Transcriptomic profiling of exRNA from spent human embryonic-cleavage culture medium reveals potential in embryo quality assessment：a retrospective study.

The exRNA transcriptome of the preimplantation embryo culture medium constitutes a comprehensive profile of embryonic RNAs, offering a non-invasive resource for elucidating developmental status. To delineate its relationship with the maternal-to-zygotic (MZT) transition and evaluate its predictive potential for developmental outcomes, we performed systematic exRNA transcriptome profiling of individually cultured cleavage-stage embryos. This study included 34 spent culture medium (SECM) samples for day 3 in vitro preimplantation cleavage embryos from 30 patients undergoing IVF/ICSI. 24 morphologically high-grade samples (Grade I, n = 12; Grade II, n = 12) and 10 low grade samples (Grade III, n = 6; abnormal PNs, n = 2; arrested embryos, n = 2) were included. Embryo transfer after the SECM collection was traced for subsequent clinical pregnancy outcome. ExRNA transcriptome from cleavage SECM showed 81.78% concordance with embryonic gene expression profiles. 1,058 differentially expressed exRNA markers were identified across embryo morphological grading groups, revealing MZT-related molecular dynamics prior to blastocyst formation. A 300-gene signature significantly correlated with embryonic developmental potential was established. The accuracy and robustness of the models were validated using clinical samples. The LogitBoost model built on exRNA profile yielded an AUC value above 0.95, indicating encouraging performance for embryo assessment. This study establishes SECM exRNA profiling as a non-invasive method for capturing key molecular events during early embryo development, particularly MZT activation. The identified exRNA biomarkers and machine learning models provide a promising framework for the objective assessment of embryo developmental potential, warranting further validation to overcome the limitations of conventional morphology-based selection in IVF.