Melatonin Supplementation During In-Vitro Maturation Restores Oocyte Quality and Developmental Competence in Aged Mice

Authors: Jeong Yoon; H. Park; K. Juhn; E. Jung; J. Jin; Y. Kang; J. Lim

Advanced maternal age is linked to heightened oxidative stress in the follicular environment, with human studies showing increased lipid peroxidation and reduced antioxidant defenses in women ≥37 years. This excess reactive oxygen species (ROS) impairs oocyte quality, while aging oocytes exhibit mitochondrial dysfunction, DNA double‑strand breaks, and epigenetic dysregulation that compromise early embryonic development. Collectively, oxidative stress emerges as a major driver of age‑related oocyte decline, prompting investigation into whether melatonin supplementation during in vitro maturation (IVM) can restore oocyte quality under aging‑associated oxidative conditions.

This study investigated whether melatonin supplementation during in vitro maturation (IVM) could improve oocyte quality and developmental potential in aged mice by reducing oxidative stress and restoring cellular function.The study used germinal vesicle (GV)-stage oocytes collected from aged mice (70–90 weeks). Oocytes were matured in vitro with or without 5 μM melatonin and then subjected to parthenogenetic activation. Researchers evaluated mitochondrial function, reactive oxygen species (ROS) levels, DNA damage, global protein synthesis, DNA methylation, embryo development, blastocyst quality, and implantation potential through multiple independent biological replicates. Melatonin supplementation significantly restored mitochondrial membrane potential (P<0.05) and markedly reduced intracellular ROS levels (P<0.0001). This was accompanied by a significant reduction in γ-H2AX-positive (gamma H2A Histone Family Member X) DNA damage foci (P<0.01). Melatonin also restored global protein synthesis and global DNA methylation (5-methylcytosine [5mC]) levels (P<0.05). Following parthenogenetic activation, developmental competence improved with melatonin treatment. The 2-cell embryo formation rate increased significantly compared with aged controls (61.2% vs. 45.8%; P<0.05). Usable embryo yield, defined as the proportion of GV-stage oocytes that developed into blastocysts, increased by 56.9%, from 32.0% in aged controls to 50.2% with melatonin (P<0.05). Blastocysts derived from melatonin-treated oocytes had higher total cell numbers than controls (115.4 vs. 98.6; P<0.05) and an improved inner cell mass to trophectoderm (ICM:TE) ratio (0.35 vs. 0.31). In addition, in vitro outgrowth assays showed significantly larger inner cell mass and trophectoderm outgrowth areas (P<0.05), indicating improved implantation potential. Developmental competence was evaluated using parthenogenetic activation rather than in vitro fertilization (IVF). Further studies using IVF and embryo transfer are needed to confirm live birth outcomes and assess the effects of melatonin on paternal chromatin and syngamy.

Overall, melatonin supplementation during in vitro maturation improved mitochondrial function, reduced oxidative stress and DNA damage, restored protein synthesis and DNA methylation, and enhanced embryo development and blastocyst quality in oocytes from aged mice.

Clinical Outcomes of Single Vitrified-Warmed Blastocyst Transfer Using Granulocyte Macrophage Colony-Stimulating Factor (GM-CSF)–Containing Versus Hyaluronan-Enriched Transfer Media: A Retrospective Comparative Study

Authors: JUNSANG PARK; J.W. Lee; H.H. Jung; Y.W. Kim; S.D. Kim; T.H. Min; J.H. Yang; E.S. Lee; J.H. Ahn; J.Y. Chung; S.Y. Kim

Granulocyte–macrophage colony‑stimulating factor (GM‑CSF) is a cytokine secreted by epithelial cells of the female reproductive tract and placenta, playing a key role in embryo maternal communication since embryos express GM‑CSF receptors but cannot produce GM‑CSF themselves. Studies have reported improved reproductive outcomes with GM‑CSF–supplemented culture media compared to standard media, yet hyaluronan‑enriched transfer media remain widely used in practice, and direct comparative evidence between these two approaches is still limited.

