JBRA Assisted Reproduction 2025;29(Suppl.2 SBRA 2025):220
Poster Presentation
29th Annual Congress of the SBRA. São Paulo/SP - Brazil, 2025
doi: 10.5935/1518-0557.20263862
P-208. Should donors with Polycystic Ovaries be considered? Insights from AI-assessed oocyte quality and recipient outcomes
Simone Mattiello1, Renata Ternus1, Gabriella Mamede Andrade1, Laís Vanzella1, Natalie Mercuri2, Parisa Mojiri2, Jullin Fjeldstad2, Nilo Frantz1, Marcos Iuri Roos Kulmann1
1 Nilo Frantz Medicina Reprodutiva – Porto Alegre - RS - Brasil
2 Future Fertility -Canada
Objective: The impact of polycystic ovary (PCO) morphology and polycystic ovary syndrome (PCOS) on oocyte quality remains debated. In oocyte donation cycles, particularly in countries where financial compensation is prohibited, donors are often IVF patients themselves, making careful donor selection crucial. While some ovarian pathologies, such as endometriosis, are generally excluded from donation, the suitability of donors with PCO or PCOS is still unclear. Recipient cycles offer a unique opportunity to assess the intrinsic quality of these oocytes, isolating donor-related effects from endometrial influences. This study aims to investigate the developmental potential of oocytes from donors with PCO or PCOS using AI-based image analysis and recipient outcomes.
Methods: This retrospective cohort study analyzed 551 donor oocytes from 38 stimulation cycles conducted between September 2024 and June 2025. Oocyte donors from 13 cycles (247 oocytes) exhibited polycystic ovaries (PCO) on ultrasound examination. Among these, three donors met the Rotterdam criteria for clinical polycystic ovary syndrome (PCOS), presenting oligo-anovulation and/or clinical/biochemical hyperandrogenism in addition to PCO morphology. The remaining 25 cycles (304 oocytes) served as controls, with donors exhibiting normal ovarian parameters. High-resolution, two-dimensional images of fresh metaphase II (MII) oocytes were captured at 200× magnification, either immediately prior to vitrification or pre-ICSI when used fresh in recipient cycles. Oocyte quality was evaluated using MAGENTATM, an AI- based image analysis tool trained to predict blastocyst development. The tool assigns a score from 0 to 10 to each oocyte, with higher scores indicating a greater likelihood of blastocyst development. To correlate AI findings with developmental potential, embryological outcomes were assessed in a subset of 135 inseminated oocytes across 19 recipient cycles using normozoospermic semen samples. Embryos were cultured to the blastocyst stage (up to Day 6) and either transferred or vitrified according to clinical protocols.
Results: Donor age was comparable between PCO and control groups (29.4±1.7 vs. 30.3±3.0 years, p=0.2883). As expected, PCO donors displayed a distinct ovarian profile, with significantly higher AMH levels (9.8±7.1 vs. 2.9±1.2 ng/mL, p<0.0001), more aspirated follicles (37.7±19.5 vs. 21.3±8.1, p<0.001), and MII oocytes (20.0±5.9 vs. 12.5±2.5, p<0.001). AI-driven oocyte quality assessment (MAGENTATM Score, MS) revealed no significant difference between PCO (7.1±2.4, n=247) and control oocytes (6.9±2.4, n=304; p=0.3265). Strikingly, oocytes from clinically confirmed PCOS donors (n=52) were also not inferior to controls, achieving even higher MS (8.2±1.7, p<0.001). In the developmental potential analysis (n=135 oocytes from 19 recipient cycles), no significant differences were observed between recipients of PCO and control donor oocytes in terms of donor age, inseminated oocytes, 2PN rates, and overall blastocyst formation (all p>0.05). However, the top-quality blastocyst rate was significantly lower with PCO donor oocytes (33.3% vs. 57.9%, p<0.05). Notably, a greater proportion of frozen oocytes was used in PCO recipient cycles compared to control cycles (88.8% vs. 50%, p<0.0001). When the analysis was restricted to frozen-only cycles (n=13), the difference in top-quality blastocyst rates disappeared (28.2% vs. 45.8%, p=0.1814), indicating that vitrification was the primary determinant of blastocyst quality.
Conclusion: Our findings indicate that oocytes from donors with PCO or PCOS do not exhibit compromised quality. Moreover, the developmental potential of PCO donor oocytes in recipient cycles was comparable to that of control donors, supporting the inclusion of PCO donors in oocyte donation programs as a viable option. However, these results should be validated in a larger sample size including pregnancy outcomes.