As assisted reproductive technology continues to develop, more infertile patients have successfully obtained live births. The question of whether extending the time of embryo culture can provide more benefits for clinical outcomes has been raised. There is no difference in outcomes between embryo transfer on the second day and the third day of the cleavage period, as shown through a systematic review analysis [1]. Most studies have reported that even high-quality cleavage-stage embryos of D2 and D3 may be at risk of abnormal and fertilization developmental arrest as a result of an inactive embryonic genome [2].

The blastocyst embryo transfer (ET) has obvious advantages in both clinical pregnancy rate and implantation rate, and has a lower miscarriage rate [3, 4]. There is low-quality evidence for successful live births and moderate-quality evidence for clinical pregnancies. Fresh embryo transfer during the blastocyst stage has a higher success rate compared to fresh transfer during the cleavage stage [5, 6]. However, it remains uncertain whether embryo transfer during the blastocyst stage can improve the cumulative live birth rate in a single oocyte retrieval cycle. The D5 blastocyst transplantation is also accompanied by a high risk of blastocyst formation failure owing to fluctuations in the culture microenvironment, such as pH, temperature and osmolarity [7, 8, 9, 10, 11]. Other results indicate that the morulae have similar advantages with the blastocysts while embryonic compaction on day 4 is a symbol of embryonic genome activation and morulae have better synchronicity with the endometrial environment [12, 13]. However, morulae were chaotic and difficult to evaluate based on just compaction rate or fragmentation [14, 15, 16, 17]. Extra out-of-incubator observations bring fluctuations to the culture system [5, 11]. These shortcomings limit the clinical application of embryo transfer on the fourth day (D4 ET). In recent years, embryo scoring systems based on the application of embryonic metabolomics and timelapse-based retrospective developmental dynamics enable more efficient evaluation of fourth-day embryos [18, 19, 20, 21].

A retrospective study showed the implantation rate (36.3% vs. 39.6%), clinical pregnancy rate (49.5% vs. 51.9%), and live birth rate (42.1% vs. 45.6%) were statistically insignificant between D4 and D5 ET [13]. Similarly, a systematic review and network meta-analysis found no significant differences in cancellation rates, miscarriage rates, ongoing pregnancy rates, and live birth rates between D4 and D5 embryo transfers [22]. These findings suggest that D4 embryo transfer can be considered a valid option in the decision-making process. However, it is important to note that there may be limitations to these studies, such as variations in study design, the number and quality of transferred embryos, and different culture conditions. Additionally, D4 embryo observations would increase extra out-of-incubator exposure which bring fluctuations to the culture system. The cumulative live birth rate after fresh day 4 transplantation in a single oocyte retrieval cycle requires further observation. In this study, propensity score matching (PSM) was used to balance the influence of intergroup confounders and to compare clinical outcomes between the D4 and D5 ET in in vitro fertilization (IVF)/intracytoplasmic sperm injection (ICSI) cycles. We also compared the cumulative live birth rates and pregnancy outcomes in the first frozen ET cycle following fresh D4 and D5 embryo transfers.

Overall, a total of 1167 patients with their first IVF/ICSI treatment were eligible for analysis. The D4 ET group and D5 ET group consisted of 974 and 193 patients, respectively. After the PSM procedure, a total of 198 patients were included in the D4 ET group and D5 ET group. The baseline characteristics, ET variables, and pregnancy outcomes before and after PSM were evaluated.

IVF-ET can obtain enough oocytes for embryo transfer by promoting ovulation [25]. Although the more oocytes and the more number of available embryos, the conversion rate of available embryos decreases [26]. In other words, most embryos do not continue to develop under the combination of genes and the environment [27]. The development of sequential culture media represents a determinant factor in sustaining extended embryo culture [26]. An increasing number of studies have shown that prolonged time in vitro culture to obtain more embryo characteristics is able to obtain better live birth rates relative to cleavage-stage embryo transfer [3, 27, 28, 29]. The embryonic self-genome activation is an important symbol of independent embryo development starting on the third day after insemination [30]. High-quality cleavage embryos have an opportunity of 40% to format transferable blastocysts [31]. Both day 4 morula and day 5 blastocyst are features of embryos able to develop independently [30]. The D5 blastocyst transplantation is also accompanied by a high risk of blastocyst formation failure, owing to fluctuations in the culture microenvironment, such as pH, temperature and osmolarity [7, 8, 9, 10, 11]. Other results indicate that the morulae have similar advantages with the blastocysts while embryonic compaction on day 4 is a symbol of embryonic genome activation and morulae have better synchronicity with the endometrial environment [12, 13].

Most studies have considered D4 ET as a flexible transplant regimen, with clinical pregnancy rates similar to D5 ET [12, 13, 20]. It seems that extending embryo culture has the potential to provide more benefits for clinical outcomes.

