L26/O-022 Oocyte quality modulates the impact of paternal age on ICSI outcomes: novel insights from an AI-based oocyte scoring system
Human Reproduction · 01.07.2026 · 41 (Supplement_1)
DOI: 10.1093/humrep/deag083.022 ↗Kurz gefasst
Fachartikel in der Zeitschrift Human Reproduction, online erschienen am 08.07.2026. Verfasst von E. Borges Junior und 8 weiteren Autor:innen.
Abstract (Originalsprache)
Study question Does oocyte quality, as assessed by an AI-derived quality score, modify the effect of paternal age (PA) on ICSI outcomes? Summary answer Oocyte quality, as represented by AI-derived quality scores, significantly modulates the effects of PA on ICSI outcomes, improving prediction of fertilization, blastulation, and pregnancy potential. What is known already While maternal age remains the primary determinant of success in assisted reproductive technology (ART), the independent effect of PA is not yet clearly defined. The oocyte’s inherent repair machinery is predominantly responsible for detecting sperm-derived lesions following chromatin remodeling; if fertilization occurs with a damaged spermatozoon. This repair capacity is modulated by several factors. AI-based tools for oocyte quality assessment offer a more objective and biologically informative evaluation of oocyte competence. However, the question remains whether AI-based oocyte evaluation can identify conditions under which the oocyte’s repair mechanisms better compensate for errors introduced by ageing sperm. Study design, size, duration This retrospective cohort study was conducted between January 2020-December 2024. Images of 13,613 denuded, mature oocytes from 3,696 ICSI cycles were assessed using an AI tool (0-10 score range). Oocytes were first stratified by PA into two groups: ≤39 (7,391) and >39 (6,221) years. Within each paternal-age group, oocytes were further stratified by the AI Score (AIS) using a pre-specified median cut-point of 6.4 per-oocyte: High AIS (>6.4, HighAIS) and Low AIS (≤6.4, LowAIS). Participants/materials, setting, methods Cycles were compared with respect to laboratory and clinical outcomes. Laboratory outcomes (fertilization, blastulation, high-quality blastocysts, and euploidy status) were analyzed at the oocyte level, whereas clinical outcomes (implantation, cumulative clinical pregnancy, and miscarriage rates) were analyzed at the cycle level. Multivariable models were performed including PA, AIS, maternal age, and female body mass index as predictive variables. Interactions between PA × AIS and PA × maternal age were also included in the models. Main results and the role of chance AIS was the strongest predictor of laboratory outcomes and, when associated with low PA, yielded the best results for fertilization (91.9%, 91.1%, 83.4%, and 84.1%; p < 0.001), blastulation (63.4%, 57.1%, 42.1%, and 38.8%; p < 0.001), high-quality blastocyst (50.1%, 45.8%, 40.7%, and 39.1%; p < 0.001), and euploidy rates (42.5%, 33.7%, 40.5%, and 27.1%; p < 0.001) for LowPA+HighAIS, HighPA+HighAIS, LowPA+LowAIS, and HighPA+LowAIS groups, respectively. Higher AIS values were associated with improved clinical outcomes among younger fathers, while also modulating the effect of high PA. Cumulative clinical pregnancy rates (66.6%, 55.7%, 55.6%, and 52.8%; p = 0.002) and implantation rates (52.6 ± 43.1%, 43.9 ± 44.5%, 44.0 ± 44.1%, and 43.2 ± 44.8%; p = 0.011) for LowPA+HighAIS, HighPA+HighAIS, LowPA+LowAIS, and HighPA+LowAIS, respectively, differed significantly across groups. The highest rates were observed when PA was ≤39 years and AIS high, followed by the intermediate groups, whereas cycles with older fathers and low AIS showed the lowest rates. Miscarriage rates were comparable across groups. Significant interactions between PA × AIS (OR 0.86; 95% CI 0.74–0.99; p = 0.031) and PA × maternal age (OR 1.16; 95% CI 0.99–1.36; p = 0.050) were observed for blastulation. Additionally, a significant PA × maternal age interaction was found for euploidy (OR 0.94; 95% CI 0.92–0.96; p = 0.023). Limitations, reasons for caution As a retrospective single-center study, residual confounding cannot be excluded. The oocyte-AI model, as applied with respect to PA, requires additional external validation. Mechanistic biomarkers of oocyte DNA repair were not evaluated, limiting biological interpretation. Studies integrating AI-based oocyte assessment with molecular gamete markers are needed to elucidate underlying mechanisms. Wider implications of the findings AI-based oocyte assessment provides novel insights into gamete interplay, demonstrating that oocyte quality may mitigate adverse PA-related effects on embryo development. These findings support the integration of AI-derived oocyte metrics to refine prognosis, enhance risk stratification, and guide individualized counseling and clinical decision-making in ICSI cycles. Trial registration number No
Zitieren
APA
Junior, E. B., Braga, D., Setti, A., Collado, M. D., Provenza, R., Fjeldstad, J., Mojiri, P., Vanzella, L., & Junior, A. I. (2026). L26/O-022 Oocyte quality modulates the impact of paternal age on ICSI outcomes: novel insights from an AI-based oocyte scoring system. Human Reproduction, 41(Supplement_1). https://doi.org/10.1093/humrep/deag083.022
Vancouver
Junior EB, Braga D, Setti A, Collado MD, Provenza R, Fjeldstad J, et al. L26/O-022 Oocyte quality modulates the impact of paternal age on ICSI outcomes: novel insights from an AI-based oocyte scoring system. Human Reproduction. 2026;41(Supplement_1). doi:10.1093/humrep/deag083.022
BibTeX
@article{Junior2026,
author = {Junior, E. Borges and Braga, D. and Setti, A. and Collado, M. Del and Provenza, R. and Fjeldstad, J. and Mojiri, P. and Vanzella, L. and Junior, A. Iaconelli},
title = {L26/O-022 Oocyte quality modulates the impact of paternal age on ICSI outcomes: novel insights from an AI-based oocyte scoring system},
journal = {Human Reproduction},
year = {2026},
volume = {41},
number = {Supplement_1},
doi = {10.1093/humrep/deag083.022}
}
EUVIMED