Which is better for mothers and babies, fresh or frozen thawed blastocyst transfer?

Background: In recent years, there have been emerging many reports on the pregnancy outcomes of fresh blastocyst transfer (BT) and freeze-thaw BT, but these couclusions are controversial and incomplete. To compare the pregnancy outcomes, maternal complications and neonatal outcomes of fresh and frozen-thawed BT in vitro fertilization or intracytoplasmic sperm injection (IVF/ICSI) cycles, we conducted a meta-analysis. Methods: A meta-analysis was conducted by searching PubMed, Embase, and Cochrane Library until January 2020. Data were extracted independently by two authors. Results: 42 studies, including 12 randomized controlled trials (RCT) met the inclusion criteria. Fresh BT showed lower implantation rate (IR), pregnancy rate (PR), ongoing pregnancy rate (OPR) and higher eptopic pregnancy rate (EPR) compared with frozen-thawed BT consistent with the results of RCT. The risks of moderate or severe ovarian hyperstimulation syndrome (OHSS), placental abruption (PA) and preterm were higher in fresh BT than in the frozen-thawed BT. The risk of pregnancy-induced hypertension (PIH) and pre-eclampsia was decreased in fresh BT , however, no signi�cant differences of risks for PIH, pre-eclampsia, OHSS, and preterm was found between the two group in the 2 RCT included. Compared with frozen-thawed BT, fresh BT appears to be associated with small for gestational age (SGA) and low birth weight (LBW). No differences in the incidences of neonatal mortality and neonatal malformations were observed between fresh and frozen-thawed BT. Conclusions: In summary, Considering the higher IR, PR, OPR, lower EPR, and the decreased risks of OHSS, PA and preterm, as well as the incidences of SGA and LBW in frozen-thawed BT, this meta-analysis indicates that frozen-thawed BT may be a better choice for mothers and babies compared with fresh BT.


Background
As cryopreservation technology develops during the past few decades, the proportion of frozen blastocyst transfer (BT) has increased [1].There have been concerns about the impact of cryopreservation on the pregnancy outcomes, maternal complications and health of born children [2].A few studies have compared the pregnancy outcomes following fresh BT and cryopreserved-thawed BT in patients undergoing IVF/ICSI cycles [3][4][5].However, the ndings are controversial.A recent metaanalysis study supported the hypothesis that single cryopreserved BT might not be the best choice compared with single fresh BT in patients undergoing IVF/ICSI cycles [3].However, another systematic review and meta-analysis study suggest that the pregnancy outcomes may be improved by performing frozen-thawed BT [4].
With regard to maternal complications, Maheshwari et al showed that frozen-thawed BT was associated with decreased risks of postpartum hemorrhage (PH), placental abruption (PA) and placenta previa (PP) and preterm compared with fresh BT, and the pregnancies arising from frozen-thawed BT seem to have lower risks of maternal complications [2].Shavit et al reached at the opposite conclusion that frozen-thawed BT may contribute to increased risk of maternal complications such as preeclampsia and gestational diabetes mellitus (GDM) [6].The latest randomized controlled trial reported that the incidence of pre-eclampsia was higher after frozen-thawed BT than fresh BT, and the risk of moderate or severe ovarian hyperstimulation syndrome (OHSS) was similar in both groups [7].
Considering the neonatal outcomes, an early review demonstrated that there were no signi cant differences in incidences of perinatal death, low birth weight of infants between fresh BT and frozen-thawed BT [8].However, another review suggested that the incidences of small for gestational age (SGA), low birth weight (LBW), and perinatal mortality were lower in women who received frozen thawed BT [2].
There is growing concern on whether children born after frozen thawed BT have increased risks of congenital malformations compared to that after fresh BT in IVF/ICSI cycles.A register-based cohort study suggested that the risk for congenital malformation of the children born after frozen thawed BT was not increased compared with fresh BT, in addition, no increased risks concerning the affected organ system were found between the two groups [9].
Due to the limited sample size, the past meta-analysis conclusions were controversial.With the emergence of new reports, there is an urgent need to perform a meta-analysis to compare the multiple outcomes following fresh BT and frozen-thawed BT to provide guidance for clinical practice.The purpose of this meta-analysis was to examine the pregnancy outcomes, maternal complications and neonatal malformations after frozen thawed BT versus fresh BT in an IVF/ICSI cycle and assess if the frozen thawed BT is a better choice than fresh BT.

Search strategy
We searched the published articles in PubMed, EMBASE and Cochrane Library databases up to August 2019, using the following terms as key words: 'humans', 'embryo*, 'cryo*', 'frozen', 'vitrif*', 'freez*', and 'fresh'.A comprehensive search strategy for MEDLINE was presented in Appendices.

Eligibility Criteria and Data Extraction
We included trials comparing clinical outcomes between patients undergoing IVF/ICSI cycles with fresh or frozen BT.Two researchers assessed the eligibility of studies and extracted the data independently.Any disagreement was resolved by discussion.Study characteristics and outcome data were generated from forty-two eligible studies.