This retrospective study compared the effectiveness of granulocyte-macrophage colony-stimulating factor (GM-CSF)-containing transfer media with hyaluronan-enriched transfer media in women undergoing single vitrified-warmed blastocyst transfer. The study included 378 transfer cycles in normal responder women aged 22–38 years between December 2023 and November 2025. All women underwent ovarian stimulation using a gonadotropin-releasing hormone (GnRH) antagonist protocol, and a single day 5 or day 6 vitrified-warmed blastocyst was transferred. Of the total cycles, 187 used GM-CSF-containing transfer media and 191 used hyaluronan-enriched transfer media. Donor oocyte cycles and double embryo transfers were excluded. Blastocysts were graded using the Gardner and Schoolcraft criteria and were equilibrated in the assigned transfer medium for approximately 97 minutes before transfer. Clinical pregnancy and miscarriage rates were compared between the two groups. Baseline demographic and treatment characteristics were similar between the groups. There were no significant differences in mean female age (33.3 ± 3.0 vs. 33.6 ± 2.7 years; p = 0.36), mean male age (35.5 ± 3.7 vs. 35.7 ± 4.0 years; p = 0.60), number of mature oocytes retrieved (9.9 ± 4.9 vs. 9.8 ± 5.5; p = 0.88), fertilization rate (90.1% vs. 91.2%; p = 0.60), post-warming survival rate (98.9% vs. 99.5%; p = 0.55), or the proportion of good-quality blastocysts transferred (48.7% vs. 53.4%; p = 0.36). Clinical outcomes were also comparable between the two groups. Biochemical pregnancy rates were numerically higher with GM-CSF-containing media than with hyaluronan-enriched media (72.7% vs. 64.9%; p = 0.10), as were ongoing pregnancy rates (61.5% vs. 53.9%; p = 0.14); however, neither difference was statistically significant. Miscarriage rates were nearly identical between the groups (11.2% vs. 11.0%; p = 0.94). As all transfers involved a single blastocyst, twin pregnancies were presumed to be monozygotic and occurred at similarly low rates in both groups (2.6% vs. 2.9%; p = 0.89). The study was limited by its retrospective, single-center design, lack of live birth data, inclusion of only normal responder women younger than 38 years, and the absence of subgroup analyses based on blastocyst quality.

Overall, GM-CSF-containing transfer media did not provide a significant clinical advantage over hyaluronan-enriched transfer media in single vitrified-warmed blastocyst transfer cycles.

Effect of Antioxidant-Enriched Embryo Culture Media on Embryo Development and ART Outcomes.

Authors: Clara Miret Lucio; M. Benavent Martínez; M. Escribá Suarez; A. García Esteve; D. González Abreu; V. López Martínez; J. Aibar Manero; M. Alavés Navarro; M. Lozano Zamora; P. Vilches Benavent; J. Crespo Simó; J. Teruel López

Single culture media are widely used in assisted reproduction, but the optimal formulation remains uncertain. Oxidative stress is a major challenge in embryo culture, and studies show that adding antioxidants to media helps protect embryos and improves their quality.

