Is it time to reconsider how to manage oocytes affected by smooth endoplasmic reticulum aggregates?

Abstract

STUDY QUESTION

Did the revised Alpha/ESHRE consensus (Vienna, 2017) bring a real answer on managing oocytes with aggregates of smooth endoplasmic reticulum (SERa)?

SUMMARY ANSWER

According to the currently available literature, a case by case approach on the time of injecting/inseminating SERa+ oocytes may be not helpful for embryologists making a decision, so we suggest fertilizing both SERa+ and SERa− oocytes and prioritizing embryos derived from SERa− oocytes.

WHAT IS KNOWN ALREADY?

In 2011, the Istanbul consensus recommended not to inject/inseminate SER+ oocytes due to adverse foetal outcomes reported in literature. At the end of 2017, a panel of experts reconsidered this recommendation and advised a case by case approach. Hence, with a lack of clear recommendations, in-vitro fertilization practitioners still have heterogeneous attitudes when managing SERa+ oocytes. In this context of controversy, an updated review could be helpful in (i) forming a common language for managing cases of SERa+ oocytes and (ii) offering the most ethical practice and best care for patients seeking infertility treatment or fertility preservation.

STUDY DESIGN, SIZE, DURATION

This review (with a last literature search on 1 June 2018) evaluated the effect of the SER dysmorphism on embryological and neonatal outcomes.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Studies were considered for inclusion if they were prospective or retrospective cohort or case–control studies. Electronic searches of the Pubmed and Embase databases were done using the keyword combination: smooth endoplasmic reticulum, SER, oocyte and zygote. Abstracts and articles written in English and limited to humans were included.

MAIN RESULTS AND THE ROLE OF CHANCE

The search returned a total of 726 studies among which 21 met the inclusion criteria. The literature does not unanimously support a negative association between SERa and embryogenesis, implantation or assisted reproductive therapy outcomes. The reviewed studies reported 112 neonatal outcomes after transfers where at least one embryo originated from oocyte affected by SERa. They included 101 healthy babies, three live births with malformations, three neonatal deaths, one stillbirth and four medical interruptions of pregnancy. After transfer of embryos exclusively derived from SERa+ oocytes, a total of 48 healthy newborns were reported along with four babies with perinatal complications (including one ventricular septal defect), one stillbirth, one neonatal death and one pregnancy termination for multiple malformations.

LIMITATIONS, REASONS FOR CAUTION

As with any review, this review was limited by the quality of the included studies especially in terms of possible methodological limitations, the limited sample size and the retrospective aspect of the studies. Among the 21 selected studies, seven were abstracts and two were case reports. Of the remaining 14 studies, only three were prospective. The tools used in identifying SERa+ oocytes may have varied from one study to another and a consequent misclassification cannot be excluded. Considering the poor resolution of light microscopy in detecting SER aggregates, we are not sure that apparently SERa− oocytes do not really exhibit such a dysmorphism if they were analysed under electronic microscopy or a time lapse system.

WIDER IMPLICATIONS OF THE FINDINGS

In the light of the existing data and the lack of a real link between fertilizing SERa+ oocytes and the occurrence of embryo aneuploidy/malformations, we think that discarding SERa+ oocytes may be not the most ethical approach even in patients with large cohorts on the day of oocyte retrieval. Avoiding the wastage of oocytes and embryos with respect to medical ethics remains a constant concern in daily IVF practice. Thus, we recommend that all mature oocytes could be fertilized and embryos originating from SERa− oocytes would be preferably transferred, even if they come from a cohort with SERa+ oocytes. The remaining embryos derived from SERa+ oocytes could be considered with a lower priority for transfer after obtaining consent from the couple if a strict follow-up of the pregnancy and the baby is performed.

STUDY FUNDING/COMPETING INTEREST(S)

We have no conflict of interest to declare and no funding was received.

REGISTRATION NUMBER

N/A.

smooth endoplasmic reticulum aggregates, oocytes, clinical outcome, perinatal complications, healthy baby

Introduction

In 2011, based on data reporting a higher risk of malformations in babies originating from oocytes with cytoplasmic smooth endoplasmic reticulum aggregates (SERa) (Otsuki et al., 2004; Ebner et al., 2008; Akarsu et al., 2009; Sá et al., 2011), the Istanbul Consensus Workshop advised against their use in IVF (Alpha Scientists in Reproductive Medicine and ESHRE Special Interest Group of Embryology, 2011). This consensus recommended discarding these oocytes at the time of fertilization. SER is a type of organelle appearing as round flat disks in the oocyte cytoplasm corresponding to large tubular SER clusters surrounded by mitochondria (Sá et al., 2011). One of the key roles of SER is calcium storage and release, which is needed for cell activation during fertilization. Moreover, complexes of endoplasmic reticulum in association with mitochondria play a crucial role in energy accumulation, protein and lipid production and synthesis of nuclear membranes throughout early embryo development (Carroll et al., 1996; Sá et al., 2011).

