EUVIMEDEUVIMED entdecken
Crossref

L26/P-625 Multimodal non-invasive blastocyst aneuploidy risk assessment integrating AI-derived embryo scores and spent culture medium cell-free DNA (cfDNA)

C. Giménez-Rodríguez, S. Marqueño, L. Bori, M.C. Savash Ishanzadeh, D. Wells, J. Bellver, M. Meseguer

Human Reproduction · 01.07.2026 · 41 (Supplement_1)

DOI: 10.1093/humrep/deag083.955 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Human Reproduction, online erschienen am 08.07.2026. Verfasst von C. Giménez-Rodríguez und 6 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Study question Does quantification of cfDNA from spent culture medium provide independent and clinically relevant information when combined with AI-based embryo assessment for embryo aneuploidy risk? Summary answer cfDNA provides complementary, independent information to AI-derived embryo scores, modestly improving discrimination and refining embryo prioritisation by enriching euploid embryos among top-ranked selections. What is known already Artificial intelligence (AI) applied to embryo imaging enables non-invasive embryo assessment and has demonstrated associations with embryo ploidy beyond conventional morphology. In parallel, cell-free DNA (cfDNA) detected in spent culture medium has been proposed as a non-invasive biomarker of embryo genetic status, although its standalone discriminatory performance and clinical utility remain debated. Whether cfDNA adds meaningful, independent information to AI-based embryo assessment beyond modest changes in global discrimination remains unclear. Study design, size, duration A total of 382 blastocysts from IVF cycles were included between January 2023 and January 2025. Embryo euploidy prediction was evaluated using AI-derived embryo image scores and cfDNA quantified from spent culture medium, with PGT-A as the reference standard. Participants/materials, setting, methods Embryos were classified as euploid or aneuploid by day 5/6 trophectoderm biopsy analysed by NGS and morphologically graded according to ASEBIR criteria. Embryo images were analysed using AI algorithms for genetic status (Life Whisperer Genetics) and developmental competence (Life Whisperer Viability). cfDNA was quantified by absolute qPCR from pooled spent culture medium samples. Multivariable logistic regression and AUC analyses were performed. Clinical relevance was further explored using ranking-based analyses, assessing euploid risk among top-prioritised embryos. Main results and the role of chance cfDNA quantity in spent culture medium was lower in euploid than aneuploid embryos (2.26 vs 3.88 pg; p < 0.05), indicating an underlying association between cfDNA levels and chromosomal status. In multivariable logistic regression, AI-derived scores remained associated with euploidy (genetic status algorithm OR 1.146, 95% CI 1.058–1.241; developmental competence algorithm OR 1.145, 95% CI 1.016–1.290). Notably, cfDNA also remained independently associated with ploidy status after adjustment for AI-derived scores and clinical covariates (OR 0.955, 95% CI 0.921–0.990). In stepwise models, conventional morphology was excluded—consistent with overlap with AI-based assessment—while cfDNA retained independent explanatory value as a complementary biological signal. Integration of cfDNA into the model resulted in only a modest increase in global discrimination (AUC 0.760 for AI alone vs 0.769 for AI plus cfDNA), with overlapping confidence intervals. However, the added value of cfDNA became evident when embryo prioritisation was evaluated. Ranking embryos by predicted euploidy probability increased euploid enrichment when cfDNA was integrated: in the top 30% of prioritised embryos, euploidy rate increased from 73.2% with AI alone to 78.6% with AI plus cfDNA, with similar directional enrichment at top 20% (80.0% to 82.7%) and top 40% (67.3% to 71.3%) thresholds. Limitations, reasons for caution This single-centre observational study is limited by its sample size. Larger, multicentre studies are required to corroborate these findings and assess generalisability. cfDNA was analysed quantitatively without molecular characterisation, and findings may not extrapolate to sequencing-based non-invasive PGT-A approaches. Wider implications of the findings Rather than acting as a standalone classifier, cfDNA provides complementary biological information that can refine AI-based embryo ranking by enriching euploid embryos among top-prioritised selections. This multimodal approach may support non-invasive embryo selection strategies, particularly where invasive PGT-A is not feasible, and warrants prospective evaluation in embryo transfer decision pathways. Trial registration number No

Zitieren

APA

Giménez-Rodríguez, C., Marqueño, S., Bori, L., Ishanzadeh, M. S., Wells, D., Bellver, J., & Meseguer, M. (2026). L26/P-625 Multimodal non-invasive blastocyst aneuploidy risk assessment integrating AI-derived embryo scores and spent culture medium cell-free DNA (cfDNA). Human Reproduction, 41(Supplement_1). https://doi.org/10.1093/humrep/deag083.955

Vancouver

Giménez-Rodríguez C, Marqueño S, Bori L, Ishanzadeh MS, Wells D, Bellver J, et al. L26/P-625 Multimodal non-invasive blastocyst aneuploidy risk assessment integrating AI-derived embryo scores and spent culture medium cell-free DNA (cfDNA). Human Reproduction. 2026;41(Supplement_1). doi:10.1093/humrep/deag083.955

BibTeX
@article{GimnezRodrguez2026,
  author = {Giménez-Rodríguez, C. and Marqueño, S. and Bori, L. and Ishanzadeh, M.C. Savash and Wells, D. and Bellver, J. and Meseguer, M.},
  title = {L26/P-625 Multimodal non-invasive blastocyst aneuploidy risk assessment integrating AI-derived embryo scores and spent culture medium cell-free DNA (cfDNA)},
  journal = {Human Reproduction},
  year = {2026},
  volume = {41},
  number = {Supplement_1},
  doi = {10.1093/humrep/deag083.955}
}
Kostenlos starten →