What Is the Optimal Timing of Thawing for Transferring Vitried-warmed Cleavage Stage of Slow-growing Embryos? A Cohort Retrospective Study

Purpose To evaluate optimal thawing time, the early thawing or the routine thawing time, for transferring vitried-warmed, and cultured overnight cleavage stage of the slow-growing embryos on Day 3 in frozen embryo transfer (FET) cycle. Methods This was a retrospective cohort study from 2017 to 2018, a total of 705 slow-growing embryos FET cycles in which the patients were aged <40. Thawing cleavage stage slow-growing Day 3 embryos on either the 2nd or 3rd day after ovulation in natural cycle or the equivalent timing of the articial cycles.


Abstract Purpose
To evaluate optimal thawing time, the early thawing or the routine thawing time, for transferring vitri edwarmed, and cultured overnight cleavage stage of the slow-growing embryos on Day 3 in frozen embryo transfer (FET) cycle.

Methods
This was a retrospective cohort study from January 2017 to July 2018, a total of 705 slow-growing embryos FET cycles in which the patients were aged <40. Thawing cleavage stage slow-growing Day 3 embryos on either the 2nd or 3rd day after ovulation in natural cycle or the equivalent timing of the arti cial cycles.

Conclusion
For slow-growing embryos, higher pregnancy outcomes were shown in early thawing strategy as compared to the routine thawing, which suggested that the improvement of endometrium-embryo synchronism may correct the time difference brought by the slow-growing embryos.

Background
Controlling the time overlap between embryo implantation and high receptivity of endometrium is the prerequisite for the success of embryo transfer [1]. Previous data showed that the implantation window phase last about 2 to 5 days [2], and higher pregnancy rate may be gained from the more paralleled development between embryos and endometrium [3,4]. Recently, several studies emphasized the precise timing of thawing and transferring embryos in FET cycle. For instance, S. Tannus et al. pointed out that extending culture of Day 5 morula to Day 6 and subsequent FET could enhance the live birth rate, indicating that the implantation potential of Day 5 slow growing embryo could be rescued to some extent by improving the endometrium and embryo synchronization [5]. However, C. Blockeel et al. reported that there was no signi cant difference in the pregnancy outcome between the early and delayed transfer of vitri ed-warmed cleavage stage embryo in FET cycles [6]. The contradictive results may be stemmed from the different stages of embryos. Till now, information regarding the optimal thawing and transferring time especially for the slow-growing cleavage stage embryos is still lacking.
Generally speaking, blastomere number as 8 could predict the quality of Day 3 embryos with obtaining satis ed live birth rate [7]. However, not all embryos exhibit the same development speed, and study reported that slow-growing embryos may account for 30% of total [8]. Slow-growing Day 3 embryos refer to embryos which have 6 or fewer blastomeres [9], which some studies reported would have negative effects on embryo-endometrial synchrony and therefore decrease embryo implantation rate [10,11]. Lewin A et al. showed that lower pregnancy rate gained from the implantation of slow-growing cleavage embryos compared to normal-growing embryos [12], while Heather Burks et al. indicated that adjusting the timing of a slow embryo transfer could lead to a comparable pregnancy outcome to that of a normal developing embryo [13]. It still lacks data that regulating the thawing and transferring time of slowgrowing Day 3 embryos may improve clinical pregnancy outcome. Therefore, the goal of this analysis is to investigate the effect of thawing and implantation time of slow-growing embryos in cleavage phase on clinical outcome during FET cycle by thawing cleavage stage Day 3 embryos on either the 2nd or 3rd day after ovulation in natural cycle or the equivalent timing for the arti cial cycles.