It has been reported that SERAs could have an impact on both embryological and clinical outcomes in ART attempts (Shaw-Jackson et al. 2014). Indeed, the presence of SERa has been associated with (i) lower oocyte maturation (Setti et al., 2016) and fertilization rates (Sá et al., 2011), (ii) lower embryo quality (Ebner et al., 2008; Sá et al., 2011; Braga et al., 2013), (iii) lower implantation and pregnancy rates (Otsuki et al., 2004; Setti et al., 2016) and (iv) increased miscarriage rates (Otsuki et al., 2004; Ebner et al., 2008; Braga et al., 2013) compared to SERa− oocytes. Higher rates of perinatal complications, birth defects and imprinting disorders have also been reported (Otsuki et al., 2004; Ebner et al., 2008; Akarsu et al., 2009; Sá et al., 2011; Mateizel et al., 2013). However, these data remain controversial and the question of discarding or not SERa+ oocytes is still open for debate especially as some studies have reported the birth of healthy babies derived from SERa+ cycles (with at least one oocyte SER+ among the cohort) and even from SERa+ oocytes (Mateizel et al., 2013; Hattori et al., 2014). Historically, Otsuki described three forms of SER clusters in positive oocytes, classified on basis of size: large (18 μm), medium (10–17 μm) and small (2–9 μm) clusters (Otsuki et al., 2004). Contrary to large and medium aggregates, visible by light microscopy, small aggregates of SER can only be assessed by electron microscopy.

The conflicting data regarding the effect of SERa on IVF outcomes have mostly originated from studies with low levels of proof and have led to heterogeneous attitudes regarding SERa+ oocytes fertilization, with only 14% of centres discarding SERa+ oocytes (Restelli et al., 2015). Therefore a revision of the Alpha/ESHRE consensus was conducted by a special interest group of embryologists in the Vienna Consensus in 2017 (ESHRE Special Interest Group of Embryology and Alpha Scientists in Reproductive Medicine, 2017). Hence, a case by case approach concerning SERa+ oocytes fertilization was recommended.

In order to evaluate whether the Alpha/ESHRE consensus of Vienna 2017 brought a real answer to the question of how to deal with SER+ oocytes and cycles, we performed a review of published studies to analyse existing data about the impact of SERa+ cycles and oocytes on clinical and neonatal outcomes.

Materials and Methods

Pubmed and Embase were searched for the most recent studies and reviews using the keyword combination: ‘Smooth Endoplasmic Reticulum’, ‘SER’, ‘oocyte’ and ‘zygote’, with the last search performed on 1 June 2018.

The literature search concerned articles or abstracts written in English on studies in humans and published in peer-reviewed journals. Articles selected for the review were then included if they reported data on the impact of SER on at least one of the following outcomes: fertilization rate, pregnancy rate, miscarriage rate, implantation rate, embryo quality and neonatal outcomes (Fig. 1). Papers trying to identify the potential predictive factors of the occurrence of SER in oocytes were also analysed.

Figure 1

Flow chart for the literature review.

Results

Design and methodology of the studies

The initial search yielded to 726 records. After selection, a total of 21 studies met the inclusion criteria, comprising seven abstracts (Bielanska and Leveille, 2011; Miwa et al., 2013; De Gheselle et al., 2014; Maldonado Rosas et al., 2015; Oudshoorn-Roessen et al., 2015; Carvalho et al., 2016; Mizoguchi et al., 2016), 12 studies and 2 case reports (Akarsu et al., 2009; Sfontouris et al., 2018) (Fig. 1). Among the 21 studies, 17 were retrospective and only 4 studies were prospective.

Table I shows the prevalence of cycles where at least one oocyte is SERa+ and SERa+ oocytes in one or more cycles. One study focused on the frequency of SERa+ per total embryos (375/5516) or per total patients (53/510) (Braga et al., 2013) and two studies assessed SERa+ oocyte frequency per total MII oocytes (Carvalho et al., 2016, Mizoguchi et al., 2016) and showed a frequency of 0.7% (46/6756) and 4.5% (128/2815), respectively. The occurrence of SERa+ cycles was variable, ranging from 4% (Maldonado Rosas et al., 2015) to 23.1% (Otsuki et al., 2018). Among SERa+ cycles, the frequency of SERa+ oocytes (defined by number of SERa+ oocytes/total number of oocytes) ranged from 17.6 (Mateizel et al., 2013) to 34.4% (Otsuki et al., 2004).

