Use of the CryoPredict algorithm to predict live birth from cryopreserved embryos.

Background Currently, the viability of cryostored blastocysts that are subsequently re-warmed is determined via the percentage of cell survival. However, the large number of cells that forms the blastocyst can make this estimate difficult and unreliable. Studies have shown that fast re-expanding blastocysts have superior pregnancy rates. Aim To determine whether the degree and speed of blastocoele re-expansion following cryopreservation and warming correlate with rates of live birth. Materials and methods A retrospective cohort study of 757 frozen embryo transfer cycles over a 4-year period at Royal Prince Alfred Hospital, Sydney. Clinical and embryology notes were retrieved. Details regarding patient demographics, stimulation cycle from which embryos were derived, frozen embryo transfer cycles, embryology and pregnancy outcomes were recorded. Results Female (P = 0.01) and male age (P = 0.02) at the time of embryo creation were inversely associated with live birth. Fertilisation method (P = 0.03), embryo type at cryopreservation (P = 0.009), embryo grade at cryopreservation (P < 0.0001), percentage of cell survival post-thaw (P < 0.0001) and the degree of re-expansion (P = 0.003) were the IVF and embryology factors significantly associated with live birth. A predictive model (CryoPredict) was created in order to individualise the probability that the transfer of a given embryo would result in live birth. Conclusions The degree and speed of blastocoele re-expansion postcryopreservation and subsequent warming can be used in conjunction with other parameters to predict live birth.