Cytoplasmic String Between ICM and mTE Is a Positive Predictor of Clinical Pregnancy and Live Birth Outcomes in Elective Frozen-thawed Single Blastocyst Transfer Cycles: a Time-lapse Study

Background: In this study, we aim to investigate whether cytoplasmic string between inner cell mass (ICM) and mural trophectoderm (mTE) is a positive predictor of clinical pregnancy and live birth outcomes. Methods: 1,267 elective frozen-thawed single blastocyst transfer (eSBT) cycles cultured in time-lapse incubation system from January 2018 to May 2019 were involved in the study. Blastocysts were grouped according to the appearance of cytoplasmic strings between ICM and mTE cells, and identied as “Present” and “Absent” groups. In Present group, they were further categorized according to the quantity of cytoplasmic strings between ICM and mTE cells. Clinical pregnancy and live birth outcomes of blastocysts were used to evaluate the effect of cytoplasmic strings between ICM and mTE. Results: The baseline demographic and laboratory features were similar between the Present and Absent groups of cytoplasmic strings between ICM and mTE (P>0.05). According to the time-lapse analysis, cytoplasmic strings between ICM and mTE were more visible among good quality blastocysts. Furthermore, blastocysts with cytoplasmic strings showed a higher clinical pregnancy and live birth rates (P<0.05), and no signicant differences were observed in abortion rate and birth weight (P>0.05). Conclusions: Although the previous conclusions of cytoplasmic strings were controversial, the present time-lapse analysis provides the evidence for the rst time that cytoplasmic strings between ICM and mTE cells would be a positive predictor of clinical pregnancy and live birth outcomes in elective frozen-thawed single blastocyst transfer cycles. eSBT, elective single blastocyst transfer; IVF, in vitro fertilization; ZP, zona pellucida; KID, known implantation data; COS, controlled ovarian stimulation; HCG, human chorionic gonadotropin; COC, cumulus-oocyte complexe; ICSI, intracytoplasmic sperm injection.

Furthermore, blastocysts with cytoplasmic strings showed a higher clinical pregnancy and live birth rates (P<0.05), and no signi cant differences were observed in abortion rate and birth weight (P>0.05).
Conclusions: Although the previous conclusions of cytoplasmic strings were controversial, the present time-lapse analysis provides the evidence for the rst time that cytoplasmic strings between ICM and mTE cells would be a positive predictor of clinical pregnancy and live birth outcomes in elective frozenthawed single blastocyst transfer cycles.

Background
The goal of assist reproductive technology is to achieve a healthy singleton birth and minimize maternal and neonatal risk [1,2]. In order to realize the goal, elective single blastocyst transfer (eSBT) was preferred during these decades [3,4]. Blastocyst transfer is considered as a more physiologically appropriate time due to the more closely mimics the time of natural implantation and may improve synchrony between the endometrium and embryo development [5,6]. It is reported that the clinical pregnancy and live birth rates of elective single blastocyst transfer (eSBT) are similar with double embryos transfer, and higher than those with single cleavage-stage embryo transfer. Furthermore, eSBT generates a reduced incidence of multiple pregnancy than double embryos transfer [7][8][9].
With the development of cryopreservation technology, more embryos were selected to be cryopreserved for further use. It is reported that frozen-thawed embryos transfer could lead to a higher live birth rate in women with polycystic ovary syndrome. Besides, frozen-thawed embryos transfer could avoid the exposure of the endometrium to the adverse sequelae of ovarian stimulation [10]. Besides, supraphysiologic hormonal milieu increases the incidence of ectopic pregnancy in fresh cycles. So, frozen-thawed eSBT strategy could not only optimize pregnancy rates, but also maintain perinatal safety [11].Blastocyst selection is the key to a successful eSBT strategy, as it is important to shorten the time to get pregnancy [12]. In the past, morphological assessment was the main criterion to select blastocysts. The blastocyst grading system by Gardner and Schoolcraft remains largely unchallenged and widely used in clinic [13]. However, this system did not include some important morphological parameters often observed in the in vitro fertilization (IVF) laboratory. Albeit of unknown importance, some of these parameters were brie y introduced in the Istanbul consensus document, i.e. the formation of cytoplasmic strings often linking together different cells and cell types [14].
Methods for embryo selection have greatly improved over the past decade, not only static morphological assessment, but also integration of no-invasive time-lapse analysis. Cytoplasmic strings could be observed between oocyte and zona pellucida (ZP), the outside surface of blastomere and traversing from inner cell mass (ICM) and mural trophectoderm (mTE) by time-lapse system. It is reported that these strings would be required to make possible a selective and bidirectional communication between mitotic mTE cells and ICM cells [15]. But, the impact of cytoplasmic strings on the clinical outcome of blastocyst is controversial [16].
In this study, we aim to determine whether the cytoplasmic strings between ICM and mTE cells is a positive predictor of clinical pregnancy and live birth in frozen-thawed eSBT cycles.

