The impact of endometrial preparation on pregnancy loss in vitrified-warmed euploid blastocyst transfer cycles: a randomized controlled trial

STUDY QUESTION: Does artificial cycle (AC) endometrial preparation result in increased pregnancy loss rates in patients undergoing vitrified-warmed euploid blastocyst transfer (frozen embryo transfer (FET))? SUMMARY ANSWER: The randomized controlled trial (RCT) demonstrates that artificial endometrial preparation is associated with significantly higher early pregnancy loss (EPL) compared to natural cycles when transferring euploid embryos. WHAT IS KNOWN ALREADY: The increased use of FET cycles has fueled discussion regarding the optimal endometrial preparation method. While earlier reviews found no significant differences in pregnancy outcomes between preparation strategies, recent studies suggest that artificial preparation may be associated with higher pregnancy loss rates and, more recently, with hypertensive disorders of pregnancy. However, these findings were not confirmed in a recent RCT using genetically untested embryos. Since up to 70% of EPLs are attributed to genetic anomalies in the embryo, analyzing FET of euploid embryos provides a more accurate evaluation of the impact of endometrial preparation on pregnancy loss. STUDY DESIGN, SIZE, DURATION: An RCT was conducted in a university hospital (May 2019-May 2024) investigating the impact of endometrial preparation on pregnancy loss in FET. A power calculation based on internal data determined that 261 patients per arm were needed to detect an increase in pregnancy loss from 9.9% in tNC-FET to 18.4% in AC-FET (80% power, 5% significance). Due to safety concerns regarding hypertensive disorders following AC-FET, the trial was discontinued after enrolling 354 patients. PARTICIPANTS/MATERIALS, SETTING, METHODS: Women aged 18-42 years undergoing FET after preimplantation genetic testing for aneuploidy (PGT-A) with or without testing for a monogenic disorder (PGT-M) or a structural chromosomal rearrangement (PGT-SR) were randomized (1:1) to either true NC-FET (spontaneous ovulation without luteal phase support (LPS)) or AC-FET (using estradiol and single-route LPS). Only the first FET per patient using a euploid blastocyst was included. The primary outcome was EPL (before 8 weeks of gestation). Secondary outcomes included clinical pregnancy rate (CPR), ongoing pregnancy, live birth rate (LBR), and obstetric outcomes. MAIN RESULTS AND THE ROLE OF CHANCE: Between May 2019 and May 2024, 354 patients were randomized (NC-FET: 178, AC-FET: 176). In the intention-to-treat analysis, EPL per started FET cycle occurred in 6.7% of patients in the NC-FET group and 14.2% in the AC-FET group (P = 0.022). Additionally, EPL per positive hCG was 9.0% in the NC-FET group and 21.6% in the AC-FET group (P = 0.006). CPRs were significantly higher in the NC-FET group (72.5% versus 52.3%, P < 0.001), as were LBRs (67.4% versus 47.2%, P < 0.001). The study was prematurely terminated after obstetrical safety concerns of AC-FET emerged from other studies. While obstetric and neonatal outcomes did not differ significantly between groups in our study, the trial was not powered for these endpoints. Despite the premature halt, the observed difference in EPL was in line with the anticipated effect size. LIMITATIONS, REASONS FOR CAUTION: Although this randomized study achieved the estimated difference in the primary outcome, the premature cessation necessitates cautious interpretation. Additionally, the absence of luteal phase progesterone measurements in both groups represents a limitation. Nevertheless, these findings offer valuable insights for counseling patients on the optimal FET preparation method. WIDER IMPLICATIONS OF THE FINDINGS: The results of this RCT reinforce concerns regarding AC-FET cycles. In addition to previously reported higher obstetrical complications, an increased risk of pregnancy loss was associated with AC-FET cycles. This protocol should therefore preferentially be used in patients with (premature) ovarian insufficiency or those resistant to ovulation induction agents. STUDY FUNDING/COMPETING INTEREST(S): No external funding was received for this study. C.R. declares speaker's fees from Organon, Ferring, Abbott, IBSA; travel support from Gedeon Richter, Ferring, Organon, Abbott, IBSA. S.S.-R. declares employment and stock ownership in IVIRMA Lisboa; grants/contracts (paid to institution) from Theramex, Organon, Gedeon-Richter, Roche Pharmaceuticals; consulting fees from Organon, IBSA, Besins; travel support (paid to institution) and speaker's fees from Organon, Ferring, Theramex, IBSA, Gedeon-Richter, Abbott, and Besins; being a member of the Executive Committee, Past-Senior Deputy of Safety and Quality SIG for ESHRE. S.M. declares grants/contracts (to institution) from Merck, Gedeon Richter, Ferring; advisory board for Oxolife; speakers' fees and travel support from Merck, Gedeon Richter, Ferring, and IBSA. W.V. declares research grants (to institution) from Merck B.V. and IBSA; consulting fees (to institution) from CooperSurgical and Cook Medical; advisory board (payment to institution) for Goodlife; speaker fees (to institution) from CooperSurgical, Merck B.V., Organon Pharma B.V., Ferring B.V., and Gedeon Richter. C.B. declares speaker's fees from Abbott, Ferring Pharmaceuticals, IBSA, Organon, Gedeon Richter, and Merck. The other authors declare no conflict of interest related to this manuscript. TRIAL REGISTRATION NUMBER: Clinicaltrials.gov: NCT03976544. TRIAL REGISTRATION DATE: 6 April 2019. DATE OF FIRST PATIENT’S ENROLMENT: 25 May 2019.