Does the freeze-all strategy improve the cumulative live birth rate and the time to become pregnant in IVF cycles?

Purpose The freeze-all strategy is widely used for ovarian hyperstimulation syndrome (OHSS) prevention. Indeed, it increases live birth rates among high responders and prevents preterm birth and small for gestational age. Why should not we extend it to all? Methods A retrospective and monocentric study was conducted between January 2008 and January 2018 comparing the cumulative live birth rates (CLBR) between patients having undergone FAS and a control group using fresh embryo transfer (FET) and having at least one frozen embryo available. Analyses were made for the entire cohort (population 1) and for different subgroups according to confounding factors selected by a logistic regression (population 3), and to the BELRAP (Belgian Register for Assisted Procreation) criteria (population 2). Results 2216 patients were divided into two groups: Freeze all (FA), 233 patients and control (C), 1983 patients. The CLBR was 50.2% vs 58.1% P = 0.021 for population 1 and 53.2% vs 63.3% P = 0.023 for population 2, including 124 cases and 1241 controls. The CLBR stayed in favour of the C group: 70.1% vs 55.9% P = 0.03 even when confounding variables were excluded (FA and C group, respectively, 109 and 770 patients). The median time to become pregnant was equally in favour of the C group with a median of 5 days against 61 days. Conclusion CLBR is significantly lower in the FA group compared to the C group with a longer time to become pregnant. Nevertheless, the CLBR in the FA group remains superior to that observed in previous studies.