Study design
In the present study, we retrospectively analyzed 3,223 transfer cycles with a time-

Clinical Protocol
According to the characteristics of patients, main controlled ovarian stimulation (COS) were performed as per our previous paper [17,18]. During COS, patients were monitored closely using transvaginal ultrasound. Human chorionic gonadotropin (HCG) was given when the leading follicle(s) was > 18 mm. Thirty-six hours after HCG injection, ultrasound-guided oocytes retrieval was carried on.

Page 4/17
The protocol of semen/cumulus-oocyte complexes (COCs) preparation, insemination and embryo culture were performed as our previous paper [19]. Density gradient centrifugation was used to optimize the semen samples. The concentration, motility and morphology of sperm were evaluated according to the fth edition of WHO guidelines. In IVF cycles, each COC was inseminated with 10,000 motile spermatozoa 3-4 h after retrieval. In intracytoplasmic sperm injection (ICSI) cycles, the COCs were denuded 2 hours later, and sperm was injected 4 hours later after retrieval. Then, zygotes were transferred to G1 Plus (Vitrolife, Sweden) and cultured in time-lapse incubation system. The system was set to take pictures for each embryo every 10 min. Pronuclei were checked 16-18 h later after insemination. On Day 3, the media was replaced by G2 Plus (Vitrolife, Sweden). On Day 5 and 6, blastocysts were cryopreserved for further use. All culture media were Vitrolife G-series media (Vitrolife, Sweden).

FET protocols
The FET protocols were referred to our previous papers. For natural cycles, the endometrial thickness, follicle growth, and ovulation were measured by transvaginal ultrasound examination and measurement of the serum progesterone levels were initiated from cycle days 10 to 12. Thawing and transferring of blastocysts was planned 3 days after ovulation. From 1 day after ovulation, progesterone is injected intramuscularly for luteal support.
For hormone replacement treatment cycles, oral estradiol (Progynova; Bayer; Leverkusen) signed 2 mg/day from 1 to 4 cycle days, 4 mg/day from days 5 to 8 days, and 6 mg/day from 9 to 12 days. On the 13th day, when the thickness of the endometrium reaches 8.0 mm or the maximum, 40 mg of progesterone is injected intramuscularly for 3 days. The blastocyst was thawed and transplanted on the 4th day after 3 days of progesterone administration.
Four weeks after transfer, the fetal cardiac activity in the uterus was observed transvaginal ultrasound to determine the clinical pregnancy. Miscarriage was de ned as loss of the gestational sac or fetal heart activity within 20 weeks after con rming clinical pregnancy.

Grouping
Blastocysts were grouped according to the appearance of cytoplasmic strings between ICM and mTE cells, and identi ed as "Present" and "Absent" groups. In Present group, they were further categorized according to the quantity of cytoplasmic strings between ICM and mTE cells as follows: single or double string(s) (Class I); three to four strings (Class II) and ve strings or more strings (Class III) (Fig. 2).