Table I

Smooth endoplasmic reticulum aggregate: description of studies and prevalence of SERa+ cycles and SERa+ oocytes.

Study		Study design		Prevalence
			SERa+cycles (n)	SERa+ cycles (%)	SERa+ oocytes (%)
Otsuki et al. (2004)	Ar	RS	18	9.4	34.4
Ebner et al. (2008)	Ar	PS	30	6.2
Akarsu et al. (2009)	Ar	CR	3
Sá et al. (2011)	Ar	RS	60	8.3	18.8
Mateizel et al. (2013)	Ar	RS	394	5.4	17.6
Braga et al. (2013)	Ar	RS	53 patients
Hattori et al. (2014)	Ar	RS	252	7	20.7
Restelli et al. (2015)	Ar	RS	130	12
Itoi et al. (2016)	Ar	RS	51	8.8
Shaw-Jackson et al. (2016)	Ar	PS	112	13.6	29.1
Setti et al. (2016)	Ar	PS	78	10.6
Itoi et al. (2017)	Ar	RS	242	15.5
Otsuki et al. (2018)	Ar	RS	43	23.1	23.1
Sfontouris et al. (2018)	Ar	CR	1
Bielanska and Leveille (2011)	Ab	RS	121		19.3
Miwa et al. (2013)a	Ab	RS	86	7.3
De Gheselle et al. (2014)	Ab	RS
Maldonado Rosas et al. (2015)	Ab	RS	131	4
Oudshoorn-Roessen et al. (2015)	Ab	PS	56		19.8
Carvalho et al. (2016)	Ab	RS	82	11.3	128/2815b
Mizoguchi et al. (2016)	Ab	RS			46/6765b

Ar = Article, Ab = Abstract, CR = Case Report, RS = Retrospective Study, PS = Prospective Study.

SERa+ cycles (%) = n SERa+ cycles/n total cycles. SERa+ oocytes (%) = n SERa+ oocytes/n total oocytes in SERa+ cycles.

^aDid not report fertilization and pregnancy rates.

^bn SERa+ oocytes/MII oocytes.

Blanks in the table reflect no available data.

Impact of SERa on fertilization and embryo development

The study of Otsuki in 2018 showed that the incidence of meiotic cleavage failure during the second polar body extrusion in oocytes with SERa+ is significantly greater than that in oocytes without SERa (OR = 5.14, 95% CI: 1.190–22.200, P = 0.028) (Otsuki et al., 2018), although their fertilization rates were comparable between SERa+ and SERa− oocytes. The same author also reported that ICSI was found to have a higher frequency of meiotic failure than conventional IVF in SERa+ oocytes. However, 10 studies assessing the fertilization rate after ICSI in SERa+ versus SERa− cycles did not find any significant difference between the two groups (Table II), while two studies reported a significantly reduced fertilization rate in SERa+ ICSI cycles (Sá et al., 2011; Restelli et al., 2015) as compared to SERa− cycles, which was also observed when adding conventional IVF data in SERa+ cycles (Oudshoorn-Roessen et al., 2015). The same team reported a significantly higher incidence of total fertilization failure in IVF SERa+ cycles than in SERa− cycles (31.8 versus 5.4%). The fertilization rate in ICSI attempts was statistically lower in SERa+ oocytes when compared to SERa− oocytes in three studies (Ebner et al., 2008; Sá et al., 2011; Hattori et al., 2014) (Table II).

Table II

Comparison of outcomes (fertilization and clinical pregnancy rates) between SERa+ and SERa− cycles and oocytes.

