L26/O-073 Blastocoel fluid RNA predicts pregnancy outcome in assisted reproduction
Human Reproduction · 01.07.2026 · 41 (Supplement_1)
DOI: 10.1093/humrep/deag083.073 ↗Kurz gefasst
Fachartikel in der Zeitschrift Human Reproduction, online erschienen am 08.07.2026. Verfasst von C. Frisendahl und 14 weiteren Autor:innen.
Abstract (Originalsprache)
Study question Can embryo excreted RNA be used to predict embryo quality in assisted reproductive technologies (ART)? Summary answer RNA in blastocoel fluid (BF) can predict pregnancy outcome of individual embryos in ART. What is known already There is an unmet need of objective methods with improved precision to assess embryo quality in ART to improve management of infertility. RNA molecules excreted by the embryo can be detected in embryo spent media (ESM) and BF, but it is unknown whether these molecules can be used to predict pregnancy outcome in ART. Study design, size, duration This multicenter prospective cohort study aimed to develop a machine-learning model to predict pregnancy outcomes based on RNA in ESM and BF. The primary outcome was the result of a urine hCG pregnancy test. Secondary outcomes included live-birth and negative pregnancy outcomes (miscarriage and biochemical pregnancy). A total of 235 ESM and 76 BF samples were collected from couples undergoing ART with single embryo transfers between 2018 and 2019 in four IVF clinics in Sweden. Participants/materials, setting, methods ESM and BF were collected from individually cultured cleavage- and blastocyst-stage embryos. Samples were sequenced to profile mRNA and small RNAs, followed by differential expression analysis using DESeq2. Genes with p < 0.05 and fold change ≥ 2 or ≤ −2 were considered significant. Machine-learning was applied to develop a prediction model for pregnancy outcome using mRNA in BF. A leave-one-out-cross-validation strategy was used for training and testing, followed by validation in an independent cohort. Main results and the role of chance To detect mRNA and small RNA molecules, we developed an innovative library preparation protocol for sequencing microlitre-scale fluids with low RNA content, enabling analysis of ESM and BF from individual preimplantation embryos. Few differentially abundant miRNAs and tRNAs were identified in ESM and BF when comparing hCG-positive and hCG-negative embryos. Larger differences were observed when comparing poor-quality (non-transferable) embryos with hCG-positive embryos, with 53 miRNAs and 60 tRNAs being differentially abundant. For mRNAs, a clear difference in mRNA content between hCG-positive and hCG-negative embryos was detected in BF but not in ESM. In total, 137 differentially abundant mRNAs were identified in BF. Similar patterns were observed when comparing live births with negative pregnancy outcomes in BF. Subgroup analysis by maternal age showed higher RNA levels in hCG-negative samples from women older than 34 years, a trend not observed in younger women. Based on these results, we developed a machine learning model to predict pregnancy outcome using mRNA profiles from BF as input features. The optimal model, based on L1-regularized logistic regression, achieved an AUC of 0.86, with 81% sensitivity, 71% specificity, and 76% accuracy. Validation in an independent cohort achieved an AUC of 0.73. Limitations, reasons for caution This is a proof-of-concept study including a limited sample size to train, test and validate the prediction model. Although a leave-one-out-cross-validation strategy has been used to avoid overfitting of the ML model, larger clinical trials are needed to validate the robustness of the method and compare it against standard care. Wider implications of the findings We show for the first time that both mRNA and smallRNA molecules can be extensively mapped in individual samples of both ESM and BF, using highly sensitive library preparation protocols. Our novel prediction model has the potential to improve embryo selection and thus increase the success rate of ART treatment. Trial registration number No
Zitieren
APA
Frisendahl, C., Gemzell-Danielsson, K., Nair, K., Koch, F., Menezes, J., Sheikhi, M., Dahlgren, K., Alnahal, W., Bohlin, T., Olofsson, J., Töhönen, V., Vafaee, F., Grace, L. P., Boggavarapu, N. R., & Faridani, O. (2026). L26/O-073 Blastocoel fluid RNA predicts pregnancy outcome in assisted reproduction. Human Reproduction, 41(Supplement_1). https://doi.org/10.1093/humrep/deag083.073
Vancouver
Frisendahl C, Gemzell-Danielsson K, Nair K, Koch F, Menezes J, Sheikhi M, et al. L26/O-073 Blastocoel fluid RNA predicts pregnancy outcome in assisted reproduction. Human Reproduction. 2026;41(Supplement_1). doi:10.1093/humrep/deag083.073
BibTeX
@article{Frisendahl2026,
author = {Frisendahl, C. and Gemzell-Danielsson, K. and Nair, K. and Koch, F. and Menezes, J. and Sheikhi, M. and Dahlgren, K. and Alnahal, W. and Bohlin, T. and Olofsson, J.I. and Töhönen, V. and Vafaee, F. and Grace, L. Parameswaran and Boggavarapu, N. Rao and Faridani, O.R.},
title = {L26/O-073 Blastocoel fluid RNA predicts pregnancy outcome in assisted reproduction},
journal = {Human Reproduction},
year = {2026},
volume = {41},
number = {Supplement_1},
doi = {10.1093/humrep/deag083.073}
}
EUVIMED