We found that both day of vitrification and quality of the inner cell mass at vitrification were significantly associated with odds of live birth in euploid embryos. In contrast, the quality of the trophoblast at vitrification, the sex of the embryo, and whether an embryo was selected for transfer based on sex were not associated with odds of live birth.

Prior studies of untested embryos demonstrated that embryos that blastulate and are vitrified earlier have more favorable pregnancy outcomes. When frozen embryos transferred after vitrification on day 7 were compared with those that were vitrified by days 5 or 6, day 7 embryos had lower rates of implantation, clinical pregnancy, and live birth [22, 23]. Within studies that compared untested embryos that were vitrified on days 5 versus 6, live birth rates were also higher in the embryos that were vitrified earlier for both fresh and frozen cycles [6, 24, 25]. Additional studies have evaluated live birth outcomes in euploid embryos with varied results. Irani et al. and Boynukalin et al. found that among single frozen euploid transfers, embryos vitrified on day 6 were associated with lower live birth rate than those vitrified on day 5, which is consistent with the results of our study [26, 27]. However, Shear et al., Ji et al. and Gonzalez et al. did not find a significant association between day of vitrification (day 5 vs day 6) and live birth in euploid embryos [28,29,30]. This is in concordance with results reported by Capalbo et al., who found that blastocyst developmental rate was not predictive of implantation rate [31]. Thus, our results are in agreement with some previous studies of both untested and euploid embryos, although there is discordance amongst existing literature. It is possible that these discrepancies are Due to methodological differences. Gonzalez et al. focused their study solely on embryos that underwent PGT-A through NGS, while our study includes PGT-A conducted via both NGS and SNP-array; furthermore, their cohort was of higher maternal age and of smaller sample size than our study. The authors also noted that their initial analysis suggested day 6 euploid embryos had lower live birth rate than day 5 euploid embryos, although this lost significance in the final analysis. The studies led by Ji et al. and Capalbo et al. both had smaller cohort sizes, which may have limited the statistical power of the studies.

Our data also showed an association between live birth and ICM quality, but not TE quality, demonstrating that not all measures of embryo morphology were associated with odds of live birth. These results align with prior studies that have found ICM to be either the only or the most significant morphological predictor of ongoing pregnancy and live birth in euploid embryos following PGT-A [4, 29, 32, 33]. In contrast, other studies have reported the superiority of TE grading in predicting live birth across both untested and euploid embryos. In a study of fresh blastocyst transfers, Ahlstrom et al. found TE quality to be the only independent predictor of live birth when analyzed along with ICM and blastocele expansion degree, although this study included fresh blastocyst transfers with unknown ploidy [34]. In a study of euploid embryos, Rienzi et al. similarly found an association between TE quality and live birth, with ICM remaining statistically insignificant [35]. Yet other studies of euploid embryos have found no significant association between either measure of morphology or live birth [26, 30, 31]. Improved grading standardization and subsequent additional research using more standardized grading would likely help better characterize the role of morphology in predicting outcomes in euploid embryos.

Although the association between pregnancy outcomes and both day of vitrification and morphology have been previously evaluated, few studies have investigated whether one measure should be prioritized over the other when selecting an embryo for transfer. Understanding whether transferring an embryo vitrified on day 5 with less ideal morphology leads to better outcomes than transferring an embryo vitrified on day 6 with better morphology is important for counseling of patients who want the highest chance of a success after a single embryo transfer. In a study of frozen transfers of untested embryos, Shi et al. concluded that day 6 embryos with a better morphology had a higher live birth rate than day 5 embryos with less ideal morphology [36]. Our study similarly found that embryos with higher ICM scores that were vitrified on later days were associated with a higher odds of live birth than embryos with lower ICM quality that were vitrified on day 5 when comparing within both male and female embryos (Supplemental Table 1). However, in another study of only euploid embryos, Gordon et al. reported that day 5 embryos with less ideal morphology had similar ongoing pregnancy and live birth rates compared to day 6 embryos with better morphology, although the former group had higher rates of miscarriage [9].