Study	Fertilization rate				Clinical pregnancy rate
	SERa+ cycles (%)	SERa− cycles (%)	SERa+ oocytes (%)	SERa− oocytes (%)	SERa+ cycles (%)	SERa− cycles (%)	SERa+ oocytes (%)	SERa− oocytes (%)
Otsuki et al. (2004)	81.6	85.2			5.6*	28.2*
Ebner et al. (2008)	72.8	71.9	58.9**	77.4**	26.7	41.1
Sá et al. (2011)	69.9*	76.0*	44.7*	75.8*	47.3	38.4	25.0	52.5
Mateizel et al. (2013)	76.2	73.5	72.9	77.0	33.0	32.4	22.6	32.4
Hattori et al. (2014)	71.2	70.3	64.0**	73.3**	20.0	17.8	39.6	22.9
Restelli et al. (2015)	75.0*	83.0*			26.0	29.0
Itoi et al. (2016)	71.8	76.0	70.1$	72.3
Shaw-Jackson et al. (2016)	66.0	65.0	65.0	69.0	38.0	27.0	32.0	40.0
Setti et al. (2016)	79.5	82.3	76.4	76.2	33.3	42.4
Itoi et al. (2017)	74.1	75.7	64.2$	74.1	38.5£	34.8	38.3	38.3
			67.9	81.8
Otsuki et al. (2018)	84.6£		86.4£	85.4£
A: Bielanska and Leveille (2011)			72.8	77			21.0	36.0
A: De Gheselle et al. (2014)	72.7	71.7			19.9	20.8
A: Maldonado Rosas et al. (2015)			70.0	73.0			69.0	72.0
A: Oudshoorn-Roessen et al. (2015)	37.1£*	61.0£*
A: Carvalho et al. (2016)	64.0	65.3			39.5	44.9
A: Mizoguchi et al. (2016)							40.0

HB = Healthy baby; CP = clinical pregnancy; nPC = perinatal complication; Mix = mixed transfer of SER+ and SER− embryos.

ICSI cycles were used unless otherwise indicated; ^£IVF cycles included; ^$IVF cycles only.

*P < 0.05, **P < 0.001.

Studies not included: case reports plus Braga et al. (2013) and Miwa et al. (2013). A = abstract only.

Concerning fertilized oocytes, Otsuki and collaborators have recently shown a significant higher mitotic cleavage failure in oocytes with SERa+ (OR = 2.56, 95% CI: 1.210–5.410, P = 0.014) in a time-lapse evaluation approach (Otsuki et al., 2018).

The impact of SERa on embryo development and subsequent quality was evaluated in twelve studies. Among them, five reported a significant effect of cytoplasmic SERa on embryo development. Regarding blastocyst formation, several articles underlined a lower blastocyst formation rate in SERa+ cycles (Ebner et al., 2008) or SER+ oocytes (Ebner et al., 2008; Sá et al., 2011). In blastocysts derived from SERa+ oocytes, a significantly lower good-quality rate at Day 5 was reported (Braga et al., 2013; Itoi et al., 2017). The presence of aggregates of SER was also significantly associated with the degree of blastocyst expansion and was negatively correlated with inner cell mass (ICM) and trophectoderm (TE) quality (Braga et al., 2013).

Impact of SERa on implantation and clinical pregnancy rates

A total of three studies analysed the impact of SERa+ on implantation rate. In two of them, a significant decrease in this parameter was demonstrated during SERa+ cycles (Otsuki et al., 2004; Setti et al., 2016). The remaining study showed a significantly lower implantation rate in the SERa+ oocyte subgroup (7.1 versus 21.3%, P < 0.05) (Hattori et al., 2014). However, in the most of the published data regarding pregnancy rates, the clinical pregnancy rate was comparable between SERa+ and SERa− cycles, as well as between SERa+ and SERa− oocytes (Table II). Only one study reported a significantly lower pregnancy rate in SERa+ cycles (Otsuki et al., 2004). Among the reviewed papers, only one reported a higher miscarriage rate in SERa+ cycles (58.3%) compared to SERa− cycles (22.3%) (Ebner et al., 2008).

Perinatal outcomes of babies originating from fresh or cryopreserved SERa+ cycles

A total of 386 detailed neonatal outcomes from SERa+, SERa− and mixed cycles were available. There were 364 healthy babies were born from fresh or cryopreserved SERa+ cycles while perinatal complications occurred in 22 pregnancies (Table III). These complications included 10 live births (LBs) with malformations, four neonatal death, two still births and six pregnancy terminations. Altogether, 48 healthy babies were born specifically from SERa+ embryos, while malformations occurred in one case (a ventricular septal defect) and there was one pregnancy termination, one neonatal death (from multiple malformations) and one stillbirth (in the context of a reciprocal translocation). There were an additional 53 healthy newborns originating from mixed transfers, accompanied by seven cases of perinatal complication. Seven studies allowed a calculation of the overall malformation rate per LB for SERa+ cycles (i.e. cycles with mixed embryos and only SERa+ embryos) compared with all SERa− cycles: 3.8% (10/260) and 2.1% (50/2376), respectively (Ebner et al., 2008; Sá et al., 2011; Mateizel et al., 2013; Hattori et al., 2014; Carvalho et al., 2016; Shaw-Jackson et al., 2016; Itoi et al., 2017).

