Non-invasive embryo evaluation and selection using time-lapse monitoring: Results of a randomized controlled study

Objective: Continuous monitoring of embryos via time-lapse (TL) provides more information on embryo kinetics and morphology compared to standard daily evaluation. Embryo selection by TL could support single embryo transfer (SET). With SET multiple gestations are avoided and perinatal outcome is improved. Our primary goal was to determine whether selection of a single blastocyst based on an algorithm comprising kinetic and morphologic scores assessed through continuous TL monitoring results in superior clinical outcome compared to embryo selection based on morphology alone. A secondary goal was to assess whether a time-lapse score based on kinetic and morphologic parameters was predictive of implantation.

Study design: Randomized controlled trial performed in two private IVF centers in Hungary. Infertile couples scheduled to undergo 1st or 2nd IVF cycles were enrolled. Female age had to be under 36 years. The intervention was embryo evaluation/selection based on TL algorithm. Patients were randomized to SET with TL monitoring (TL-eSET) vs. SET with standard evaluation (control-eSET). Assuming an increase in pregnancy from 44% to 58%, a sample size of 202 per group was calculated based on the interim analysis at 10% information fraction. The primary outcome of the study was pregnancy rate. Secondary outcomes were miscarriage rates, live birth, perinatal outcome and the ability of a time-lapse score constructed based on kinetic and morphologic parameters to predict implantation. Chi-square tests, likelihood-ratio tests and exact tests were used for the analysis of categorical variables. Continuous variables were compared using independent group t-test and analysis of variance.

Results: The study was closed after three years. Eventually 161 patients were randomized and analyzed (N = 80 TL-eSET and N = 81 control-eSET). Pregnancy rate did not significantly differ between the groups though there was a trend favoring TL selection (TL-eSET: 46.3% vs control-eSET: 34.6%, p = 0.150; OR: 1.628 (95% CI: 0.857-3.092)). The time-lapse score based on morphologic and kinetic parameters was significantly higher for blastocysts that implanted vs. those that did not (14.5 ± 1.8 vs. 12.1 ± 2.9, p = 0.0001). There were no adverse effects of the intervention.

Conclusions: Selection of a single blastocyst based on information derived from time-lapse monitoring can aid embryo selection for SET.