Why we should transfer frozen instead of fresh embryos: the translational rationale

Abstract

Epidemiologic studies have shown an increased rate of adverse perinatal outcomes, including small for gestational age (SGA) births, in fresh in vitro fertilization (IVF) cycles compared with frozen embryo transfer cycles. This increase is not seen in the donor oocyte population, suggesting that it is the peri-implantation environment created after superovulation that is responsible for these changes. During a fresh IVF cycle, multiple corpora lutea secrete high levels of hormones and other factors that can affect the endometrium and the implanting embryo. In this review, we discuss both animal and human data demonstrating that superovulation has significant effects on the endometrium and embryo. Additionally, potential mechanisms for the adverse effects of gonadotropin stimulation on implantation and placental development are proposed. We think that these data, along with the growing body of epidemiologic evidence, support the proposal that frozen embryo transfer should be considered preferentially, particularly in high responders, as a means to potentially decrease at least some of the adverse perinatal outcomes associated with IVF.

Keywords: In vitro fertilization; frozen embryo transfer; implantation; placentation; superovulation.

FIGURE 1

Effect of the peri-implantation environment on mouse placentation and fetal growth. (A) Fetal and placental weights of embryos at embryonic day 19 resulting from the transfer of naturally conceived blastocysts to control and superovulated recipient mice. Fetal and placental weights were significantly smaller in the superovulated group. Data are expressed as mean ± SEM. (B) Placental histology differed significantly in the offspring of control compared with superovulated recipients. Placentas from superovulated recipients showed attenuated branching with limited invasion of the junctional zone by the labyrinth zone with a decrease in the junctional:labyrinth zone ratio. ^aP<.05 vs. natural mating. ^bP<.001 vs. natural mating. Adapted from Mainigi et al. (63).