Correlation of maternal factors in egg freezing cycles with in vitro fertilization outcomes of both in vivo and rescue in vitro matured oocytes.

Terry Schlenker: Writing – review & editing, Investigation, Data curation, Conceptualization. Ye Yuan: Writing – review & editing, Writing – original draft, Visualization, Supervision, Software, Project administration, Methodology, Formal analysis, Data curation. Elizabeth A. Jannaman: Formal analysis, Data curation. William B. Schoolcraft: Supervision, Resources, Project administration, Funding acquisition. Jason E. Swain: Writing – review & editing, Supervision, Resources, Project administration, Methodology, Investigation, Data curation, Conceptualization.

A total of 110 patients and 134 IVF cycles were included in this study. The characteristics of the study population, including reasons for oocyte vitrification, trigger type, maternal age at vitrification, maternal age at warming, day 3 follicle-stimulating hormone (FSH), antimüllerian hormone (AMH), antral follicle count (AFC), total FSH, posttrigger estradiol, progesterone, and luteinizing hormone (LH) are summarized in Table 1 . On average, 10.5 IVO oocytes and 3.1 rescue-IVM oocytes were warmed per cycle from the 134 cycles. The median maternal age at the time of warming was 40 years. The oocyte survival, fertilization, and cleavage rates were similar, as well as the D5GQB, total GQB, and euploidy rates were significantly lower in rescue-IVM oocytes than in IVO oocytes. Of five single euploid blastocyst transfers derived from rescue-IVM oocytes, three resulted in fetal cardiac activity and normal live birth. These were not different compared with patients with euploid blastocyst transfers from in vivo matured oocytes. Data are presented in Table 2 . Table 2 Oocyte survival after vitrification and warming, early embryo development, blastocyst euploid rate, and clinical pregnancy rate after euploid blastocyst frozen embryo transfer of rescue in vitro maturation oocytes in comparison with in vivo matured oocytes. Rescue-IVM IVO P value Survival rate 90.8% (376/414) 89.1% (1,255/1,409) .362 Fertilization (2PN zygote) rate 75.8% (285/376) 76.3% (958/1,255) .276 Cleavage rate 73.1% (275/376) 75.8% (951/1,255) .308 Day 5 good-quality blastocyst rate 3.2% (9/285) 11.6% (111/958) <.00001 Total good-quality blastocyst rate 16.8% (48/285) 42.8% (410/958) <.00001 Euploid rate 33.3% (16/48) 53.9% (221/410) .009 Fetal cardiac activity 60.0% (3/5) 63.9% (23/36) 1 Note: IVO = in vivo matured MII; 2PN = 2-pronuclear; rescue-IVM = rescue in vitro maturation. Oocyte survival after vitrification and warming, early embryo development, blastocyst euploid rate, and clinical pregnancy rate after euploid blastocyst frozen embryo transfer of rescue in vitro maturation oocytes in comparison with in vivo matured oocytes. Note: IVO = in vivo matured MII; 2PN = 2-pronuclear; rescue-IVM = rescue in vitro maturation. Using a Spearman correlation, all patient characteristics and developmental data were compared ( Fig. 1 ). Maternal age at vitrification was negatively correlated with the D5GQB (rho = −0.19, P =.047) and total GQB (rho = −0.34, P =.0002) rates of IVO oocytes and total GQB rate (rho = −0.19, P =.043) of rescue-IVM oocytes. For IVO oocytes, human chorionic gonadotropin (hCG)-only trigger had a negative correlation with D5GQB (rho = −0.18, P =.038) and tended to be negatively correlated with the cleavage (rho = −0.17, P =.060) and total GQB (rho = −0.17, P =.060) rates. Lupron-only trigger tended to be positively correlated with the D5GQB rate (rho = 0.17, P =.054). The cleavage rate of rescue-IVM oocytes was negatively correlated with AMH (rho = −0.23, P =.015) and AFC (rho = −0.24, P =.013). No correlation was noted for the patient’s day 3 FSH, total FSH, use of a dual trigger, posttrigger estradiol, progesterone, and LH with any developmental data ( Fig. 1 and Supplemental Table 1 , available online). As show in Supplemental Figure 1 (available online), simple linear regression analysis confirmed that the impact of age at vitrification on these developmental data either was significant (IVO total GQB rate, P <.001) or tended to be significant (IVO D5GQB rate, P =.059; rescue-IVM total GQB rate, P =.077). The rescue-IVM cleavage rate tended to be negatively affected by AMH ( P =.071) and was poorly predicted by AFC ( P =.135). Figure 1 The Spearman correlation matrix comparing patient characteristics and developmental data. The scale represents the Spearman rho values describing the positivity of each correlation. Significant correlations between patient characteristics and developmental data were highlighted ( P <.05). AFC = antral follicle count; AMH = antimüllerian hormone; D3 = day 3; D5 = day 5; E2 = estradiol; FSH = follicle-stimulating hormone; hCG = human chorionic gonadotropin; IVO = in vivo matured MII; LH = luteinizing hormone; P4 = progesterone; r-IVM = rescue in vitro maturation. The Spearman correlation matrix comparing patient characteristics and developmental data. The scale represents the Spearman rho values describing the positivity of each correlation. Significant correlations between patient characteristics and developmental data were highlighted ( P <.05). AFC = antral follicle count; AMH = antimüllerian hormone; D3 = day 3; D5 = day 5; E2 = estradiol; FSH = follicle-stimulating hormone; hCG = human chorionic gonadotropin; IVO = in vivo matured MII; LH = luteinizing hormone; P4 = progesterone; r-IVM = rescue in vitro maturation. We also compared embryo development among the hCG-only, Lupron-only, and dual trigger groups. Although no difference was found in the fertilization rate ( Fig. 2 A), a significantly reduced cleavage rate was found in the hCG-only group compared with that in the dual trigger group ( Fig. 2 B). The hCG-only group also had significantly reduced D5GQB and total GQB rates compared with the Lupron-only group ( Fig. 2 C and D). The ages at vitrification among the three groups were not different ( Fig. 2 E), eliminating the primary confounding factor in these comparisons. The development of rescue-IVM oocytes followed a similar trend. However, because of the small sample size, no significant difference was detected ( Supplemental Fig. 2 ). Among the five blastocysts used for embryo transfer, three were from the Lupron-only group and resulted in two live births, and two were from the dual trigger group and resulted in one live birth. Figure 2 The effects of trigger protocols on the fertilization ( A ), cleavage ( B ), day 5 good-quality blastocyst (D5GQB) ( C ), and total good-quality blastocyst ( D ) rates of in vivo matured oocytes in egg freezing cycles and the maternal age at vitrification (mean ± standard deviation) ( E ) for each group are shown. The P value of the Fisher exact test is presented. hCG = human chorionic gonadotropin. The effects of trigger protocols on the fertilization ( A ), cleavage ( B ), day 5 good-quality blastocyst (D5GQB) ( C ), and total good-quality blastocyst ( D ) rates of in vivo matured oocytes in egg freezing cycles and the maternal age at vitrification (mean ± standard deviation) ( E ) for each group are shown. The P value of the Fisher exact test is presented. hCG = human chorionic gonadotropin.