This prospective randomized study evaluated whether adding antioxidants to embryo culture media improves embryo quality and compared embryo quality between two single culture media using an Artificial Intelligence-based embryo assessment score (iDAScore). The study was conducted in two phases between December 2024 and November 2025. In Phase 1, 70 oocyte donors had their ovaries randomized so that oocytes from each ovary were cultured in either medium A or medium B, which contained antioxidants. This phase included 90 recipients. In Phase 2, 78 whole donors were randomized to either medium A or B, involving 100 recipients. In Phase 1, 2,625 oocytes were retrieved, of which 2,172 (82.74%) were mature. In Phase 2, 2,677 oocytes were retrieved, with 2,116 (79.04%) being mature. Mature oocytes were fertilized by intracytoplasmic sperm injection (ICSI) and cultured under identical laboratory conditions in either medium A or medium B. In Phase 1, 913 mature (MII) oocytes were injected (451 in medium A and 462 in medium B), while 1,023 mature oocytes were analyzed in Phase 2 (568 in medium A and 455 in medium B). Fertilization rates were similar between the two media in both phases (Phase 1: 83.59% vs. 84.41%; p = 0.401 and Phase 2: 82.19% vs. 76.7%; p = 0.12). Post-ICSI oocyte degeneration rates also did not differ significantly between groups (Phase 1: 7.09% vs. 4.97%; p = 0.224 and Phase 2: 3.6% vs. 6.15%; p = 0.068). In Phase 1, a significantly higher proportion of embryos reached the usable blastocyst stage in medium A than in medium B (64.19% vs. 51.53%; p = 0.005). However, no significant difference was observed in Phase 2 (62.09% vs. 61.06%; p = 0.86).A total of 50 embryo transfers were performed in Phase 1 and 75 in Phase 2, with an average of approximately 1.2 blastocysts transferred per patient. No significant differences were observed between the two media in β-human chorionic gonadotropin positivity, clinical pregnancy rates, or clinical miscarriage rates in either phase. Embryo quality assessed using the Artificial Intelligence-based iDAScore was also similar between the two culture media (5.96 ± 2.70 for medium A vs. 5.75 ± 2.50 for medium B; p = 0.46).

Overall, no significant positive effect of antioxidant-containing culture media on in vitro embryo development was observed in this cohort. They suggested that the use of donor patients with high reproductive potential may have contributed to the favorable outcomes regardless of the culture medium. Further studies in patients with poorer reproductive potential are needed to better assess the impact of culture media composition.

Comparison of Clinical Pregnancy Outcomes Between Ultra-Fast and Conventional Warming Protocols Under Fatty Acid–Supplemented Thawing Conditions: A Pseudo-Randomized Prospective Study

Authors: Kento Usui; Y. Fukatsu; A. Abe; A. Okada; S. Tsuchiya; T. Nakagawa; S. Maekawa; S. Kita; A. Takahashi; A. Usui

Ultra‑fast warming has emerged as a way to simplify procedures and reduce embryologist workload while maintaining outcomes comparable to conventional protocols. Current evidence supports similar pregnancy rates between the two approaches, but high‑quality clinical studies under fatty acid–supplemented thawing conditions are lacking. Experimental data suggest that the benefits of fatty acid supplementation may be diminished with shortened warming, though most findings come from in vitro models. Whether these biological effects translate into clinical pregnancy outcomes remains uncertain, highlighting the need for prospective trials focused on live birth and pregnancy outcomes.

This prospective, single-center comparative study evaluated whether ultra-fast warming provides pregnancy outcomes comparable to conventional warming while improving laboratory efficiency during frozen–thawed blastocyst transfer using a fatty acid-supplemented thawing medium. The study included 449 frozen–thawed blastocyst transfer cycles conducted between November 2024 and November 2025. Patients aged 43 years or older and embryos with grade C inner cell mass or trophectoderm were excluded. Participants were assigned to conventional warming (232 cycles) or ultra-fast warming (217 cycles) based on whether their age at oocyte retrieval was odd or even. Conventional warming used a three-step, 10-minute protocol, while ultra-fast warming used a one-step, 1-minute protocol. The primary outcomes were embryo survival, re-expansion, biochemical pregnancy, clinical pregnancy, and ongoing pregnancy rates. Baseline characteristics were similar between the two groups, including age at oocyte retrieval (33.9 ± 3.8 vs. 34.0 ± 4.1 years), age at embryo transfer (34.0 ± 4.1 vs. 33.9 ± 3.8 years), number of previous embryo transfer attempts (1.9 ± 1.4 vs. 1.9 ± 1.3), and iDAScore® v2.0 of transferred embryos (6.70 ± 2.1 vs. 6.40 ± 2.35). Embryo survival after warming was 100% in both groups. However, the re-expansion rate was significantly higher with conventional warming than with ultra-fast warming (82.8% [192/232] vs. 74.1% [160/217]; p < 0.05). Delayed or incomplete re-expansion occurred more frequently following ultra-fast warming.Pregnancy outcomes were comparable between the two groups. Biochemical pregnancy rates were 57.3% with conventional warming and 52.5% with ultra-fast warming (p = 0.661). Clinical pregnancy rates were 53.4% and 47.5%, respectively (p = 0.343), while ongoing pregnancy rates were 45.7% and 42.4% (p = 0.343). Although all pregnancy outcomes numerically favored conventional warming, none of the differences were statistically significant. The study was limited by the absence of live birth data and the exclusion of older patients and lower-grade embryos. Larger multicenter studies including a broader range of patients and embryo grades are needed to further evaluate the two warming protocols.

