Is large for gestational age in singletons born after frozen embryo transfer associated with freezing technique or endometrial preparation protocol? A longitudinal national French study

Study question: Is large for gestational age (LGA) observed in babies born after frozen embryo transfer (FET) associated with either the freezing technique or the endometrial preparation protocol?

Summary answer: Artificial cycles are associated with a higher risk of LGA, with no difference in rate between the two freezing techniques (vitrification versus slow freezing) or embryo stage (cleaved embryo versus blastocyst).

What is known already: Several studies have compared neonatal outcomes after fresh embryo transfer (ET) and FET and shown that FET is associated with improved neonatal outcomes, including reduced risks of preterm birth, low birthweight, and small for gestational age (SGA), when compared with fresh ET. However, these studies also revealed an increased risk of LGA after FET. The underlying pathophysiology of this increased risk remains unclear; parental infertility, laboratory procedures (including embryo culture conditions and freezing-thawing processes), and endometrial preparation treatments might be involved.

Study design, size, duration: A multicentre epidemiological data study was performed through a retrospective analysis of the standardized individual clinical records of the French national register of IVF from 2014 to 2018, including single deliveries resulting from fresh ET or FET that were prospectively collected in fertility centres. Complementary data were collected from the participating fertility centres and included the vitrification media and devices, and the endometrial preparation protocols.

Participants/materials, setting, methods: Data were collected from 35 French ART centres, leading to the inclusion of a total of 72 789 fresh ET, 10 602 slow-freezing FET, and 39 062 vitrification FET. Main clinical outcomes were presented according to origin of the transferred embryos (fresh, slow frozen, or vitrified embryos) and endometrial preparations for FET (ovulatory or artificial cycles), comparing five different groups (fresh, slow freezing-ovulatory cycle, slow freezing-artificial cycle, vitrification-ovulatory cycle, and vitrification-artificial cycle). Foetal growth disorders were defined in live-born singletons according to gestational age and sex-specific weight percentile distribution: SGA and LGA if <10th and ≥90th percentiles, respectively. Analyses were performed using linear mixed models with the ART centres as random effect.

Main results and the role of chance: Transfers led to, respectively, 19 006, 1798, and 9195 deliveries corresponding to delivery rates per transfer of 26.1%, 17.0%, and 23.5% after fresh ET, slow-freezing FET, and vitrification FET, respectively. FET cycles were performed in either ovulatory cycles (n = 21 704) or artificial cycles (n = 34 237), leading to 5910 and 10 322 pregnancies, respectively, and corresponding to pregnancy rates per transfer of 31.6% and 33.3%. A significantly higher rate of spontaneous miscarriage was observed in artificial cycles when compared with ovulatory cycles (33.3% versus 21.4%, P < 0.001, in slow freezing groups and 31.6% versus 21.8%, P < 0.001 in vitrification groups). Consequently, a lower delivery rate per transfer was observed in artificial cycles compared with ovulatory cycles both in slow freezing and vitrification groups (15.5% versus 18.9%, P < 0.001 and 22.8% versus 24.9%, P < 0.001, respectively). Among a total of 26 585 live-born singletons, 16 413 babies were born from fresh ET, 1644 from slow-freezing FET, and 8528 from vitrification FET. Birthweight was significantly higher in the FET groups than in the fresh ET group, with no difference between the two freezing techniques. Likewise, LGA rates were higher and SGA rates were lower in the FET groups compared with the fresh ET group whatever the method used for embryo freezing. In a multivariable analysis, the risk of LGA following FET was significantly increased in artificial compared with ovulatory cycles. In contrast, the risk of LGA was not associated with either the freezing procedure (vitrification versus slow freezing) or the embryo stage (cleaved embryo versus blastocyst) at freezing. Regarding the vitrification method, the risk of LGA was not associated with either the vitrification medium used or the embryo stage.

Limitations, reasons for caution: No data were available on maternal context, such as parity, BMI, infertility cause, or maternal comorbidities, in the French national database. In particular, we cannot exclude that the increased risk of LGA observed following FET with artificial cycles may, at least partially, be associated with a confounding effect of some maternal factors. No information about embryo culture and incubation conditions was available. Most of the vitrification techniques were performed using the same device and with two main vitrification media, limiting the validity of a comparison of risk for LGA according to the device or vitrification media used.

Wider implications of the findings: Our results seem reassuring, since no potential foetal growth disorders following embryo vitrification in comparison with slow freezing were observed. Even if other factors are involved, the endometrial preparation treatment seems to have the greatest impact on LGA risk following FET. FET during ovulatory cycles could minimize the risk for foetal growth disorders.

Study funding/competing interest(s): This work has received funding from the French Biomedicine Agency (Grant number: 19AMP002). None of the authors has any conflict of interest to declare.

Trial registration number: N/A.