antagonist cotreatment, GnRH agonist trigger, blastocyst culture, and SET- FET, the risks for iatrogenesis may be reduced to the lowest level possible, while maintaining good live birth rates. These elements may be combined into a single strategy, which has become known as segmented-ivf. Segmented-IVF consists of four primary segments, (1) controlled ovarian stimulation with GnRH antagonist co-treatment and GnRH agonist trigger, (2) embryo culture and cryopreservation (3) FET cycle programming, and (4) frozen-thawed SET. Segmented-IVF will simplify certain aspects of COS management (i.e., start date of stimulation, premature progesterone rise), will allow for flexibility in triggering, and, therefore, in the scheduling of oocyte retrievals (i.e., the avoidance of weekends). Segmented-IVF will allow the inclusion of other procedures that may serve to enhance reproductive outcomes, procedures which would adversely affect reproductive outcomes if performed in conventional IVF. These 'outcome-benefitting' procedures include blastocyst biopsy for comprehensive genetic screening and analysis [93], endometrial receptivity tests to determine the optimal day of transfer [94], endometrial procedures to induce

9 9 Table 2 Obstetric outcomes in fresh ET compared to FET. Authors Wikland et al [58] Ishihara et al [90] Li et al [52] Ozgur at al [25] Retrospective cohort; single center, Turkey, January 2012 to December 2013 Retrospective population; Australian and New Zealand Assisted Reproduction Database (ANZARD), January 2009 and December 2011 Retrospective population; Japanese ART registry database, Study design Retrospective cohort; single center, Scandinavia, January 2006 to May 2008 Embryo stage Blastocyst Blastocyst Blastocyst Blastocyst Fresh Frozen Fresh Frozen Fresh Frozen Fresh Frozen Patient age (yrs) Gestational age (weeks) Live birth per transfer (%) 36.3 (203/563) 32.1 (103/321) 20.5 (6883/33,559) 26.0 (30,952/118,866) 27.8 (13,049/46,890) 24.8 (4955/19,978) 36.3 (203/563) 56.8 (217/382) Live birth per pregnancy (%) 68.6 (203/296) 60.0 (103/172) 74.0 (6883/9302) 72.1 (30,952/42,942) 77.5 (13,049/16,845) 75.8 (4955/6537) 83.4 (171/205) 85.7 (114/133) 0.0 (0/199) 1.0 (1/103) 0.70 (42/5981) 0.72 (196/27,408) 2.6 (317/12,241) 2.0 (93/4721) 3.4 (6/176) 2.6 (3/116) Very preterm delivery, <32 weeks (%) Birthweights, 2500 (%) 95.5 (190/199) 93.2 (96/103) 87.3 (5223/5981) 91.0 (24,940/27,408) 90.9 (11,121/12,241) 93.8 (4426/4721) 93.0 (159/171) 93.0 (106/114) Birthweights (g) Weight difference (g) implantation-promoting immune response [95,96], and hysteroscopic or laparoscopic surgical procedures to correct intrauterine anomalies [97]. However, implementing segmented-ivf as the routine IVF treatment may increase costs and the relative risks of treatment. The cost of segmented-ivf may depend on individual clinics policies, procedures, and operating environment. The main differences in costs between segmented and routine IVF are the costs of endometrial preparation/fet cycle monitoring (i.e., ultrasound scans and blood tests), cryopreservation (including storage and freeze thaw procedures) of blastocyst(s), and patient clinic communication. Implementing routine segmented-ivf may also require a revision of how the key performance indicators of the laboratory are managed, as FET pregnancy outcomes are disconnected in time from the in vitro culture conditions that produced the cryopreserved embryos or blastocysts. As with blastocyst culture, a major risk of segmented-ivf is cycle cancelation, due the loss of embryo competence and/or embryo degeneration. Many informed infertile patients might have strongly perceived concerns about segmented-ivf, which include the risk of cycle cancelation, the time between oocyte collection and embryo transfer, and whether FET is as effective as fresh ET [98]. Results, however, are showing that the risk of cycle cancelation may be low, as >95% of blastocysts survive warming, and in a freeze-all situation equates to a <1% cycle cancelation rate. While natural cycles are most patientfriendly, artificial cycles provide a relatively rigid schedule, with the benefit of being able to provide patients with the date of their embryo transfer with some certainty [99]. Providing patients on the day of oocyte collection with their schedule to the planned day of embryo transfer and providing messages and calls on highlighted dates will help to manage patient stress and anxiety. Because patients overriding expectation from treatment is pregnancy, patients need to be fully informed about the clinics success rates with fresh ET and FET as well as about the perinatal and possible long-term benefits of FET. Moreover, it has been shown that patients may be in a better psychological state-of-mind at the time of FET than at the time of fresh ET [100]. Thus, the counseling role of the clinician is vital in the decision-making processes of patients in respect to accepting segmented-ivf. The primary treatment goal for infertile patients is the delivery of high term live birth rates per started treatment. The secondary, but no less important goal is providing patients with a safe and individualized treatment. The dire and possible lifelong consequences of OHSS and ectopic pregnancy cannot be overemphasized. The GnRH antagonist protocol with GnRH agonist trigger for final oocyte maturation, therefore, has become the backbone of our ART programs. This protocol provides a simple, but versatile platform for ovarian stimulation. Furthermore, based on our experiences, our reported results, and the support of results from other published works, including comparisons with spontaneous pregnancies, have motivated us to move from a standard IVF program to a segmented-ivf program, i.e., fresh ET to FET. The reported results provide evidence that the closer to natural hormonal conditions and endometrial development observed in FET might be the main reasons for the improved reproductive outcomes observed. The inability to determine the time of