Table III

Perinatal outcomes for babies born from fresh or cryopreserved SERa+, SERa− and mixed cycles.

Study	Type of embryo transfer	Newborn (n)	Perinatal outcomes
			HB	PC	Type of PC
Otsuki et al. (2004)	SERa−	1		1	1 Beckwith–Wiedmann syndrome
Ebner et al. (2008)	SERa−	4	3	1	1 diaphragmatic hernia
	Mixed	2		2	2 neonatal deaths
Akarsu et al. (2009)	SERa+	2		2	2 multiple malformations (1PT, 1 neonatal death)
Sá et al. (2011)	SERa+	1		1	1 ventricular septal defect
	SERa−	23	23
	Mixed	5	5*
Mateizel et al. (2013)	SERa+	10	10a
	SERa−	65	62	3	3 malformations£
	Mixed	28	25*	3	1 47XX+18 (PT), 2 malformations£
Miwa et al. (2013)	SERa−	14	14
Hattori et al. (2014)	SERa+	14	14a
	SERa−	16	16
	Mixed	2	2
Restelli et al. (2015)	SERa−	28	26	2	1 stillbirth, 1 aneuploidy (PT)
Shaw-Jackson et al. (2016)	SERa+	5	5$
	SERa−	23	21*	2	1 trisomy 18 (PT), 1 neonatal death
	Mixed	4	4c
Itoi et al. (2017)	SERa+	15	15$b
	SERa−	55	54$ bb	1	1 neonatal malformation
Sfontouris et al. (2018)	SERa+	1		1	1 stillbirth (reciprocal translocation with a cryptic 2q31 deletion)
Bielanska and Leveille (2011)	SERa+	1	1
	SERa−	16	16
	Mixed	15	133	2	1 heart defect + omphalocele (PT), 1 Down syndrome (PT)
Carvalho et al. (2016)	SERa−	29	28	1	1 malformation
	Mixed	4	4*
Mizoguchi et al. (2016)	SERa+	3	3
Total		386	364	22

HB = healthy baby; PC = perinatal complication; Mix = mixed transfer of SER+ and SER- embryos; PT = pregnancy termination.

^$One twin pregnancy, *including two twin pregnancies, ³three twin pregnancies, ^a12 from vitrified embryo transfer, ^bfive in IVF10 in ICSI, ^b24 in IVF and 31 in ICSI, ^Ctwo cryo-cycles resulting in two healthy births.

^£Included the following: 47XY+21, one duplicated kidney, two brachialis paresis, one hypospadias.

Predictive factors of SER occurrence

It was demonstrated by some authors that duration of ovarian stimulation and administered doses of gonadotropin (Otsuki et al., 2004; Sá et al., 2011; Hattori et al., 2014) as well as serum estradiol concentration on the day of ovulation triggering (Otsuki et al., 2004; Hattori et al., 2014) positively correlate with the presence of SERa+ in MII oocytes. Along the same lines, a large number of retrieved oocytes significantly increases the risk of occurrence of SERa (Mateizel et al., 2013; Setti et al., 2016). One study reported an association between the diameter of the SER aggregates and their occurrence in mature oocytes (Ebner et al., 2008). Moreover, oocytes degenerating after the ICSI procedure showed the largest SER aggregation diameter (51.2 ± 18.5 μm) in comparison to their fertilized and viable counterparts (22.6 ± 10.1 μm).

Discussion

The aim of this review was to cover the progression of the recent available literature data on the impact of oocytes affected by SER clusters on embryological, clinical and neonatal outcomes and we also intended to assess if the revised Alpha/ESHRE consensus (Vienna, 2017) brought a real answer on managing these oocytes. We support that discarding SERa+ oocytes, as recommended by the Istanbul Consensus (2011), may not be the most reasonable strategy. Moreover, a case by case approach as suggested by the Vienna consensus (2017) may not be helpful enough to apply when dealing with SERa+ oocytes in daily IVF laboratory practice.

