論文

Objective: To generate an effective embryo prediction model and identify a non-invasive evaluation method by analyzing microRNAs (miRNAs) in embryo culture medium. Design: Analysis of microRNA profiles from spent culture medium of blastocysts with good morphology that did or did not result in pregnancy. Setting: Clinical and experimental research. Patients: Sixty patients who underwent thawed embryo transfer of blastocysts after intracytoplasmic sperm injection. Intervention(s): None. Main outcome measure(s): The association of miRNA abundance levels secreted by blastocysts in culture medium and implantation success. Results: Our RNA sequencing analysis found a total of 53 differentially expressed miRNAs in the culture media of pregnancy and non-pregnancy groups. Twenty-one miRNAs were analyzed for their potential to predict implantation success. Eight miRNAs (hsa-miR-191-5p, hsa-miR-320a, hsa-miR-92a-3p, hsa-miR-509-3p, hsa-miR-378a-3p, hsa-miR-28-3p, hsa-miR-512-5p, and hsa-miR-181a-5p) were further extracted from the results of a logistic regression analysis of qPCR Ct values. A prediction model for high-quality blastocysts was generated using the eight miRNAs, with an average accuracy of 0.82 by 5-fold cross validation. Conclusion: We isolated blastocyst miRNAs that may predict implantation success and created a model to predict viable embryos. Increasing the number of investigated cases and further studying the effect of each miRNA on embryonic development is needed to refine the miRNA-based predictive model.