Study design
This was a retrospective cohort study initiated at the Center for Reproductive Medicine in the Peking University Shenzhen Hospital, China between January 2017 and July 2018. A total of 705 FET patients in which patients aged < 40 were included and 1486 embryos were formed, of which 1366 embryos were eventually transferred. All patients underwent only one FET cycle. In our research, the nal transplanted slow-growing embryos were divided into early thawing group and routine thawing group according to the physician's decision, all of which were cultured overnight and were transferred one day later. This study has been approved by the Ethics Review Committee of Shenzhen Hospital of Peking University.
Ovarian stimulation, IVF or ICSI treatment, Embryo freezing and thawing Embryos were gained from controlled ovarian stimulation with conventional in vitro fertilization (IVF) or intracytoplasmic sperm injection (ICSI). All embryos were assessed according to the routine evaluation system [9] and were cryopreserved on Day 3 of embryo culture. All the cryopreservation was performed using vitri cation protocols. For Day 3 vitri cation, embryos with at least 4 blastomeres and ≤ 25% fragmentation were selected for cryopreservation. Only the slowing growing embryos which were composed of 4 to 6 blastomeres were included in this study.
Embryos were thawed at the following two different time points and classi ed into two different groups (Fig. 1). As for the natural cycle and controlled ovarian stimulation cycle, embryos were thawed after 2 days (early thawing group) or 3 days (routine thawing group) of ovulation. As for the hormone replacement treatment (HRT) cycle, embryos were thawed after 3 days (early thawing group) or 4 days (routine thawing group) of progesterone administration.
Embryos were then warmed 1 day before transferring. The surviving embryos were cultured overnight and transferred at the 2nd day. Morphological survival was con rmed by counting the intact cells on the number of cells present at cryopreservation. Embryos with at least 50% intact cells were considered surviving and were further cultured. Further cleavage was evaluated the next morning and characterized into top quality, good quality and poor quality according to the previous standard [6]. Embryos with further cleavage and showing signs of compaction or blastulation were considered as top quality. Good quality embryos had at least eight blastomeres and further cleavage of at least two cells, while poor quality embryos had less than eight blastomeres and/or no further cleavage and/or limited further cleavage of only one cell. The further cleavage rate was de ned as the percentage of embryos with at least division of two cells after one-day culture on the total number of transferred embryos.

Preparation of the endometrium
As for natural cycle and natural cycle with ovulation stimulation cycle: Ultrasound examination was applied to monitor the follicular growth from day 8-10 of the menstrual cycle and consecutively examined every 2-4 days. After the luteinizing hormone (LH) surge, dominant follicle collapse was con rmed as the ovulation day (day 0) [14]. Besides, ovulation day was double con rmed according to serum progesterone (P) level. If dominant follicle disappeared, P serum level was checked. In the natural cycle with hCG, after ultrasound evidence of follicles ≥ 18 mm and endometrium thickness ≥ 7 mm, 6,000 IU urinary hCG (Choriomon, IBSA, Lugano, Switzerland) was administered to trigger ovulation.
As for HRT cycle and GnRH-a pretreatment plus HRT cycle: before the start of the cycle, the basal hormone value of endocrine evaluation was con rmed by the measurement of estradiol (E2), P, luteinizing hormone (LH) and follicle stimulating hormone (FSH). Then a dose of 2 mg estradiol valerate (Progynovaw, Bayer-Schering Pharma AG, Berlin, Germany con rm) twice a day was given to the patients for 7 days, followed by 6 days of estradiol valerate at a dose of 2 mg three times per day. On Day 13 of estradiol valerate treatment, endometrial thickness was measured by ultrasound and hormonal analysis was performed by serum E2 and P examination. In FET cycle, the same embryonic time convention is usually followed, and the date of ovulatory in the natural cycle is equivalent to the 1st day of P administration in the arti cial cycle [15]. If the serum P level was higher than.1.5 ng/ml, the cycle was cancelled. If endometrial thickness was ≥ 7 mm, P supplementation was started. The choice for using GnRH-a was depended on the physician's decision.

Assessment of pregnancy outcome
Conception was de ned as arbitrary serum beta hCG is higher than 50 IU/L at 14 days following embryo transfer. Clinical pregnancy was con rmed by ultrasound scan with at least one fetus with heart beat by 7 weeks pregnancy. The implantation rate was that the number of gestational sacs observed divided by the number of embryos transferred and the live birth was de ned as at least one baby born.

