Abstract
Background: The availability and use of frozen-thawed embryos after controlled ovarian hyperstimulation for assisted reproduction have increased with improvements in vitrification techniques and the rise of gonadotropin-releasing hormone (GnRH) antagonist protocols. Although evidence has shown that frozen-thawed embryo transfers (FETs) result in higher live birth rates than fresh embryo transfers, it is uncertain whether this association exists in cycles employing the GnRH antagonist protocol. Objective: To test the hypothesis that FETs are more likely to result in a live birth than fresh embryo transfers in a GnRH antagonist protocol cycle and to investigate whether frozen blastocyst transfer increases live birth rates compared to fresh blastocyst transfer. Design: A retrospective historical cohort study was conducted using data collected from the Department of Reproductive Medicine of Liuzhou Maternity and Child Healthcare Hospital for 1,437 patients who underwent the GnRH antagonist protocol between 1 January 2015, and 31 December 2020. The primary outcome was the live birth rate, which was compared between fresh embryo transfer and FET, and the secondary outcomes were clinical pregnancy rate and miscarriage rate, which were compared between the two groups. Analyses were adjusted to account for the age of the patient, number of embryo transfers, day of embryo transfer, and type of infertility. Results: Fresh embryo transfers accounted for 1,026 (71.4%) of the 1,437 patients who underwent the GnRH antagonist protocol in our analysis, while FETs accounted for 411 (28.6%). Patients with fresh and frozen-thawed embryos had comparable median body mass index (body mass index; 22.3 [IQR, 24.6-20.0] vs. 22.0 [IQR, 24.5-19.9]). There was a significant difference in the median age of the fresh embryo transfer group (34.0 [IQR, 39.0-30.0]) and the Frozen-thawed embryo transfer group (32.0 [IQR, 37.0-29.0]). Blastocysts were transferred in 14.6% of the fresh embryo transfer cycles and 45.5% of the FET cycles, whereas they account for 10.4% and 13.0% of all patients, respectively. The mean number of embryos transferred was 2 (IQR, 2.0-1.0) for the fresh embryo transfer group and 1 (IQR, 2.0-1.0) for the FET group, with a significant difference in the mean number of embryos transferred. The live birth rate after fresh embryo transfer vs. FET was 28.7% vs. 34.5% (absolute difference, 5.9%; adjusted relative risk [aRR], 1.15 [95% CI, 0.88-1.51]). The clinical pregnancy rates were 39.9% vs. 46.0%, respectively (absolute difference, 6.1%; aRR, 1.10 [95% CI, 0.85-1.43]). The miscarriage rates were 22.5% vs. 23.8%, respectively (absolute difference, 1.3%; aRR, 1.13 [95% CI, 0.75-1.70]). Conclusion: In this retrospective study of women who underwent assisted reproduction using GnRH antagonists, FETs resulted in a higher live birth rates and clinical pregnancy rates than fresh embryo transfers, which parts of these differences were attributable to embryo stage. However, the interpretation of the findings is limited by the possibility of selection and confounding biases.