Pregnancy Outcomes for Single Frozen-thawed Blastocyst Transfer of Day 5 and Day 6 With Different Quality: a Large Matched-cohort Study

Background: Whether the day of blastocyst expansion affects pregnancy outcomes in frozen-thawed blastocyst transfer (FBT) cycles remains controversial. The conicting results may be attributed to the potential confounders of FBT cycles between day 5 and day 6. The aim of this study is to compare the clinical pregnancy rate (CPR) and live birth rate (LBR) of FBT cycles on day 5 versus day 6 accounting for the baseline comparability and determine the transfer order of FBT cycles. Methods: This was a retrospective match-cohort study, FBT cycles of day 5 and day 6 were matched by the general characteristics of patients, 2207 matched pairs of FBT cycles were included from January 2016 to December 2019 in our reproductive medicine center. Results: The CPR and LBR of day 5 blastocyst were signicantly higher than day 6 with the same quality. In the FBT cycles with good quality, the CPR of day 5 and day 6 was 61.30%, 57.56% respectively (P=0.045), the LBR was 44.79%, 36.16% respectively (P<0.001). While in the FBT cycles with poor quality, the CPR of day 5 and day 6 was 48.61%, 40.89% respectively (P=0.006), the LBR was 31.71%, 25.74% respectively (P=0.019). The CPR of FBT cycles on day 6 with good quality was statistically higher than day 5 with poor quality (57.56% versus 48.61%, P=0.001). Maternal age, AMH, endometrial thickness, embryo quality and the day of blastocyst expansion was independently correlated with CPR and LBR. The FBT cycles of day 5 exhibited signicantly higher CPR [adjusted odds ratio (OR)=1.246, 95% condence intervals (Cl): 1.097-1.415, P=0.001] and LBR (adjusted OR=1.435, 95% Cl: 1.258-1.637, P<0.001) compared with day 6. Conclusions:

Results: The CPR and LBR of day 5 blastocyst were signi cantly higher than day 6 with the same quality.
In the FBT cycles with good quality, the CPR of day 5 and day 6 was 61.30%, 57.56% respectively (P=0.045), the LBR was 44.79%, 36.16% respectively (P<0.001). While in the FBT cycles with poor quality, the CPR of day 5 and day 6 was 48.61%, 40.89% respectively (P=0.006), the LBR was 31.71%, 25.74% respectively (P=0.019). The CPR of FBT cycles on day 6 with good quality was statistically higher than day 5 with poor quality (57.56% versus 48.61%, P=0.001). Maternal age, AMH, endometrial thickness, embryo quality and the day of blastocyst expansion was independently correlated with CPR and LBR. The Conclusions: The embryo quality should be the primary indicator in the FBT cycles. The blastocyst of day 5 should be preferred if the embryo quality of day 5 and day 6 is equal.
Trial registration: Not applicable.

Backgroud
Blastocyst culture has also been widely used in most reproductive centers because of the better development potential versus cleavage embryo [1]. Meanwhile, single embryo transfer can reduce the risk of multiple pregnancy without sacri cing pregnancy rate signi cantly [2,3]. Therefore, single blastocyst embryo transfer seems to be the best choice for the patients following assisted reproductive technology treatment [4,5].
Transfer the blastocyst with best development potential is an essential step to achieve successful gestation. As we all know, embryo quality was positively associated with the pregnancy outcomes [6,7]. Besides the embryo quality, numerous studies have aimed to compare the pregnancy outcomes of the blastocyst expanding on day 5 and day 6. De nitive evidence has shown that the pregnancy rate of day 5 blastocyst transfer was signi cantly higher than day 6 blastocyst transfer in fresh cycles [8]. Current studies proposed that the decreased pregnancy rate might attribute to the embryo-endometrium asynchronization of day 6 blastocyst [9,10].
Recent studies have shown better pregnancy outcomes in frozen-thawed embryo transfer cycles compared with the fresh cycles in normal responders [11][12][13][14]. This improvement of pregnancy rate in frozen-thawed embryo cycles may attribute to the better embryo-endometrium synchronization, which was impaired in fresh cycles [11,[14][15][16]. However, whether the day of blastocyst expansion affects pregnancy outcomes in frozen-thawed blastocyst transfer (FBT) cycles remains controversial. Some reports showed that the pregnancy rate was signi cantly higher of blastocyst cryopreserved on day 5 than day 6 [17][18][19][20]. While other studies reported equivalent clinical outcomes between day 5 and day 6 blastocyst in FBT cycles [21,22]. The con icting results of the above reports may be attributed to the potential confounders of FBT cycles between day 5 and day 6, such as maternal age at retrieval, body mass index (BMI), type of infertility, ovarian reserve (basal follicle-stimulating hormone (FSH), anti-Müllerian hormone (AMH), antral follicle count (AFC)), endometrial thickness, the number of embryos transferred and blastocyst grade at cryopreservation.
In our present study, we performed a retrospective matched-cohort comparison with a large sample size, taking into account the potential confounding factors mentioned above. The clinical pregnancy rate (CPR) and live birth rate (LBR) of FBT cycles were compared between embryos cryopreserved on day 5 and day 6. Additionally, subgroup analysis strati ed by embryo quality was performed to compare the CPR and LBR. We expect that our results could guide clinical practice to select the blastocyst with best development potential and improve the outcomes of assisted reproductive technology.