Risk of Bias Assessment
We assessed the risk of bias from included studies following the guidance suggested by the Cochrane Collaboration, regarding the generation of sequence allocation, allocation concealment, blinding, and incomplete outcome data for each trial included in the review.Funnel plots were adopted to investigate whether the difference was due to publication or reporting bias.

Outcome Measures
The pregnancy outcome: Implantation rate re ected the number of gestational sacs seen per embryo transferred.Pregnancy was identi ed through increased serum hCG level within 10 days after blastocyst transfer.Ongoing pregnancy was de ned as pregnancy proceeding beyond the 10th gestational week.Clinical pregnancy was considered as the presence of a gestational sac with fetal heart activity, as assessed by ultrasound at 7 weeks of gestation.Miscarriage included any pregnancy that did not become ongoing pregnancies.Multiple pregnancy was de ned as a gestation with more than one fetus.Live birth was calculated by birthing events per embryo transfer.We recorded the following maternal complications: GDM, pregnancyinduced hypertension (PIH) and pre-eclampsia, moderate or severe OHSS, preterm, PP, PA, PH and preterm rupture of membrane.Preterm was de ned as live births < 37 weeks' gestational age.Very preterm was de ned as live births < 32 weeks' gestational age.The neonatal outcomes included gestational age at delivery, birth weight, stillbirth, perinatal mortality and neonatal mortality.Large for gestational age (LGA) was de ned as birthweight higher than the 90th percentile of referential birthweight.SGA was de ned as birthweight lower than the 10th percentile of referential birthweight.Very small for gestational age (VSGA) was de ned as weighing below the 3rd percentile of referential birthweight.High birth weight baby (HBW) was de ned as weight of > 4000 g at birth.Very high birth weight baby (VHBW) was de ned as weight of > 4500 g at birth.Low birth weight baby (LBW) was de ned as weight of < 2500 g at birth.Very low birth weight baby (VLBW) was de ned as weight of < 1500 g at birth.We also analyzed the neonatal malformations including congenital anomaly and chromosomal aberrations, different organ system malformations.

Statistical analysis
All statistical analysis was conducted using Rev Man software.For the included studies, the dichotomous data results for each of the studies eligible for meta-analysis were expressed as a risk ratio (RR) with 95% con dence intervals (CI).These results were combined for meta-analysis with use of the Mantel/Haenszel model along with the random effects model.
Statistical heterogeneity was assessed with a chi-squared test and quanti ed with the I 2 statistic.An I 2 value greater than 50% may be considered to represent substantial heterogeneity.p < 0.05 was considered statistically signi cant.

Results
A total of 3645 available publications were retrieved in our search.Of these, 3473 were excluded after reading the title and the abstract.Finally, 42 articles, including 12 randomized controlled trials (RCT) and 30 non-randomized controlled trials (NRCT) were considered to be eligible by one or both reviewers (Supplemental Fig. 1).Table 1 gives the details of all included studies.
In conclusion, fresh BT tends to lead to SGA and LBW.Frozen thawed BT has the opposite effect.The stillbirth, perinatal mortality and neonatal mortality showed no statistically signi cant differences between the two groups.

Neonatal malformations
From the data we have summarized, no risk differences in congenital anomaly and chromosomal aberration rates of newborns were detected between fresh BT and frozen-thawed BT (RR 1.06 95% CI 0.97-1.15,P = .19Heterogeneity: I 2 = 0%) (Fig. 6A).Further inspecting the risk of the different organ system malformations in newborns including circulatory system (Fig. 6B), respiratory system (Fig. 6C), nervous system (Fig. 6D), gastrointestinal system (Supplemental Fig. 6A), genitourinary system (Supplemental Fig. 6B), eye, ear, face (Supplemental Fig. 6C), and musculoskeletal system (Supplemental Fig. 6D), no increased risk in frozen-thawed BT were found.The above data indicate that freeze-thaw BT is not a risk factor for neonatal malformations.