SERa prevalence and impacts on IVF outcomes

Our review shows that up to 23.1% cycles could be SERa+ (extreme values are ranging from 4 to 23%) and that 17.6–29.1% of oocytes could be SERa+ per affected cycle (Carvalho et al., 2016; Shaw-Jackson et al., 2016). According to the review performed by Shaw-Jackson and colleagues, up to 10% of cycles could be SERa+. A higher prevalence reported in our study could be explained by a great detection rate of SERa+ cycles (23.1%) reported by Otsuki et al. (2018). While all of the available studies assessed SERa only in ICSI cycles, one originality in the latter study is that it investigated the prevalence of SERa including conventional IVF cycles thanks to an early fertilization check combined with a time-lapse recording system (Otsuki et al., 2018). We noted that Itoi and colleagues have also assessed the presence of SER in conventional IVF cycles thanks to an early denudation procedure but without time-lapse monitoring (Itoi et al., 2016). Not including the assessment of SERa during conventional IVF cycles may partly explain the general underestimation of the prevalence of SER aggregates. Moreover, the discrepancy in the percentage of SERa+ cycles between IVF laboratories may be explained by different resolutions of the tools used in detecting SERa, such as different types of microscopes or the use of time-lapse systems. Compared to the latest review on the topic which was performed by Shaw-Jackson et al. (2014), the current review included additional studies assessing the prevalence of SERa+ oocytes during the IVF procedure which makes this prevalence higher than that previously reported (Itoi et al., 2017; Otsuki et al., 2018). Data about the risk of recurrence of SERa oocytes are discordant. On one hand the presence of a second SERa+ oocyte for the same patient in another cycle reaches 40% (Ebner et al., 2008; Sá et al., 2011) while on the other hand, a risk of reappearance is almost 10–15% according to others (Miwa et al., 2013).

With regards to ART outcomes, some studies were reassuring as they reported similar fertilization and/or pregnancy rates between SERa+ and SERa− oocytes, as well as similar malformation rates between SERa− and SERa+ cycles and oocytes (Ebner et al., 2013; Mateizel et al., 2013; Hattori et al., 2014; Itoi et al., 2017). Similarly, Restelli found that, once the step of fertilization has been passed, cleavage and implantation rates were very similar between the compared two groups (Restelli et al., 2015). In sum, it was established by more than one study that the overall capacity of MII oocytes to develop into good-quality cleavage stage embryos and their ability to reach blastocyst stage were not different between SERa+ and SERa− cycles (Ebner et al., 2008; Hattori et al., 2014; Maldonado Rosas et al., 2015; Itoi et al., 2016; Setti et al., 2016). This supports the idea that the intrinsic developmental competence of the SERa− oocytes within a SERa+ cycle is not compromised. On the contrary, SERa+ oocytes are significantly associated with negative embryo development outcomes when compared to their sibling SER− metaphase II oocytes (Ebner et al., 2008; Sá et al., 2011; Braga et al., 2013; Itoi et al., 2017). To date, only one study has reported significant lower clinical pregnancy rates when comparing SERa+ and SERa− cycles (Otsuki et al., 2004). The same team has shown a lower implantation rate in SERa+ cycles, corroborated by Setti’s team (Setti et al., 2016). Regarding embryo development until the blastocyst stage, one study has reported a decrease in the blastulation rate in SERa+ cycles compared to negative counterparts (Ebner et al., 2008) while no difference in blastocyst formation have been described by others (Hattori et al., 2014).

Impact of SERa on neonatal outcomes

Early data related to the possible detrimental effects of transferring embryos originating from SERa+ oocytes on neonatal outcomes were alarming compared to controls (Otsuki et al., 2004; Ebner et al., 2008; Akarsu et al., 2009). Abnormal outcomes previously described were: a case of Beckwith–Wiedemann syndrome (Otsuki et al., 2004), a case of diaphragmatic hernia (Ebner et al., 2008), ventricular septal defect (Sá et al., 2011), multiple malformations leading to LB or pregnancy termination (duplicated kidney, brachialis paresis, hypospadias, heart defect and omphalocele) (Bielanska and Leveille, 2011) and chromosomal aberrations (47XX+18, trisomy 18, 47Xy+21, Down Syndrome). One must keep in mind that these complications occurred after transferring embryos deriving from SERa− oocytes originating from SERa+ cycles.

Focus on the SERa size

The size of the SERa seems to play an important role in ART outcomes since a correlation was found between its diameter and ICSI outcomes (Ebner et al., 2008). However, no target size of SERa from which adverse IVF outcomes are more likely to appear has been clearly established. Interestingly, it was reported that during denudation procedures and when unfertilized oocytes are cultured, SER aggregates that initially did not contain these features could appear in oocytes. As previously described by Otsuki, there are three forms of SERa, classified on basis of size: large, medium and small aggregates. Contrary to the first two cases, visible by light microscopy, small aggregates of SER can only be assessed by electron microscopy. Furthermore, culturing oocytes with medium size SERa lead to larger aggregates after 18 h (Otsuki et al., 2004). These data suggest, first that the three forms may have the same origin, not always visible under light microscopy which is still the most commonly used tool in IVF laboratories nowadays. So, injecting oocytes with apparently no SER aggregates under light microscopy does not mean that these oocytes are free of small SERa that cannot be seen under light microscopy magnification. Secondly, it implies that all the oocytes originating from the same stimulated cycle among which at least one oocyte presents visible SERa under light microscopy may be concerned by SERa. Here again the question which arises: which oocyte to keep and which one to discard especially when managing SERa negative oocytes in a SERa positive cycle?