Outcome measures
The pregnancy outcome of FET cycle was studied, including conception, clinical pregnancy, implantation, biochemical miscarriage, rst trimester pregnancy loss, second trimester pregnancy loss, ongoing pregnancy and livebirth.

Statistical analyses
The p-value, risk ratio and con dence interval were analyzed by Social Science 19 (SPSS, Inc., Chicago, IL, USA). The categorical data are expressed as percentage and analyzed by using the chi-square test or Fisher's accurate test according to the sample size. The odds ratio (OR) and 95% con dence interval (CI) were compared to evaluate the difference. According to the normality of the results, the continuous variables were analyzed by independent t test or Mann-Whitney u test, and all the data were two-tailed test. Logistic regression analysis was used to adjust for confounders, including regimen of endometrial preparation and the timing of thawing embryos. The signi cance level was set at p < 0.05.

Results
During the study period, 705 FET patients were initially included, of which 470 were in the early thawing group and 235 were in the routine thawing group. In the early thawing group, 2 patients were lost followup and 468 patients were included ultimately. In the routine thawing group, there was no lost follow-up patients, and 235 patients were included nally (Fig. 2). As indicated in Table I, there was no signi cant difference in the basic characteristics including age, height, weight, BMI, indication for IVF/ICSI and basic hormones between the early thawing group and the routine thawing group.
The cycle characteristics of the two groups was presented in Table II. There was no signi cant difference between the two groups in the choice of IVF or ICSI. The endometrial thickness on the transferring day was 11.3 ± 2.3 mm in early thawing group versus 11.2 ± 2.2 mm in routine thawing group. With an average of about 2 embryos per patient, there was no signi cant difference in the number of embryos transferred.
The embryonic characteristics was exhibited in Table III. The ratio of embryos transferred was 92.0% and 91.8% in the two groups, which showed no signi cant difference. There was no signi cant difference in the percentage of the further cleavage embryos between the two groups. After thawing and culturing overnight, the quality of embryos was scored and characterized as top quality, good quality and poor quality. There was no signi cant difference in the embryo quality distribution between the two groups.
As shown in Table IV, clinical outcomes, including conception per woman, clinical pregnancy rate, implantation rate, biochemical miscarriage rate, rst trimester pregnancy loss, second trimester pregnancy loss, ongoing pregnancy and live birth (single and twin) rate were compared between two groups. The clinical pregnancy rate in the early thawing group (152/468(32.5%)) was signi cantly higher than that in the routine thawing group 55/235(23.4%)) (OR 1.39 (CI 1.06-1.81), p = 0.013), while there was no statistically signi cant difference in pregnancy loss between the two groups (OR 0.89 (CI 0.53-1.47), p = 0.65). Besides, the implantation rate, ongoing pregnancy rate and live birth rate in the early thawing group were also signi cantly higher than that in the routine thawing group as depicted in the table IV.
The quality of embryos which obtained clinical pregnancy in the two groups was compared and the result demonstrated there was no signi cant difference in the percentage of top, good and poor quality between the two groups at the transfer day (Table V).
Additionally, logistic regression analysis of mixed factors with Clinical pregnancy were showed in table VI. Several variables, including regimen of endometrial preparation and the timing of thawing embryos, were employed in the logistic regression analysis to reduce the in uences on clinical pregnancy outcomes. The results showed that the other three groups were shown to be associated with similar clinical pregnancy outcome compared with the natural cycle. Besides, the clinical pregnancy outcome of thawed embryos on day-2 was signi cantly higher than that on day-3.

