Rescue in vitro maturation of germinal vesicle oocytes after ovarian stimulation: the importance of the culture media.

study questionWhat are the nuclear and initial developmental outcomes of rescue-IVM oocytes across multiple commercially available media? SUMMARY ANSWER: Of the 11 different media tested, Medium K provided the highest rescue rate in the shortest time and elicited the highest normal oocyte activation rate (NOAR). WHAT IS KNOWN ALREADY: Following ovarian stimulation, 10-15% of the oocytes are immature with about 10% states in prophase I, i.e. the germinal vesicle (GV) stage. A consensus is still needed whether to rescue these oocytes for clinical use or not, as reports of live births from rescued-metaphase-II (MII) oocytes continues to emerge. These oocytes may be valuable to poor prognosis patients when alternatives are not available and to oncofertility patients. Nonetheless, although rescue in vitro maturation (r-IVM) is experiencing a comeback, clear good practice recommendations regarding inherent protocols are lacking. Moreover, no commercially available culture media exist to support this potentially valuable rescue strategy. STUDY DESIGN, SIZE, DURATION: A prospective experimental study was conducted in a 1-year period at a private IVF center and entailing two consecutive phases. In study phase I, 1570 GV oocytes retrieved after ovarian stimulation from 490 young donors (maximum 4 per donor) were randomly cultured for 24 h in 11 commercially available culture media in a time lapse incubator. The two media eliciting the highest rescue rate in the shortest time were selected for study phase II. In this phase, 105 r-MII oocytes, obtained from 190 GVs rescued in the two chosen media, underwent artificial oocyte activation (AOA) and were further cultured in a time lapse incubator for 24 additional hours in Medium J until time of pronuclear fading (tPNf). PARTICIPANTS/MATERIALS, SETTING,

methodsAll donors (26.1 ± 3.8 years) underwent GnRH antagonist ovarian stimulation protocols with agonist trigger. The oocyte in vivo maturation rate was 80%. In the case of immature oocytes, two to four GV per woman were donated for research and cultured in an time lapse incubator. Time of GV breakdown (tGVBD) and time of the first polar body extrusion (t1PB) were annotated. AOA was conducted according to a previously published protocol. The primary outcome of study phase I was the rescue rate per cultured GV and the nuclear maturation dynamics in each medium studied. The primary outcome of study phase II was the NOAR of r-IVM oocytes in each selected medium. In this phase, time of pronuclear appearance and fading (tPNa and tPNf) and S-phase duration were also annotated. MAIN RESULTS AND THE ROLE OF CHANCE: The GVs cultured in 3 of the 11 tested media showed rescue rates ≥55%. However, the time to reach the r-MII in two of them, namely Medium G and Medium K, was significantly shorter (19.4 ± 0.2 h, 95%CI: 19.0-19.8 h). These two media were selected for study phase II. Following AOA, the NOAR obtained after rescue-IVM was significantly higher in the latter (n = 37/53, 70% vs. n = 21/52, 41%; P = 0.006, Power = 77%). No significant differences were observed in tPNa, tPNf or S-phase duration. LIMITATIONS, REASONS FOR CAUTION: The study was conducted in good-prognosis young oocyte donors and should be confirmed in poor-prognosis and/or advanced maternal age infertile women. Metaphase I (MI) immature oocytes were not included. AOA, which was used to assess initial oocyte competence (i.e. to resume meiosis and form a pronucleus), is useful for initial cytoplasmic competence but is an incomplete approach for the comprehensive assessment of cytoplasmic competence or further embryo development. In addition, tests such as microtubule and nuclear staining in r-MII to assess chromosome misalignment, organelle distribution and activity, non-invasive hyperspectral and AI oocyte analyses, more detailed morphodynamic assessments and, finally, analysis of meiotic segregation errors are mandatory to ensure the safety of r-MII oocytes prior to their potential clinical translation. Furthermore, due to the lack of knowledge regarding the qualitative and quantitative formulation of the commercially available culture media, to unravel the physiological mechanisms underlying the outcomes achieved is challenging. WIDER IMPLICATIONS OF THE

findingsAmong commercially available media not specifically designed for r-IVM, the media with glucose as main source of energy may show reduced rescue effectiveness. Conversely, the media with pyruvate as main source of energy, and combined with low lactate concentrations may elicit more favourable conditions to support both oocytes' nuclear and initial cytoplasmic competence performance in GVs obtained after ovarian stimulation. This study provides first suggestions about which of the suboptimal systems would have the least detrimental effect on oocyte competence, thereby setting the stage for future appraisals for most effective rescue-IVM protocols. STUDY FUNDING/COMPETING INTEREST(S): This study was supported by the Instituto de Salud Carlos III, granted with the European Union (PI22/00924); the 'Agencia Valenciana de la Innovación' under the 'Consolidación de la cadena de valor' program of 2025 (INNCAD/2024/159) that is co-founded by the European Union through the 'Programa Operativo FEDER' de la Comunitat Valenciana 2021-2027, and by IVIRMA Valencia. The authors report no conflicts of interest related with the content of this manuscript. TRIAL REGISTRATION NUMBER: N/A.