Blastocysts Derived from Oocytes Retrieved in Cycles with Elevated Late-Follicular Progesterone Per Follicle Show Reduced Reproductive Potential; An Analysis Based on Progesterone Levels Per Individual Follicle

Since the report by Poter et al. in 1984 [1] describing the gonadotropin-releasing hormone agonist (GnRH-a) long protocol, ovarian stimulation has become a standard practice in assisted reproductive technology (ART). In the 1990s, multiple studies reported that elevated serum progesterone (P) during the long protocol was associated with reduced pregnancy rates. [2-4] at that time-when fresh embryo transfer predominated-this decline was attributed to an asynchrony between progesterone-induced endometrial decidualization and the developmental pace of the embryo. [5]

In 2018, Racca and colleagues showed that elevated latefollicular P was associated with a reduced cumulative pregnancy rate; higher P was also linked to lower embryo utilization on day 3 and day 5. [6] Similarly, Huanf et al. reported that trigger-day P ≥2.5 ng/mL was associated with a lower proportion of good-quality day-3 embryos. [7] Although that study did not assess pregnancy outcomes, fewer good-quality embryos likely translate into poorer clinical results. Most prior work stratified outcomes by triggerday serum P alone. Physiological sources of P include the corpus luteum, adrenal cortex, placenta, and the follicle itself. Because circulating P is often ascribed to the corpus luteum, elevated P during stimulation has traditionally been labeled “premature luteinization.” However, before ovulation granulosa and theca cells have not fully luteinized. An alternative explanation is that the large amount of intrafollicular P diffuses (“seeps”) into the circulation along its steep concentration gradient. Reports indicate that follicular-fluid P can be ≥1,000-fold higher than serum, [8] supporting this hypothesis. To normalize circulating P for follicle burden, we defined progesterone-per-follicle (P/fol) as trigger-day serum P divided by the number of follicles ≥15 mm and examined its association with clinical pregnancy after freeze-all and subsequent frozen-thawed embryo transfer (FET).

Study population and cycle selection

We retrospectively screened 26,279 oocyte pick-up (OPU) cycles (1 January 2020–31 December 2024). To ensure precise ascertainment of oocyte maturity, only ICSI cycles were considered; cycles planned for conventional insemination were excluded. We further excluded cycles requiring surgical sperm retrieval (MESA [microsurgical epididymal sperm aspiration] or TESE [testicular sperm extraction]), cycles using artificial oocyte activation, and fresh-transfer cycles. Among the remaining records, cycles yielding no blastocysts were excluded. The primary analysis set comprised 8,507 cycles; a physiological analytic subset (see below) included 6,908 cycles (Figure 1).

Analytic subset

To minimize confounding from excessive follicular response or peri-ovulatory hormonal surge, we restricted to cycles that, on the trigger day (two days before OPU), met all of the following: <10 follicles ≥15 mm, LH <12 IU/L, and E2 ≤3,000 pg/mL (Figure 1).

Ovarian Stimulation

Baseline assessment occurred on menstrual cycle day (MC) 2–3 (estradiol [E2], follicle-stimulating hormone [FSH], and luteinizing hormone [LH]; transvaginal ultrasound). Stimulation protocols included natural; minimal (clomiphene citrate [CC] or letrozole [LTZ] alone); mild (CC or LTZ plus recombinant FSH); GnRH antagonist; and modified progestin-primed ovarian stimulation (PPOS). Less common GnRH-agonist long/short protocols were grouped as “other.”

• Natural: no medication except for the maturation trigger [9].

• Minimal: CC 50–100 mg/day or LTZ 2.5–5.0 mg/day from MC3 to the day before trigger.

• Mild: CC 100 mg/day or LTZ 5 mg/day for 7 days (MC3– MC9) plus recombinant-FSH with dose adjustments based on MC3 FSH [10].

• PPOS: as in mild, plus oral medroxyprogesterone acetate 10 mg twice daily from MC6 until trigger [11].

• GnRH antagonist: CC 100 mg/day or LTZ 5 mg/day for 7 days with daily rFSH 150–300 IU from MC3; ganirelix acetate 0.25 mg/day added based on monitoring and continued until trigger. Final oocyte maturation was induced when ≥2 follicles reached ≥20 mm using one of: recombinant hCG 250 μg, intranasal GnRH agonist 300 μg twice 30 min apart, or a dual trigger. OPU occurred 35–36 h later [12].

Hormone assays

Baseline E2, FSH, and LH levels (MC3) were measured using commercially available chemiluminescent enzyme immunoassay (CLEIA) kits (AIA-Pack CL®, TOSOH Corporation, Tokyo, Japan). On MC10, serum E2, LH, and P were measured using the same analytical platform. Anti-Müllerian hormone (AMH) was measured using a commercially available immunoassay kit (VIDAS® AMH Assay, bioMérieux Japan, Tokyo, Japan).