Therefore, the strategy of fresh embryo transfer in our center is that if the number of embryos assessed as good or fair on the third day was more than the intended number of embryos to be transferred, we selected the most viable morulae for transplantation on the fourth day. If there were at least four good embryos, then we selected blastocysts for transplantation on the fifth day. This explained why the two groups before PSM data were quite different showing a large deviation. It became obvious that patients had lower AFC, estradiol on hCG day, FORT, number of oocytes retrieved, number of normal fertilization, number of transferable embryos, and number of high-quality embryos on day 3 in the D4 ET group before matching. The D4 ET group showed a higher live birth rate (54.21% vs. 44.88%, p = 0.015), with a lower birth rate of one embryo transferred (21.36% vs. 43.01%, p 0.001) before PSM. Basic data and single embryo transfer rate affected the comparison of live birth rate between the two groups. Multivariate logistic regression analysis showed that D4 ET increased live birth rate in fresh cycles relative to D5 ET before PSM (OR = 1.552, 95% CI: 1.036~2.323, p = 0.033). After PSM, these characteristics were similar in the two groups, except for the number of high-quality embryos on day 3, which was lower in the D4 ET group (3 [2, 3.5] vs. 4 [2, 4], p = 0.035). The live birth rate of D4 ET tended to increase (56.57% vs. 45.37, p = 0.108). High-quality embryos are generally considered to be a favorable factor for a successful live birth. All the results obtained imply that D4 ET has similar effectiveness and feasibility relative to D5 ET.

The important drawback that affects D4 ET still remains making it a less-preferable approach is that extra out-of-incubator observations bring fluctuations to the culture system, although the main reason is the difficulty in evaluating the D4 morula [32]. Temperature decline in embryological culture dishes was time-dependent outside the incubator [9]. The reduction of the observation frequency to four times (on days 1, 3, 5 and 6) could enhance embryo quality and blastocyst formation rate relative to the daily observation after insemination (day 1 to day 6; six times) [10]. Even disruptions in culture conditions as simple as incubator door opening leads to measurable, significant changes in morphokinetics [11, 33]. Clearly, D4 embryo observation will increase extra out-of-incubator exposure which bring fluctuations to the culture system. Unexpectedly, in this study no change was found in subsequent blastocyst formation rate (33.57 vs. 34.05, p = 0.618; 35.10 vs. 33.80, p = 0.468) and cumulative live birth rate (70.02 vs. 73.58, p = 0.322; 69.70 vs. 72.73, p = 0.638) between the two groups before and after PSM in the fresh cycles. The fluctuations caused by D4 ET had a minimal effect on embryonic development, possibly due to the application of time-lapse imaging (TLI). TLI can offer more information about embryo development compared with static observations and is expected to enhance the identification of good-prognosis embryos for clinical application profiting from continuous embryo monitoring in an undisturbed environment [19, 34, 35]. There was no significant difference in endometrial thickness(8.8 [8, 10] vs. 8.9 [8, 9.6], p = 0.689; 8.6 [8, 10] vs. 8.9 [8, 9.7], p = 0.993), one embryo transfer rate(28.35 vs. 25.84, p = 0.639; 22.86 vs. 24.44, p = 0.724), clinical pregnancy rate (54.88 vs. 61.80, p = 0.243; 57.14 vs. 73.33, p = 0.129), and live birth rate (43.90 vs. 50.56, p = 0.263; 45.71 vs. 55.56, p = 0.382) between the two groups before and after PSM in the first frozen ET cycles after fresh ET.

From these results, we concluded that D4 ET had a similar clinical outcome as compared with D5 ET. However, it is important to note that this particular study is a retrospective study. The number of participants after PSM was relatively low, which could limit the generalizability of the findings. Moreover, when analyzing the data, there was always the possibility of the role of chance, where the observed results may be due to random variation rather than true association. Therefore, it was important to interpret the results of this study with caution and consider these limitations when drawing conclusions. A prospective study is necessary to further compare D4 and D5 embryo transfer. Additionally, D4 embryos were only evaluated based on fragmentation and compaction rates. In future studies, we aim to propose an embryonic evaluation system that combines developmental dynamics and morphology. This has the potential to promote the wider and more effective application of D4 embryo transfer in clinical settings, rather than it being considered as an alternative option.

In conclusion, D4 ET had no significant adverse effect on clinical outcome in fresh cycles and first frozen ET cycles relative to D5 ET, although difficulties in grading a morula stage embryo and fluctuations to the culture system from out-of-incubator observations may alter our results.

The datasets generated and analyzed during the current study are not publicly available, since the dataset will be used for other retrospective analyses. The data are available from the corresponding author upon reasonable request.

YG, BZ and SW contributed to conception and design of the study. FD and HD organized the database. MD, LT and HD performed the statistical analysis. YG and HD wrote the first draft of the manuscript. FD, YG, LT and BZ wrote sections of the manuscript. All authors contributed to manuscript revision. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

All procedures used in this study were conducted in accordance with the principles of conducting experiments with human participants as outlined in the Declaration of Helsinki. This study adopted the consent of the Ethics Committee of Xingtai Infertility Specialized Hospital (approval number: 2021-04). All process in IVF-ET were obtained by written informed consent.

We would like to express our gratitude to all those who helped us during the writing of this manuscript. Thanks to all the peer reviewers for their opinions and suggestions.

The work was supported by the Health Commission of Hebei Province (Grant No. 20221863).

The authors declare no conflict of interest.