A multi-centre evaluation of a novel 4-cell embryo classification system based on intercellular contact points

Abstract Study question Is there any prognostic value to classifying 4-cell embryos according to intercellular contact points (ICPs)? Summary answer A significant association was found between the new blastomere arrangement classifications and blastulation, as well as blastocyst quality. No significant association was found for pregnancy. What is known already Current literature states that the geometric arrangement of blastomeres in 4-cell embryos is usually classified as either “tetrahedral” or “non-tetrahedral/planar”. Though tetrahedral embryos have been associated with greater developmental capacity, their prevalence has varied between studies. At ESHRE 2021, Hickman et al. proposed a more fine-grained classification system; each class was associated with a 4-digit code, where the Nth digit in said code gave the number of cells with N-1 ICPs. This gave rise to 6 classes: tetrahedral (0004), pseudotetrahedral (0022), planar (0040), closed-Y (0121), open-Y (0301) and linear (0220). In our study, we evaluate this new system. Study design, size, duration This study was a retrospective analysis of 844 4-cell embryos across 3 clinics in 3 countries. Focal stacks of the embryos were captured on Embryoscope/MIRI time-lapse incubators between 2018 and 2020. Embryos were annotated by their respective clinics using the original “tetrahedral/planar” system. Additionally, data on blastulation, blastocyst grade (Gardner scale) and biochemical pregnancy rates were obtained for each embryo. Participants/materials, setting, methods The embryos were annotated according to the new system by a panel of three researchers without knowledge of the clinics’ or each other’s classifications. Afterwards, the panelists assembled for discussion and a unanimous consensus was reached for each embryo. The cell arrangement classes were analysed with respect to blastulation rates, blastocyst grade and pregnancy rates using chi-squared tests. The distribution of classes across the clinics was also analysed, as well as agreement among the panelists. Main results and the role of chance Under the new system, tetrahedral and pseudotetrahedral embryos saw significantly higher rates of blastulation (P = 0.017) and good-quality blastocysts (p = 0.043; here, “good-quality” means that the grade contains no ‘C’s or numbers less than 3) compared to other arrangements. No association was found with respect to biochemical pregnancy (p = 0.77). There was significant variation between the clinic-provided classifications (p < 0.001) with tetrahedral-to-planar ratios ranging from 1 to 9. Under the new system, there was no significant difference between the proportions of each arrangement (p = 0.66). The arrangements in order of decreasing prevalence across all the data were tetrahedral (63.7%), pseudotetrahedral (25.8%), planar (8.0%), closed-Y (2.2%) and linear (0.3%). No open-Y embryos were observed. Upon comparison with the clinic-provided tetrahedral/planar classifications, there were clear differences in the treatment of pseudotetrahedral embryos - some clinics predominantly classified them as “tetrahedral” while others as “planar”. Prior to any discussions, the panelists’ annotations unanimously agreed on the classifications of 63% of the embryos; with 33% of the embryos having 2 different classifications and 4% of embryos causing total disagreement. The majority of disagreements regarded pseudotetrahedral embryos being confused for either tetrahedral or planar embryos. Limitations, reasons for caution It can be at times difficult to visualise the 3D structure of embryos from focal stacks. As a result, some annotations may be erroneous, though the use of a panel-based approach helped to mitigate this. Moreover, further studies will need to take place to validate the findings in this work. Wider implications of the findings The findings demonstrate the prognostic utility of Hickman et al.’s ICP-based classification system. Moreover, the findings suggest that much of the variability seen in the prevalence of tetrahedral embryos under the tetrahedral/planar system came from the mislabelling of pseudotetrahedral embryos which comprised 25.8% of the study population. Trial registration number N/A