Oocyte retrieval and embryo culture

OPU was performed transvaginally under ultrasound guidance using 20- or 21-gauge needles; aspiration was by manual syringe or pump. Anesthesia (intravenous, local, or none with NSAID analgesia) was per patient preference. Cumulus–oocyte complexes were denuded immediately and nuclear maturation assessed. Semen was obtained as fresh or previously cryopreserved ejaculate. To minimize heterogeneity, the present analysis was restricted to ICSI cases, and Piezo-ICSI was used. [13] Embryos were cultured individually to the blastocyst stage under mixed-gas conditions (5% O₂, 6% CO₂, 89% N₂) in a time-lapse incubator. [14] Vitrification occurred on day 5 or 6. [15] Blastocysts were graded by Gardner and Schoolcraft; those ≥4BB were considered good-quality [16] and were eligible for cryopreservation.

Frozen-thawed embryo transfer and outcomes

Blastocysts were thawed by standard protocol. Endometrium was prepared using either ovulatory cycles (natural or LTZ-induced) or hormone-replacement cycles. [17] One or two blastocysts were transferred. Serum hCG was measured 9–10 days after transfer; clinical pregnancy was defined by an intrauterine gestational sac on ultrasound 16–17 days after transfer.

Statistical analysis

Multivariable logistic regression tested the association between log-transformed P/fol and clinical pregnancy, adjusting for female age, male age, gravidity, AMH, BMI, and follicle count (≥15 mm). Two-sided p<0.05 was considered significant.

The distribution of ovarian stimulation protocols among 6,908 cycles was shown in Figure 2. Among 6,908 cycles in the analytic subset, protocol distribution was: natural 4%, minimal 16%, mild 39%, GnRH antagonist 4%, modified PPOS 31%, and other 6% (Figure 2).

Cycle characteristics are summarized in Table 1. Mean (±SD) female age was 39.2 ± 4.5 years and the male partner age was 41.1 ± 7.5 years. Mean body mass index (BMI) was 21.2 ± 2.9 kg/m², gravidity 1.5 ± 1.4, and AMH 1.9 ± 2.1 ng/ml. Overall ART outcomes are summarized in Table 2. On the trigger day, the mean number of follicles ≥15 mm was 3.6 ± 2.2. The mean number of oocytes retrieved was 4.4 ± 3.7, of which 3.4 ± 3.7 were mature (MII). Fresh ejaculated sperm was used in 87.3% of cycles. Mean serum P was 0.55 ng/mL (range 0.05–50.00), and mean P/fol was 0.19 ng/ mL (range 0.01–26.96). The mean number of vitrified blastocysts per cycle was 1.5 ± 1.6. Across 6,908 transfers, 1,408 clinical pregnancies were observed.

Table 1: Baseline demographics and clinical characteristics of the study cohort.

#; mean ± Standard deviation

Table 2: ART outcomes of the cycles under consideration.

#; mean ± Standard deviation

**MII; metaphase II

***follicles; follicles (≥15mm) on the day of maturation trigger

****BT; blastocyst transfer

On the trigger day, the distribution of P/fol was highly rightskewed (Figure 3A). After log transformation, the variable approximated a normal distribution (Figure 3B), fulfilling assumptions for linear regression; we therefore proceeded with correlation and regression analyses. Log-P/fol showed a strong positive correlation with the number of oocytes retrieved (r = 0.416; 95% CI, 0.397–0.436; p < 1×10⁻²⁸⁰; Figure 3C), whereas the association with the oocyte retrieval rate was only weakly positive (Figure 3D). In contrast, higher Log-P/fol was inversely associated with the number of mature (MII) oocytes, the fertilization rate, the number of blastocysts, and the number of good-quality blastocysts (Figure 4A–4D).

In multivariable analysis, higher P/fol remained an independent negative predictor of clinical pregnancy (adjusted OR 0.61; 95% CI 0.46–0.79; p<0.001), alongside female age, AMH, gravidity, and follicle count (Table 3).

Table 3: Independent predictors of clinical pregnancy: multivariable logistic regression (adjusted odds ratios, 95% CI).

In prior reports, elevated circulating P during ovarian stimulation-particularly in the late follicular phase/around triggerhas been likely to adverse reproductive outcomes, even when embryos vitrified at the blastocyst stage and transferred later. The prevailing view is that late-follicular P elevation compromises embryo competence, thereby reducing live birth rates. [6] Consistent with this, higher trigger-day P (e.g., ≥2.5 ng/mL) correlates with a lower proportion of good-quality embryos[7], and morphological signs of granulosa-cell luteinization are associated with diminished blastocyst formation and fewer high-quality blastocysts. [18] These data support an oocyte/embryo-centric effect rather than a purely endometrial milieu.