Overall, under fatty acid-supplemented thawing conditions, ultra-fast warming achieved pregnancy outcomes comparable to conventional warming, although conventional warming resulted in significantly higher embryo re-expansion rates and showed a numerical trend toward better pregnancy outcomes.

Key Predictors of Usable Blastocyst Yield in Highly Purified Human Menopausal Gonadotropin (hp-hMG) Only Cycles: A Multicenter Machine Learning Analysis

Authors: Nevo Itzhak; A. Luz; R. Hourvitz; M. Youngster; A. Hochberg; M. Baum; E. Maman; D. Chuderland Ben Arie; M. Yu; W. Zhou; K. Barletta; E. Hariton; N.P. Polyzos; A. Hourvitz

Estimating usable blastocyst yield is crucial for counseling and laboratory planning, yet clinicians often rely on imperfect proxies such as age, ovarian reserve markers, or oocyte counts, which fail to capture attrition from retrieval to Day 5. Current AI tools mainly predict oocyte yield or pregnancy outcomes, while evidence quantifying usable blastocyst numbers remains scarce—leaving a gap for precise, quantitative tools to guide expectations.

This retrospective multicenter cohort study evaluated the clinical and laboratory factors that best predict the number of usable blastocysts in women undergoing in vitro fertilization (IVF) or intracytoplasmic sperm injection (ICSI) cycles using highly purified human menopausal gonadotropin (hp-hMG) alone. The study analyzed routine clinical and laboratory data collected from IVF clinics in Europe, North America, and Asia between 2014 and 2024. The prediction model was developed using 974 hp-hMG-only cycles and validated in an independent test set of 345 cycles with complete fertilization and usable blastocyst outcome data. The primary outcome was the number of usable blastocysts. The model incorporated maternal age, body mass index (BMI), anti-Müllerian hormone (AMH), antral follicle count (AFC), trigger-day estradiol (E2) and progesterone (P4) levels, follicle counts, starting gonadotropin dose, total and mature (metaphase II [MII]) oocyte counts, and fertilization (two-pronuclear [2PN]) count. An Extreme Gradient Boosting (XGBoost) regression model was used to predict usable blastocyst yield, and feature importance analysis identified the variables with the greatest influence on prediction accuracy. In the independent test set (n = 345), the mean usable blastocyst yield was 2.56. The model demonstrated good predictive performance, with a coefficient of determination (R²) of 0.55 and a mean absolute error (MAE) of 1.23.

Fertilization (2PN) count was identified as the strongest predictor of usable blastocyst yield. Maternal age and the number of mature (MII) oocytes were the next most influential factors and improved prediction even after fertilization data were included. Other important predictors, in descending order of contribution, were total oocytes retrieved, trigger-day estradiol level, and antral follicle count. Body mass index had a relatively small impact on prediction. The study was limited by its retrospective design, inclusion of only hp-hMG-only stimulation cycles, and possible differences in laboratory culture and embryo grading practices across centers. Prospective external validation is needed before the model can be applied in clinical practice.

Overall, integrating fertilization data with maternal age, ovarian reserve markers, and treatment characteristics enabled accurate prediction of usable blastocyst yield, highlighting the potential of artificial intelligence to support treatment planning in assisted reproduction.

ESHRE 2026, July 5^th-8^th, London, UK.