Predictive factors of SERa occurrence

The mechanism underlying SERa formation remains currently unknown and data about the risk of recurrence of SERa oocyte are discordant and not conclusive enough to claim a unique constitutive origin. On one hand, the presence of a second SERa+ oocyte for the same patient in another cycle reaches 40% (Ebner et al., 2008; Sá et al., 2011) while on the other hand, the risk of reappearance is almost 10–15% for others (Miwa et al., 2013). SERa occurrence could be constitutive and/or linked to predictive factors. The serum estradiol level on the day of hCG administration, induced by the growth of ovarian follicles, was higher in cycles with SERa (Otsuki et al., 2004; Hattori et al., 2014; Itoi et al., 2016). Similarly, it has been suggested that this dysmorphism could be a consequence of ovarian stimulation since clusters of SER have not been observed in germinal vesicle oocytes (Van Blerkom and Henry, 1992) except in very rare cases (Otsuki et al., 2018). Moreover, previous studies have shown a positive correlation between the risk of SERa+ oocytes and the number of retrieved oocytes as well as the size of follicles in ovarian stimulation cycles. However, the occurrence of SERa did not increase in patients with polycystic ovary syndrome nor in case of ovarian hyperstimulation syndrome. In the same trend, in a murine model, no obvious changes in microorganelles, including smooth and rough endoplasmic reticulum in mouse oocytes, were detected after induced ovulation (Lee et al., 2017). Otsuki et al. suggested that the presence of SERa in the oocyte could be a sign of prolonged cytoplasmic maturation. Indeed, SERa appear and grow during oocyte prolonged culture (for 2–5 days) (Otsuki et al., 2004). A hypothetical cytoplasmic deterioration occurs in ageing human oocytes during extended culture, possibly promoting formation of SER clusters. Hence, there are at least two different interpretations of oocyte aging: the first one in case of advanced maternal age, and the second one is in-vitro aging, due to extended culture during ART procedures. On the contrary, ultrastructural studies of MII oocytes submitted to in-vitro aging shows a significant decrease in mitochondria-smooth endoplasmic reticulum aggregates (Bianchi et al., 2015). Hence, regarding these contradictory results, an impact of the delay between ovulation triggering and oocyte retrieval or the fertilization procedure on the SERa occurrence could be relevant and should be further explored.

Molecular status of SERa+ oocytes

To address the need to explain the molecular mechanisms underlying the observed detrimental effects of SERa+ oocytes on subsequent embryo development, Stigliani and colleagues recently underwent for the first time a transcriptomic approach comparing gene expression profiles between SERa+ and SERa− oocytes (Stigliani et al., 2018). They pointed to a set of genes that were significantly dysregulated in case of SER aggregates among 24 SERa+ oocytes: (i) down regulation of several genes involved in cytokinesis and mitotic/meiotic regulation, spindle assembly and chromosome partition or gene involved in mitochondrial structure and respiratory activity and (ii) an up-regulation of other genes involved in organization of cytoskeleton and microtubules. At a molecular level, the presence of SERa in human metaphase II oocytes alters the expression profile of crucial genes in comparison to normal counterparts. These data are in line with recent observations reporting that SERa affect spindle size and generate cortical actin disorganization (Dal Canto et al., 2017), which may lead to aberrant embryo cleavage. As the latter is involved in cytokinesis, loss of its integrity may have an impact on cell cleavage. This observation is consistent data reported by Otsuki and colleagues on the increase of mitotic cleavage failure in oocytes with SERa (Otsuki et al., 2018). However, there is no available study to answer the question of whether these potential effects on spindle assembly and chromosome segregation are correlated with a higher aneuploidy risk in the subsequent embryo. Considering the existent phenomenon of self-correction in oocytes, there is a real need to verify if the risk of producing aneuploid embryo still persists despite the ability of the oocyte to correct certain chromosome abnormalities.