In our cohort using freeze-all with subsequent FET, higher trigger-day progesterone-per-follicle (P/fol) independently predicted lower implantation (clinical pregnancy) rates (adjusted OR 0.61; 95% CI 0.46–0.79; p<0.001; Table 3). Because this association persists despite transfer in a newly prepared endometrium, endometrial asynchrony alone2 is insufficient; the harm likely occurs at the oocyte/embryo level.

We propose two, non–mutually exclusive hypotheses to explain this association. First, direct P toxicity within a high-P follicular microenvironment may impair oocyte competence. In line with this, trigger-day P ≥2.5 ng/mL is linked to fewer goodquality embryos,[7] and lower P exposure in culture upregulates competence-related transcripts such as OCT4/MATER, promoting cleavage. [19] Second, serum P rise may be a surrogate for compromised follicle physiology-i.e., follicles that poorly retain P (“seepage”), reflecting suboptimal granulosa–theca and cumulusoocyte complex function. Given reports that late-follicular progesterone elevation does not affect embryo euploidy rates, [20] it is plausible that the adverse outcomes reflect intrinsic follicular dysfunction-i.e., follicles that cannot effectively retain progesterone within the antrum-rather than an effect on aneuploidy. We therefore examine the mechanisms by which circulating P-i.e., P that “seeps” from developing follicles into the bloodstream which we term “seeping P”-arises during stimulation and why this phenomenon may adversely influence ART outcomes.

We propose the following mechanism for the observed “seeping P.” Within the two-cell, two-gonadotropin paradigm, inadequate LH signaling limits CYP17A1 activity, curtailing conversion of pregnenolone/progesterone to androstenedione, thereby allowing intrafollicular P to accumulate and seep into the circulation.[21,22] Oocyte-secreted factors (OSFs), including growth differentiation factor 9 (GDF9) and bone morphogenetic protein 15 (BMP15) restrain premature granulosa/cumulus luteinization and reduce P production across species, indirectly supporting a role in limiting serum P elevation.[23-25] Lower follicular-fluid GDF9 with age and associations of higher BMP15 with better ART outcomes suggest that reduced OSF activity characterizes poorer-quality follicles; this provides a plausible basis for our finding that high S-P/fol relates to lower clinical pregnancy rates. Notably, late-follicular P does not appear to alter embryo euploidy, [20] further implicating intrinsic follicular dysfunction rather than aneuploidy effects.

Beyond preventing premature ovulation (LH-centric vigilance), ovarian stimulation should proactively monitor P and P/fol to preempt late-follicular P rise. Earlier triggering, individualized dosing, and protocols that blunt progesterone escape may protect embryo quality. These data motivate prospective trials to test whether P-conscious stimulation improves implantation and livebirth rates.

The authors have no relevant conflicts of interest to declare.

This study was conducted with funding from the authors’ group without any external sources.

The authors thank medical colleagues and laboratory staff at Sugiyama Clinic Shinjuku for collecting clinical data. Englishlanguage editing was assisted by ChatGPT (OpenAI, San Francisco, CA, USA). AI tools were not used for study design, statistical analysis, data interpretation, or the peer-review process. The authors alone are responsible for the integrity and accuracy of all content.

Institutional Review Board of Sugiyama Clinic (18-001).