This retrospective cohort study was approved by the research committee in our center. It included patients who had both normal in vivo matured MII (IVO) oocytes and immature oocytes that underwent rescue-IVM as well as reached the MII stage. Their oocytes were vitrified for various reasons ( Table 1 ) and thawed between 2020 and 2023 to undergo IVF cycles (n = 110 patients). Oocytes that were immature at the time of retrieval were cultured in a proprietary medium that aimed to promote oocyte nuclear maturation. Oocytes that reached MII were vitrified. The median maternal age at the time of vitrification was 36 years. All oocytes were vitrified using the same vitrification/warming protocol. After warming, intracytoplasmic sperm injection and resulting culture conditions were identical for IVO as well as rescue-IVM oocytes. The resulting blastocysts were biopsied and underwent preimplantation genetic testing for aneuploidy before vitrification. Euploid blastocysts from rescue-IVM oocytes were only used for frozen embryo transfer for patients who did not have euploid blastocysts from IVO oocytes. Oocyte survival after vitrification and warming, fertilization (2-pronuclear zygote), cleavage, day 5 good-quality blastocyst (D5GQB, grade ≥3BB) formation, total good-quality blastocyst (GQB) formation, blastocyst euploid rate, and clinical pregnancy with fetal cardiac activity were assessed. Patient characteristic information was obtained retrospectively from the patient electronic medical record and used for correlative analysis with the developmental data. Correlative analyses were completed using a Spearman nonparametric correlation matrix with GraphPad Prism 9.4.1. The direction and strength of the relationship between two variables were represented as rho; a negative value reflected a negative relationship, whereas a positive value reflected a positive relationship. For two variables that were strongly correlated ( P <.05), an additional simple linear regression was performed to show the relationship of the two variables. Developmental data were analyzed as binomial proportions and compared using the Fisher exact test. Significance was noted at P <.05. Table 1 Characteristics of the study population. Variable n Median IQR Vitrification reason — — —  Planned fertility preservation 85 n/a n/a  Male factor (no sperm) 11 n/a n/a  Oncological fertility preservation 3 n/a n/a  Other reasons 8 n/a n/a Diagnosis — — —  Polycystic ovary syndrome 4 n/a n/a  Endometriosis 4 n/a n/a Trigger type — — —  hCG-only 26 n/a n/a  Lupron 32 n/a n/a  Dual 76 n/a n/a Age at vitrification (y) 116 36 35–39 Age at warming (y) 114 40 37–42 D3 FSH (IU) 112 7.1 5.8–8.9 AMH (ng/mL) 116 1.845 1.3–2.9 AFC 116 14 10.2–20 Total FSH (IU) 134 2,700 1,800–3,300 Posttrigger E2 (pg/mL) 134 3,145 2,381–4,880 Posttrigger P4 (ng/mL) 134 6 3.9–8.5 Posttrigger LH (mIU/mL) 99 79.7 56.3–125.1 Note: Data are presented as medians and IQRs unless otherwise indicated. AFC = antral follicle count; AMH = antimüllerian hormone; D3 = day 3; E2 = estradiol; FSH = follicle-stimulating hormone; hCG = human chorionic gonadotropin; IQR = interquartile range; LH = luteinizing hormone; n/a = not available; P4 = progesterone. Characteristics of the study population. Note: Data are presented as medians and IQRs unless otherwise indicated. AFC = antral follicle count; AMH = antimüllerian hormone; D3 = day 3; E2 = estradiol; FSH = follicle-stimulating hormone; hCG = human chorionic gonadotropin; IQR = interquartile range; LH = luteinizing hormone; n/a = not available; P4 = progesterone.