Statistical Analysis
Statistical analysis was performed by the Statistical Package for Social Sciences (SPSS) version 19.0 (SPSS Inc., USA). T-test was used as a continuous variable to compare difference, such as age, Body Mass Index (BMI), duration of subfertility, FSH, AMH, oocytes MII, 2PN, and so on. Chi-square test was used for binary and categorical variables, such as primary subfertility, insemination, clinical pregnancy, and so on. Statistical signi cance was established at P < 0.05.

Patient and cycle characteristics
A total of 1,267 frozen-thawed eSBT cycles at Reproductive Medicine Center, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology were involved in the cohort study. According to the appearance of cytoplasmic strings between ICM and mTE cells, they were allocated as "Present" group with 1,002 cycles and "Absent" group with 265 cycles. So, 79.08% transferred blastocysts could be observed with cytoplasmic strings between ICM and mTE cells, and 20.92% could not (Fig. 3A).
Patients in the two groups were similar in baseline demographic features. The difference between the average maternal ages, maternal BMI, ratio of primary subfertility, duration of subfertility, the value follicle stimulating hormone (FSH) and anti-mullerian hormone (AMH) of the two groups were not signi cant (P > 0.05, Table 1). Therefore, the baseline demographic features of the two groups were similar. The cycle characteristics, such as insemination method, oocytes, MII, 2PN, FET protocol and the endometrial thickness were all similar (P > 0.05, Table 1).

The frequency of strings observed in different blastocyst stages
Most transferred blastocysts that meet the criterion were cultured on Day 5 and Day 6. Few blastocysts culture was prolonged to Day 7 (N = 3). As shown in Fig. 4A, the percentage of the appearance of cytoplasmic strings between ICM and mTE cells were similar from Day 5 to Day 7 (P < 0.05).
The blastocyst development stage and quality of ICM and TE were was referenced to David Gardner as follow: Stage 2: the volume of the blastocoel is not less than half of the total volume of the blastocyst; Stage 3: the blastocoel completely occupies the total volume of the blastocyst; Stage 4: the blastocyst expands, the blastocoel completely lls the blastocyst, the total volume of the embryo becomes larger, and the zona pellucida becomes thinner; Stage 5: the hatching blastocyst; Stage 6: the blastocyst hatched from ZP [13]. In most cases, the vitri cation criterion is larger than Stage 2. So, only 21 cases at Stage 2 were vitri cated. As shown in Fig. 4B, except Stage 2, other stages were present a similar percentage of appearance of cytoplasmic strings between ICM and mTE cells, from 72.7-90.00% (P < 0.05). Only 28.6% blastocysts at Stage 2 could be observed with the cytoplasmic strings between ICM and mTE cells.
The ICM/TE grades are also referenced to David Gardner as follow: A. a large number of cells and closely packed; B. a small number of cells and loosely packed; C. very few cell numbers [13]. The percentages of stings appearance were signi cantly less in blastocysts with a Grade C of both ICM and TE than that with Grades A and B ( Fig. 4C and D). So, the appearance of strings was rare in low-quality blastocysts.

Pregnancy and perinatal outcomes
The main outcome of the study was the clinical pregnancy rate and live birth rate. As shown in Fig. 5A, 59.1% blastocysts were implanted in Present group, and 48.2% blastocysts were implanted in Absent group. So, appearance of strings may lead to a better clinical pregnancy rate (P < 0.05). Among three classi cations of blastocysts, Class I and II present a similar clinical pregnancy (58.8% and 53.0%, P > 0.05), but Class III (97.5%) is much higher than other groups (P < 0.05).
The blastocysts in Present group also achieve a higher live birth rate compared with Absent group (48.5% versus 35.5%, P < 0.05, Fig. 5B). The results of live birth showed the similar trend with clinical pregnancy rate. The live birth rates of blastocysts in Class I and II were 50.7% and 40.0%, respectively (P > 0.05). But, the rate in Class III was 77.5%, much higher than the other two groups (P < 0.05).
As shown in Fig. 5C, the abortion rate was not signi cantly different in all groups. Though, the abortion rate is Class III reached 20.0%, the case number was too little to cause a signi cant difference (N = 12).
For perinatal outcome, the neonatal birth weight was statistically similar in all groups, from 3.31 to 3.37 kg (P > 0.05, Fig. 5D). So, the cytoplasmic strings between ICM and mTE cells did not affect the neonatal birth weight.