1. Porter RN, Smith W, Craft IL, Abdulwahid NA, Jacobs HS (1984) Induction of ovulation for in-vitro fertilization using buserelin and gonadotropins. Lancet 2(8414): 1284-1285.
2. Mio Y, Sekijima A, Iwabe T, Onihara Y, Harada T, et al. (1992) Subtle rise in serum progesterone during the follicular phase as a predictor of the outcome of vitro fertilization. Fertil Steril 58: 159-166.
3. Fanchin R, de Ziegler D, Taieb J, Hazout A, Frydman R (1993) Premature elevation of plasma progesterone alters pregnancy rates of in vitro fertilization and embryo transfer. Fertil Steril 59: 1090–1094.
4. Shulman A, Ghetler Y, Beyth Y, Ben Nun I (1996) The significance of an early. (premature) rise of plasma progesterone in in vitro fertilization cycles induced by a “long protocol” of gonadotropin releasing hormone analogue and human menopausal gonadotropins. J Assist Reprod Genet 13: 207-211.
5. Jiang W, LI D, Zhu L, Wang L, Chen L, et al. (2022) Elevated serum progesterone levels on the hCG trigger day have a negative impact on the live birth rate in the first fresh IVF-ET cycle. J Obstet Gynaecol 42: 3503-3508.
6. Racca A, Santos-Riberio S, Munck ND, Mackens S, Drakopoulos P, et al. (2018) Impact of late－follicular phase elevated serum progesterone on cumulative live birth rates: is there a deleterious effect on embryo? Hum Reprod 33: 860-868.
7. Huanf B, Ren X, Wu Li, Zhu L, Xu B, et al. (2016) Elevated progesterone levels on the day of oocyte maturation may affect top quality embryo IVF cycle. PLoS One 11: e0145895.
8. Fauser BC, Van Heusden AM (1997) Manipulation of human ovarian function: physiological concepts and clinical consequences. Endocr Rev 18: 71-106.
9. Gakka Y, Nakagawa K, Watanabe H, Kuroda K, Horikawa T, et al. (2022) Comparison of Pregnancy outcomes between fresh embryo transfer in a natural IVF cycle and IUI cycle among infertile young women. J Reprod Infertil 23: 93-99.
10. Sugiyama R, Nakagawa K, Nishi Y, Ojiro Y, Juen H, et al. (2013) Using a mild ovarian stimulation protocol combined with clomiphene citrate and recombinant follicle-stimulation hormone to determine the optimal number of oocytes needed to achieve pregnancy and reduce the concerns patients. Reprod Med Biol 12: 105-110.
11. Nakagawa K, Horikawa T, Shiobara K, Ishiyama K, Kataoka H, et al. (2024) Mild ovarian stimulation using the aromatase inhibitor letrozole (LTZ) is a good stimulation protocol for to obtain euploid blastocysts for advanced maternal age women. J Gynecol Res Obstet 10: 31-39.
12. Nakagawa K, Oba M, Ehara K, Ishigaki N, Ino N, et al. (2018) Clinical outcomes of assisted reproductive technology treatment by using a self-injection of recombinant human chorionic gonadotropin as the final maturation trigger. Reprod Med Biol 17: 203-208.
13. Hiraoka K, Isuge M, Kamada Y, Kaji T, Suhara T, et al. (2021) Piezo-ICSI for human oocyte. J Vis Exp Apr 20:170.
14. Nakagawa K, Shirai A, Nishi Y, Sugiyama R, Kuribayashi Y, et al. (2010) A study of the effect of an extremely low oxygen concentration on the development of human embryos in assisted reproductive technology. Reprod Med Biol 9: 163-168.
15. Sugiyama R, Nakagawa K, Shirai A, Sugiyama R, Nishi Y, et al. (2010) Clinical outcomes resulting from the transfer of vitrified human embryos using a new device for cryopreservation (plastic blade). J Assist Reprod Genet 27: 161-167.
16. Gardner DK, Lane M, Steavens J, Schlenker T, Schoolcraft WB (2000) Blastocyst quality and the prediction of human implantation and pregnancy. Fertil Steril 73: 1155-1158.
17. Nakagawa K, Ojiro Y, Jyuen H, Nishi Y, Sugiyama R, et al. (2014) Prostaglandin therapy during the proliferative phase improves pregnancy rates following frozen embryo transfer in a hormone replacement therapy. J Obstet Rynaecol Res 40: 1331-1337.
18. Wang M, Xi Q, Yang Q, Li Z, Yang L, et al. (2021) The relationship between a novel evaluation parameter of premature luteinization and IVF outcomes. RBM online 442: 323-331.
19. O Shea LC, Mehta J, Lonergan P, Hensey C, Fair T (2012) Developmental competence in oocyte and cumulus cells: candidate genes and networks. Syst Biol Reprod Med 58: 88-101.
20. Neves AR, Santos Riberio S, García-Martínez S, Devesa M, Soares SR, et al. (2021) The effect of late-follicular phase progesterone elevation on embryo ploidy and cumulative live birth rates. RBM online 43: 1063-1069.
21. Hillier SG, Whitelaw PF, Smyth CD (1994) Follicular oestrogen synthesis: the 'two-cell, two-gonadotrophin' model revisited. Mol Cell Endocrinol 100: 51-54.
22. Erickson GF, Shimasaki S (2001) The physiology of folliculogenesis: the role of novel growth factors. Fertil Steril 76: 943-949.
23. Paulini F, Melo EO (2011) The role of oocyte-secreted factors GDF9 and BMP15in follicular development and oogenesis. Reprod Domest Anim 46:354-361.
24. Trombly DJ, Woodruff TK, Mayo KE (2009) Roles for transforming growth factorbeta superfamily proteins in early folliculogenesis. Semin Reprod Med 27:14-23.
25. Kedem A, Fisch B, Garor R, Ben-Zaken A, Gizunterman T, et al. (2011) Growth differentiating factor 9 (GDF9) and bone morphogenetic protein 15 both activate development of human primordial follicles in vitro, with seemingly more beneficial effects of GDF9. J Clin Endocrinol Metab 96:1246-1254.