To our knowledge, this report is the first to demonstrate that rescue-IVM oocytes can survive vitrification and warming as well as reach the cleavage stage similarly to their sibling in vivo matured counterparts, although the blastocyst development and euploid rates were lower than those in the IVO oocytes. Importantly, euploid blastocysts derived from rescue-IVM oocytes have reproductive potential similar to the IVO oocytes in the patient population examined in this study. Our findings on vitrified and warmed rescue-IVM oocytes agreed with the earlier observations on rescue-IVM oocytes without vitrification and warming ( 2 , 3 ). Rescue in vitro maturation and second-day intracytoplasmic sperm injection can result in good-quality euploid blastocyst as well as live birth. Although the percentage of patients who may benefit from this practice is low and may vary by program, this approach may be the only chance for patients with poor prognosis to conceive as well as can offer an additional resource for patients banking oocytes. However, the impact of the increased workload and associated costs in the IVF laboratory should not be ignored. Thus, the reproductive potential of these oocytes and resulting embryos as well as their impact on patient success rates may need to be balanced with other logistic considerations. The negative correlations of age at vitrification with blastocyst development of IVO and rescue-IVM oocytes were not surprising. Interestingly, the AMH level tended to negatively impact the embryo cleavage of rescue-IVM oocytes in this study. The AMH level is considered as the most important marker for ovarian reserve but a weak predictor for clinical pregnancy outcome ( 4 ). Our study supported the conclusion that AMH and AFC are poor predictors for embryo development in egg freezing cycles. We also found a negative impact of the hCG-only trigger on embryo development compared with the Lupron-only and dual triggers ( Fig. 2 ). Lupron trigger induces endogenous LH production to promote oocyte maturation and appears to be more favorable for oocyte quality. However, Lupron trigger may not work well for patients who cannot produce sufficient LH; therefore, dual trigger with a low-dose hCG was widely applied and resulted in improved live birth rates in high responders ( 5 ). Interpreting the effects of trigger types on the IVF outcomes is often confounded by mixed patient characteristics and individualized trigger method. It is possible that hCG trigger resulted in a suboptimal oocyte quality that could not recover well from vitrification and had declined embryo development. Studies with a larger sample size and more defined patient characteristics are needed to confirm this observation.

T.S. has nothing to disclose. Y.Y. has nothing to disclose. E.A.J. has nothing to disclose. W.B.S. has nothing to disclose. J.E.S. has nothing to disclose.