Discussion
Nowadays, the main criterion for select blastocyst to transfer is morphological feature, such as the grades for ICM and TE based on Gardner system. However, some essential morphological parameters may be neglected, i.e. cytoplasmic strings among cells. It is reported that some TE cells at the polar-mural junction extend cell projections to the surface of the ICM cells through the cytoplasmic strings by observing the ultrastructure of blastocysts [20]. Salas-Vidal and Lomeli (2004) reported that lamentous actin was detected in the cytoplasmic extensions or lopodia in mouse blastocysts [15]. There existed two kinds of lopodia: 1) short lopodia extend both from the ICM and the mural TE into the blastocoel cavity; 2) long, thin lopodia traverse the blastocoel from the mural TE to a central ICM cell. The short lopodia exist abundantly in blastocoel; and about 40% blastocysts could be observed with the long lopodia.
The cytoplasmic strings between ICM and mTE cells could commonly be observed at blastocyst stage by time lapse system. It is reported that 55-60% of blastocele could obverse this phenomenon [21]. In present study, even 79.08% transferred blastocysts could be observed with cytoplasmic strings between ICM and mTE cells. The reason that present percentage is higher in our study may be the cases were transferred blastocysts, rather than all blastocysts. Besides, these blastocysts were selected as the most likely to get pregnancy, so we may conclude the cytoplasmic strings were more visible among good quality blastocysts. Therefore, it is crucial to explore the origin and effect of these strings.
The cytoplasmic strings may be caused due to the immigration from polar TE cells to mural TE [22]. Salas-Vidal and Lomelı́ (2004) reported that the cytoplasmic strings between ICM and mTE cells may suggest cellular activity [15]. Besides, lopodia are dynamic, appearing to extend and retract in coordination with cell division [21]. The strings were also veri ed by immunolocalization of the FGFR2 and ErbB3 receptors to indicate a direct communication between the mural TE and the ICM cells [15]. It is also reported that the cytoplasmic string could receive the required mitotic signals from the ICM. Other studies found that lopodia was abundant of E-cadherin, and determined cell elongation behaviors required by compaction. When α-catenin and β-catenin were knocked down, the lopodia quality was reduced dramatically [21,23]. So, E-cadherin may be the main ingredient of lopodia. Besides, these cytoplasmic strings not only exist in vitro culture, but also in uteri of volunteers [24]. So, cytoplasmic strings between ICM and mTE cells is a common physiological phenomenon that occurs during cell growth. The presence is conducive to the communication between the mural TE and the ICM cells, conducive to blastocyst development.
In present study, the group of presence of cytoplasmic strings between ICM and mTE cells was more visible with a higher grade of ICM and TE. Therefore, we may conclude that cytoplasmic strings between the mTE and the ICM cells may be an indicator of better quality blastocysts. This may due to the enhanced cellular communication between ICM and mTE cells. Therefore, it is interesting to explore the effect of these strings on the clinical and perinatal outcome of blastocysts.
In order to compare the pregnancy outcomes between blastocysts with present and absent cytoplasmic strings between ICM and mTE cells, clinical pregnancy, live birth and abortion rates were evaluated. A higher clinical pregnancy and live birth rates were observed (P < 0.05), but abortion rate was not signi cantly different (P > 0.05 Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no con ict of interest.   Classi cation of cytoplasmic strings between ICM and mTE cells. Threads were classi ed into three groups, based on their quantity: single and double string(s) (I), three to four strings (II) and more than four strings (III). ICM, inner cell mass; mTE, mural trophectoderm.