Study design
We conducted a retrospective match-cohort study that included single FBT cycles from January 2016 to December 2019 in our reproductive medicine center. This study was approved by the Institutional Review Board of Tongji Hospital. All data were collected from the electronic medical record, the FBT cycles which met the following inclusion criteria were included: (1) cryopreservation of blastocyst only by vitri cation, (2) endometrial preparation with hormone replacement treatment, (3) only one blastocyst was transferred. FBT cycles with missing data and that generated from preimplantation genetic screening (PGS) cycles were excluded.
The included FBT cycles were divided into day 5 and day 6 group according to the day of cryopreservation. The potential confounding factors including maternal age at retrieval, BMI, type of infertility, ovarian reserve (basal FSH, AMH, AFC) and endometrial thickness were matched between the two groups. Data related to pregnancy outcomes and neonatal outcomes were also obtained.

Embryo culture
Oocytes were inseminated either by conventional in vitro fertilization (IVF) or intra-cytoplasmic sperm injection (ICSI). On day 3 of fertilization, one or two cleavage embryos with the best quality were transferred. If the transfer could not be performed in speci c condition, one or two with the best quality were vitri cated. Other supernumerary embryos were cultured until day 5 or day 6 in sequential media regardless of their quality. On day 5, only fully expanded blastocyst were transferred or cryopreserved, and the rest were cultured until day 6. On day 6, all the fully expanded blastocyst were preserved. All the treatments were performed after written informed consent were signed. Blastocyst were graded on day 5 and day 6 according to the degree of blastocoel cavity and the morphology of inner cell mass and trophectoderm, as described by Gardner scoring system [23]. Blastocyst with Gardner scoring of 3BB or better were considered good quality.

Blastocyst vitri cation and thawing procedures
The blastocyst were treated with laser to induce arti cial collapse before vitri cation. After blastocoel collapse, the blastocyst were exposed to equilibration solution for 5-10 min at room temperature. Then, the blastocyst were transferred to vitri cation solution and incubated for 1 min, loaded onto Crotop and plunged into liquid nitrogen immediately.
Blastocyst with the best quality was thawed 2-4 h before embryo transfer. The blastocyst were immediately immersed in thawing solution prewarmed to 37°C for less than 1 min. Then, the blastocyst were transferred to dilution solution and incubated for 3 min at room temperature. After the incubation in No.1 washing solution for 5 min at room temperature, the blastocyst were transferred to No.2 washing solution for 5 min at 37°C, and then transferred to culture dishes with pre-equilibrated blastocyst medium. The solutions for vitri cation and thawing all purchased from Kitazato (Japan).

Endometrial programming and blastocyst transfer
Patients started with an oral administration of estradiol (Progynova, Bayer) at a dosage of 2 mg/day for endometrial preparation. The dosage of estradiol was increased to 6 mg/day gradually. A transvaginal ultrasound assessment was performed to exam the endometrial thickness and adjust the dosage of estradiol about 10 days later. When conditions were appropriate (endometrial thickness reached 7 mm or more), intramuscular progesterone (40 mg) was administrated. Both day 5 and day 6 blastocyst were scheduled to be transferred on the day 6 of progesterone exposure.