Discussion
Great advances have been made in cryopreservation culture technique for embryo since the success of the rst pregnancy of frozen-thawed embryo transfer (FET) in 1983 [10].This technique has been applied as a supplement to IVF and embryo transfer.FET was accepted by every center and has become an essential part of IVF/ICSI treatment.Therefore, the increased use of FET has intensi ed the awareness of the safety of the technique [11].The meta-analysis compared the outcomes of fresh BT and frozen-thawed BT undergoing IVF/ICSI cycles, with comprehensive respects of the pregnancy outcomes, maternal complications, neonatal outcomes and malformations.
With respect to pregnancy outcomes, our study showed that frozen BT was associated with increased IR, PR and OPR and a decreased EPR compared with fresh thawed BT, which was consistent with the results of RCT.There was no difference in CPR, MR, MPR, and LBR.However, according to RCT, no difference in MR, MPR and LBR and a decreased CPR were tested in fresh BT compared with fresh thawed BT.Recently, Zeng et al showed that there was no difference in IR, CPR, MR, and MPR, but an increased LBR was found in fresh BT comparing with cryopreserved thawed BT [3].Roque et al showed frozen-thawed BT signi cantly improved CPR and OPR in patients in IVF/ICSI cycles [4].The incidence of EP between the two groups varied in different studies.The inconsistent conclusions may be related to differences in the data included.A prevous study reported that EPR was higher in frozen thawed BT [12].However, our study suggested that frozen thawed BT was related to lower EPR consistent with these studies [13,14].However, in these reports sub-category analysis wasn't performed according to the stage of embryo transfer including cleavage and blastocyst stage embryos.Hence, in view of the increased IR, PR, OPR and decreased EPR following frozen thawed BT, we believe that frozen thawed BT have a better pregnancy outcome than fresh BT.Embryo implantation is one of the important steps for reproductive success, and implantation failure remains an unsolved problem in IVF/ICSI cycles.The primary responsible cause of failure is the impairment of the endometrial receptivity (ER), whereas the embryo itself is responsible for the failure [15].A study suggested impaired ER is more apt to occur in fresh ET cycles after ovarian stimulation, when compared with FET cycles with arti cial endometrial preparation.Impaired ER apparently accounted for most implantation failures in the fresh group [16].The another explanation for better results in pregnancies subsequent to frozen BT is that the physical effects of freezing and thawing embryos may lter out weaker embryos and allow only good quality ones to survive, resulting in better fetal growth [17].
From the perspective of maternal complications, our research demonstrated that the risks of OHSS, PA and preterm in fresh BT are increased compared to in frozen-thawed BT.On the contrary, the risk of PIH and pre-eclampsia in fresh BT is decreased compared to in frozen-thawed BT.According to 2 RCT, no difference in PIH and pre-eclampsia, OHSS and preterm were found between fresh BT and frozen-thawed BT.Owing to the few numbers of RCT reporting maternal complications, the insu cient evidence may lead to inconsistent results.There were no difference in the GDM and PP of the fresh and cryopreservedthawed BT.OHSS is an iatrogenic condition resulting from an excessive ovarian response to superovulation medication.
According to a previous meta-analysis, no difference was found in OHSS between fresh BT and frozen thawed BT [18].However, the previous data were insu cient.A few recent reviews demonstrated that singleton pregnancies after transfer of frozen thawed embryos were associated with lower risks of preterm birth (< 37 weeks), very preterm birth (< 32 weeks) when compared with those after fresh embryos transfer, which agrees with our research [2,19,20].
In regard to neonatal outcomes, our study suggested that there were lower risks of SGA, LBW in singleton pregnancies after frozen thawed BT compared with fresh BT, which was consistent with the previous meta-analysis [2].However, the stillbirth, and perinatal mortality neonatal mortality is not statically different between two groups.Moreover, with respect of neonatal malformations, there was no difference between fresh BT and frozen thawed BT.In conclusion, singleton pregnancies after frozen thawed BT seem to have better neonatal outcomes than those after fresh BT, owing to lower risks of SGA and LBW.The reasons for better neonatal outcomes of frozen BT compared with fresh BT are not known yet.In contrast to IVF with fresh embryo transfer, FET is usually performed in minimally stimulated or natural cycles.This lowers the risk of SGA and LBW after FET, which may attribute to a luteal phase that mirrors the natural cycle, with favorable effects on the endometrium and early implantation [20].Another probable explanation was that controlled ovarian hyperstimulation (COH) was associated with poorer neonatal outcomes assessed by SGA and LBW in a rent study [21].The results favoring frozen thawed BT instead of fresh BT may relate to the adverse effects of COH on ER [22,23].Therefore, elective cryopreservation of viable embryos could be an alternative to avoid the deleterious effects of COH in embryo endometrium synchrony [16,24].

Strengths and limitations
The major strength of this systematic review is the comprehensive literature search, identifying study objects from a huge number of relevant publications; another strength is the many aspects of pregnancy outcomes, maternal complications and neonatal outcomes the study evaluated between frozen or fresh BT to know which is better for mothers and babies.In addition, we conducted RCT and NRCT meta-analysis respectively to improve the quality of evidences.But the present metaanalysis also has some limitations; one of which was the signi cant heterogeneity about the pregnancy outcome in the metaanalysis.We tried to nd the source of heterogeneity by running a subgroup analysis to examine the source of heterogeneity but failed.Besides, the baseline characteristics of patients differ more or less among the included studies, including countries, age, smoking, duration of infertility, type of infertility, endometrial thickness, and cryopreservation type.

Conclusion
In summary, considering the higher IR, PR, OPR, lower EPR, and the decreased risks of OHSS, PA and preterm, as well as the incidences of SGA and LBW in frozen-thawed BT, this meta-analysis indicates that frozen-thawed BT may be a better choice for mothers and babies in IVF/ICSI cycles compared with fresh BT.

Figure 1 Forest
Figure 1

Figure 2 Forest
Figure 2

Figure 3 Forest
Figure 3

Figure 4 Forest
Figure 4

Figure 5 Forest
Figure 5