In addition, calcium signalling seems to be disrupted in oocytes affected by SERa, with a significant lower calcium oscillation frequency, a longer duration of the first peak of calcium and a higher total amount of calcium released during fertilization in comparison to SERa− oocytes (De Gheselle et al., 2014). A causal link between intracellular calcium perturbation upon oocyte activation and the occurrence of both organogenesis abnormalities and chromosomal aberrations among offspring is difficult to establish. It was hypothesized that SERa+ oocytes could exhibit epigenetic disorders which will be transmitted to the subsequent embryo. But this still to be not sufficient to explain the heterogeneity of the reported perinatal complications.

Suggestions for managing oocytes with SERa since the revised Alpha/ESHRE consensus (Vienna, 2017)

Based on current evidence, in 2017, the Expert Panel revised the Istanbul consensus and suggested that the decision to inject SER+ oocytes should be reviewed by the clinical team on a case by case basis. Moreover, they advised a strict follow up of the pregnancy and of the subsequent baby after transferring embryos originating from SERa+ oocytes. Our study reported, in the current review, 48 healthy babies born after transfers of embryos originating from SERa+ oocytes, with only four babies suffering from perinatal complications. In the present study, we have reported a percentage of 5.7% (22/386) of perinatal complications in all SERa+ cycles (with SERa+, SERa− and mixed embryos). The previous review of the literature performed by Shaw-Jackson and colleagues, yielded 8.6% (16/187) perinatal complications. To date there is no clear association between SERa and foetal malformations. In this context of controversy, we think that concluding on a case by case approach may not be helpful enough to embryologists in making decisions when managing SERa+ oocytes, especially in certain activities such as female fertility preservation with daily cases of SERa+ oocytes/cycles. In such a category of patients, discarding SERa+ oocytes could be a real waste as it decreases their chances to achieve a LB especially when the number of oocytes is restricted, which is often the case. This allegation is also relevant for patients with poor prognosis due to advanced female age or showing poor ovarian response to ovarian stimulation. A previous study estimated a transfer cancellation rate of 18% during 130 ICSI cycles including SERa oocytes when compared to matched controls (8%) (Restelli et al., 2015).

Since the strategy of managing SERa+ oocytes should be standardized whatever the field is (female fertility preservation or couple infertility management with IVF or ICSI procedures), and based on the current available data on the subject, we suggest not to discard SERa+ oocytes. According to the current study, it can be assumed that a considerable number of embryos originating from oocytes with SERa could have the same implantation and development potential as those without SERa. Hence, we think that these embryos could be considered for reproductive purposes when no embryo deriving from SERa− oocytes are available. In sum, we advise injecting/inseminating all oocytes (SERa+ and SERa−). On the day of embryo transfer, embryos originating from SERa− oocytes, in both SERa positive or negative cycles, would be prioritized. But when only embryos originating from SERa+ oocytes are remaining, the couple should be informed about the state of the art on the topic and these embryos could be transferred when the couple’s consent is obtained. In that specific case, embryo transfers should be strictly carried out with close follow up during foetal development as well as after birth. Furthermore, data concerning neonatal parameters and the health of obtained children should be published.

Conclusion

We believe that our study may offer further information that can support embryologists to make decisions on whether SERa+ oocytes should be used for IVF treatment. In the light of this review, we confirm a high prevalence of this oocyte dysmorphism. Moreover, reassuring literature data concerning embryo transfer in case of SERa− oocytes originating from SERa+ cycles encourage avoiding waste of oocytes and embryos. On an ethical standpoint, we think that discarding SERa+ oocytes may not be the most reasonable strategy. Furthermore, a case by case approach such as that recommended by the Vienna consensus may be not enough helpful and clear enough to apply when dealing with SERa+ oocytes and SERa− oocytes in SER+ cycles in IVF laboratories. Given the lack of published studies on the outcomes of transferring embryos originating from SERa+ oocytes, and because our objective is to offer the most ethical practice and the best care for patients, we recommend fertilizing SERa+ oocytes. The selection should be performed on the day of embryo transfer: embryos from SERa− oocytes would be preferably transferred. When only embryos deriving from SERa+ oocytes remain, the couple will be provided with appropriate information concerning the available literature data on the topic in order to help them make a decision. If their consent is obtained, a strict follow-up of foetal development and the baby at birth is mandatory.

Acknowledgements

We would like to thank Mr Alexandre Boutet, a member of the ‘Bibliothèque Inter-Universitaire de Santé’, Paris for his help in the database research.

Authors’ roles

L.F. did the literature search, data analysis and wrote the article. A.S. was involved in the data analysis and wrote the article. All of the authors contributed to the scientific discussions and reviewed the article. K.P.C. designed the study, revised the article and approved the final draft.

Funding

No funding was received for this study.

Conflict of interest

None declared.

Author notes

The authors consider that the first two authors should be regarded as joint first authors.