Outcome measures
The primary outcome was clinical pregnancy, which was de ned as the observation of a gestational sac with fetal heart pulsations on ultrasound scanning 4 to 5 weeks after blastocyst transfer. Live birth rate was de ned as a delivery of a viable infant at ≥28 weeks of gestation, which was the secondary outcome in our study. The neonatal outcomes included gestational age, birth weight and gender ratio.

Statistical analysis
All statistical analysis were performed using SPSS 26.0, and data were presented as the mean± standard deviation (SD). Comparison of continuous variables was conducted using Student's t test. Person's χ2 test was performed to compare qualitative variables. Multivariate logistic regression analysis was used to identify the potential factors that might in uence the clinical outcomes of FBT cycles. Age at retrieval, BMI, type of infertility, ovarian reserve (basal FSH, AMH, AFC), endometrial thickness, embryo quality, and the day of blastocyst expansion were all taken into consideration in this model. Odds ratios (OR) and 95% con dence intervals (Cl) were calculated, P<0.05 was de ned as statistical signi cance.

Patient characteristics
A total of 2207 matched pairs of FBT cycles were included. The patient characteristics were presented in Table 1. The baseline characteristics including maternal age, BMI, type of infertility, basal FSH, AMH, AFC and endometrial thickness were comparable between the FBT cycles of day 5 and day 6 group. However, the embryo quality showed signi cant difference between the two groups. The rate of good quality embryo in day 5 group (80.43%) was much higher than day 6 group (51.25%) (P < 0.001). The good-quality blastocyst was de ned as Gardner scoring was 3BB or better (B3-B5, AA, AB, BA, BB). Poor quality was de ned as B3-B5 embryo worse than 3BB (BC, AC).
Data are presented as the mean ± standard deviation (SD) or n, number in the bracket is the percentage of cases.
Comparison of continuous variables is conducted using Student's t test and qualitative variables using Person's χ2 test.

Pregnancy outcomes
As shown in Table 2, the CPR of day 5 group (58.81%) was signi cantly higher than day 6 group (49.43%) (P < 0.001). The LBR of day 5 group also showed obvious advantage compared with day 6 group (42.23% versus 31.08%, P < 0.001). Furthermore, the FBT cycles of day 5 and day 6 group were divided into two subgroups respectively, basing on embryo quality. Our results illustrated that the CPR of day 5 group was signi cantly higher than day 6 group regardless of embryo quality (Fig. 1). In good-quality subgroup, the CPR of FBT cycles on day 5 and day 6 was 61.30%, 57.56% respectively (P = 0.045), the LBR was 44.79%, 36.16% respectively (P < 0.001). While in poor-quality subgroup, the CPR of FBT cycles on day 5 and day 6 was 48.61%, 40.89% respectively (P = 0.006), the LBR was 31.71%, 25.74% respectively (P = 0.019). Interestingly, the CPR of FBT cycles on day 6 with good quality was statistically higher than day 5 with poor quality (57.56% versus 48.61%, P = 0.001). The LBR of FBT cycles on day 6 with good quality (36.16%) was also higher than day 5 with poor quality (31.71%), but the difference was not statistically signi cant (P = 0.099). Data are presented as the mean ± standard deviation (SD) or n, number in the bracket is the percentage of cases, unless gender ratio.
Comparison of continuous variables is conducted using Student's t test and qualitative variables using Person's χ2 test.

Neonatal outcomes
The neonatal outcomes were described in Table 2, no signi cant difference was observed in the birth weight, gestational age and gender ratio between the FBT cycles of day 5 and day 6 (P = 0.62, P = 0.37, P = 0.55 in sequence)