While many studies have investigated the influence of PGT-A on sex ratio, few have studied the association between embryo sex and pregnancy outcomes among euploid embryos. In a large-scale analysis of births following IVF, Shaia et al. found that the use of PGT increased the chance of having a male infant in comparison to those who did not use PGT even after excluding transfers conducted for sex selection [37]. This finding is corroborated by Kulmann et al. and Zhang et al., who reported a greater proportion of male births following PGT-A in the absence of sex selection [38, 39]. The authors attributed these results to the extended culture of PGT-A favoring male embryo development, as well as potential genetic factors that favor male embryo survival [37, 38]. This observation may also be explained by studies showing that trophectoderm and general morphological scoring were associated with male embryos within both untested and euploid embryos, resulting in more male embryos chosen based on morphologic factors [38, 40,41,42,43,44]. Our study also documented a difference between embryos of different sex when evaluating the association between morphologic measures and live birth. Day of vitrification was the only morphologic feature significantly associated in male embryos, which is a departure from our finding that ICM grade is associated with live birth amongst all embryos. In contrast, only ICM grading was significantly associated with live birth in female embryos. This suggests a possible inherent difference in development and morphology between female and male embryos, as has been reported in previous literature, which may explain why the differing associations with morphology in male and female embryos did not translate to differences in live birth [37, 38]. However, it is difficult to elucidate why these different associations between male and female embryos did not influence pregnancy outcomes, and further investigations with larger sample sizes are needed to better understand the relationship between sex and embryo development, morphology, and pregnancy outcomes after embryo transfer.

An analysis conducted by Bakkensen et al. found that allowing sex selection after PGT-A did not impact the sex ratio of offspring, although the specific association between sex selection and live birth was not investigated [20]. Previous literature has also demonstrated that implantation and live birth rates are similar between male and female euploid embryos, and that untested embryos of both sexes are equally likely to be euploid and develop at similar rates [16, 17]. Similarly, our study found neither embryo sex nor whether an embryo was selected for transfer based on sex to be significantly associated with pregnancy outcomes. This finding is valuable for guiding counseling for parents who may have a preference for which sex embryo to transfer.

The number of embryo transfers evaluated is a strength of this study, which allowed for reasonably precise analyses among different subgroups of our cohort, with most 95% confidence interval for ORs narrower than 0.3–0.4 in width. That said, there are some limitations of this study. We were only able to evaluate outcomes in our single center, which may decrease the generalizability of these results to other clinics. The subjective nature of morphologic grading and the widespread use of the Gardner criteria for morphological grading outside of our center increase the difficulty of extrapolating our results to other facilities [45]. Furthermore, a small number of embryos were biopsied and vitrified at other centers. Additionally, factors such as indication for undergoing PGT-A, type of endometrial preparation, and maternal characteristics outside of age and BMI were not adjusted for in this study. While maternal age is a factor that has been shown to influence embryo development and outcomes, we did not adjust for age in this study because prior studies have shown that pregnancy outcomes in PGT-A-tested embryos are similar by maternal age [46, 47]. Studies showing age-dependent differences in pregnancy outcomes have reported conflicting outcomes. A meta-analysis led by Vitagliano et al. demonstrated an association between maternal age > 35 years and a decrease in the composite outcome of ongoing pregnancy rate and live birth rate, independent of embryo ploidy [48]. Conversely, Harris et al. showed that cumulative live birth was significantly less in those with maternal age < 35 years and more likely in those with age 38–40 years, compared with no PGT-A [49]. However, a subgroup analysis limited to FET found that women ages 35–40 had an increase in live birth in cycles using PGT-A compared to morphology alone, while this relationship was insignificant in those ages < 35 [49]. To determine the clinical significance of these findings and to provide clarity to the conflicting evidence, additional investigation into the relationship between maternal age and pregnancy outcomes in PGT-A embryo transfers is critical. The authors also recognize that blastocoele expansion, which was not explored in this study, is a component of embryo grading. However, it was not included in this study as our center’s embryology lab uses a less conventional grading system that was difficult to standardize at the time of data collection. There are also a low number of ICM and TE grade 3 embryos included in this study Due to the dataset available at our center; to mitigate potential biases, we grouped grade 2 and grade 3 embryos together. Lastly, although all attempts are made for day of vitrification to be determined by an embryo’s eligibility for trophectoderm biopsy, laboratory factors may have contributed to some embryos being biopsied and vitrified on an earlier or later day in our clinic than they would have been in another center.

In conclusion, euploid embryos with better ICM quality and those that reached the blastocyst stage earlier were associated with higher odds of live birth, suggesting that prioritizing embryos for transfer by these parameters may provide patients with the best chance of live birth.