Discussion
The synchronous development of embryo and endometrium is an important prerequisite for the success of embryo transfer. As for FET cycle, the development of endometrium and embryo are isolated [14]. Slow-growing embryos have lower implantation potential compared to normal embryos [10,16]. When the embryos develop slowly, at the same time, elevated progesterone in vivo would lead to endometrium decidualization, and therefore a "time difference" would exist in the implantation of the embryos according to the routine time. The time difference between the slowing-growing embryos and the accelerated development endometrium may be main reasons why the implantation rate of the slowgrowing embryo is lower than that of the normal embryo. For instance, M. W. Healy et al. compared the pregnancy outcome of implantation of slow-growing Day 5 and Day 6 embryos when premature progesterone promoted on the trigger day and found that lower live birth rate was shown on Day 6 embryos (45.6% versus 34.0%), indicating that endometrium-embryo synchronism decreases when slowgrowing embryos encounter advanced endometrium over time [17]. In Kevin S. et al. study, the clinical pregnancy of blastocyst with similar implantation potential on the 5th, 6th and 7th day after ovulation was compared, lower clinical pregnancy rate was found on Day 7 blastocysts cryopreserved [18]. It seemed that the reason for low pregnancy rates of later developing blastocysts may origin from the asynchrony of embryos and endometrium instead of the slow development of the embryos. In our study, early thawing group obtained higher pregnancy outcome by transferred on the 2nd day after ovulation in natural cycle or the equivalent timing for the arti cial cycles, just as the normally developed embryos are transferred after thawing on the 3rd day, which effectively resynchronizes the embryo to the development of the endometrium.
In addition, some studies extended the embryo culture to a certain degree of development, and compared them with the normal growing embryo of pregnancy outcomes. Heather Burks et al. analyzed the implantation between those who reached 8 cells on Day 3 (normal embryo group) and those who gained 8 cells on Day 4 (delayed embryo group) and found similar pregnancy outcomes [13]. The above study suggested that the synchronization of endometrium and embryos may correct the time deviation brought by the slow development of the embryos, which had similar goal as this study. However, in that study, there was signi cant difference in infertile factors caused by ovarian reserve function between the delated Day 4 embryo group and the normal Day 3 embryo group (44.4% versus 16.4%, p = 0.003), which indicated that the limited comparability between the two groups. In our study, embryos with the same developmental starting point in two groups were thawed and transferred at different times respectively, which could strictly control variables.
In this research, the quality of embryos was evaluated after thawing and overnight-culture and characterized as top quality, good quality and poor quality. In order to exclude the effect of embryo differences on clinical pregnancy outcome, we evaluated the quality of embryos which gained clinical pregnancy and found that there were no signi cant differences between embryos with top quality, good quality or poor quality, which indicated that these pregnancy outcomes were not caused by embryonic factors. At the same time, our result exhibited that there was no signi cant difference in the miscarriage rate between two thawing strategies, while higher implantation rate was shown in early thawing group, which illustrated that the improvement of clinical pregnancy rate was achieved by increasing implantation rate instead of reducing miscarriage rate.
To the best of our knowledge, this is the rst comparison of clinical outcomes between early thawing and routine thawing of day 3 slow-growing embryos. However, there are a few limitations in our study. Firstly, the methods of preparing endometrium were signi cantly different in the current study. A randomized controlled trial reported that similar outcomes shown in different types of endometrial preparation strategies for FET cycles, included natural cycle, natural cycle with ovulation stimulation cycle, HRT and GnRH-a pre-treatment plus HRT cycle [19], there is no current evidence that higher clinical outcomes can be obtained by using a certain method for the preparation endometrium of FET cycles [15,20]. Besides, as showed on our multivariate regression analysis showed, endometrial preparation protocol inconsistency did not affect clinical pregnancy outcome. Secondly, single-center retrospective analysis limited the strength of the evidence of the current conclusion.

Conclusions
In conclusion, higher pregnancy outcomes obtained by early thawing of slow-growing embryos deserve clinicians' attention and is of guiding value for the determining of thawing and transferring programs. Higher quality randomized controlled studies are needed in the future to further con rm the role of early thawing and transfer of slow-growing embryos.

Declarations
Availability of data and materials Not applicable.   1 Data are expressed as mean values+ standard deviation (SD). 2 The data is shown as the ratio of the mentioned value to the total number of samples (percentages between brackets).