Discussion
Single FBT has been a widely accepted by increasing number of reproductive centers, and it is important to choose the blastocyst with best development potential to transfer. As known, endometrium receptivity is critical for both implantation and gestation [24]. In our study, only the FBT cycles using hormone replacement treatment for endometrial preparation were included, which enhanced embryo-endometrium synchronization. We conducted a matched-cohort study to investigate the embryonic factor of blastocyst development potential, including the day of blastocyst expansion and embryo quality. To the best of knowledge, this is the rst matched-cohort analysis to compare the pregnancy potential of frozen-thawed blastocyst cryopreserved on day 5 versus day 6.
The comparison of pregnancy outcomes of FBT cycles showed that the CPR and LBR of day 5 were signi cantly higher than day 6, which consisted with most previous study [17-20, 25, 26]. Furthermore, the FBT cycles of day 5 and day 6 were divided into two subgroups according to the embryo quality (good or poor). The signi cant advantage of FBT cycles on day 5 was still maintained after accounting the embryo quality in our present study. However, whether the blastocyst with good quality cryopreserving on day 5 had superior pregnancy potential was still under debate. Some original articles showed higher pregnancy potential of day 5 blastocyst than day 6 [17,19], while others exhibited equal pregnancy potential between day 5 and day 6 [22,26]. Only a few research have performed a comparison of the pregnancy outcomes of FBT cycles with poor embryo quality, and both suggested the superiority of blastocyst on day 5 [19,26].
Besides the original articles, two systematic reviews also analyzed the different outcomes of FBT cycles on day 5 and day 6 [20,25]. On the whole, both of them approved of the superiority of blastocyst on day 5. However, controversy on the pregnancy outcome of blastocyst cryopreserved on day 5 versus day 6 with the same quality still exited. This inconsistent result partly caused by the cryopreservation method of the studies they included. Slow-freezing method was more widely used than vitri cation ten years ago or even earlier. This hypothesis was also con rmed by Bourdon  Based on the literatures and our present results, we conclude that the FBT cycles of day 5 showed higher pregnancy potential than day 6 with the same embryo quality. However, an interesting result was also found in our study. The good-quality blastocyst of day 6 showed higher pregnancy potential than poorquality blastocyst of day 5, which indicate embryo quality was an important factor for successful gestation. Limited studies have reported the speci c data we concerning, even showed con ict results. The study of Yang et al. showed higher CPR of blastocyst on day 6 with good quality than day5 with poor quality [26], which was similar with our results. However, the study of Ferreux et al. supported the superiority of blastocyst on day 5 even with poor quality [19]. These con ict results may attribute to the heterogeneity of the number in FBT cycles. We only analyzed the single FBT cycles, while some studies also included the frozen-thawed cycles of transferring two blastocysts. In the routine operation, clinicians tended to transfer two blastocysts if the embryo quality were poor. This may partly increase the successful gestation of blastocyst with poor quality. Additionally, the embryo culture strategy such as the number of cleavage embryo transfer, whether cleavage embryo cryopreservation and blastocyst transfer were performed in fresh cycles may also affect the clinical outcomes of following FBT cycles.
Besides the day of blastocyst expansion, the results of multivariate logistic regression analysis showed that maternal age at retrieval, AMH, endometrial thickness and embryo quality was independently correlated with CPR and LBR. It was indisputable that maternal age and embryo quality were closely related to successful gestation. Age had a negative effect on pregnancy chance [27][28][29][30], while morphological embryo quality was positively associated with the pregnancy outcomes [6,7], and our data also reinforced it (Fig. 1). AMH is also an independent predictor of live birth outcome [31,32]. In addition, some studies reported that increased endometrial thickness is associated with better pregnancy outcome [33,34].
Furthermore, the neonatal outcomes were also analyzed in our study. No signi cant difference was observed in the birth weight, gestational age or gender ratio between the FBT cycles of day 5 and day 6. Our results were consistent with those reported by Wang et al. [35].
As a retrospective analysis, there some unavoidable limitations in our study. Such as the difference of operators, culture system (Vitrolife or Cook) may produce biases. But the operators and culture medium were xed during the same period, which were exactly the same in the FBT cycles of day 5 and day 6.

Conclusions
Based on the results of this large matched-cohort study, we conclude that both the day of blastocyst expansion and embryo quality independently correlated with the pregnancy outcomes. The blastocyst cryopreserved on day 5 exhibited higher CPR and LBR than day 6 with the same embryo quality. But the good-quality blastocyst of day 6 had higher chance of successful gestation than poor-quality blastocyst of day 5. Therefore, the embryo quality should be the primary indicator in the FBT cycles, and the blastocyst of day 5 should be preferred if the embryo quality is equal. Availability of data and materials The datasets generated and analyzed during the current study are not publicly available due to the privacy of patients but are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests. designed the study and revised the manuscript critically. All authors have approved the nal article.