Spontaneous ovulation, hormonal profiles, and the impact of progesterone timing variation on outcomes in natural proliferative phase frozen embryo transfer cycles with single euploid blastocyst transfer.

This retrospective study, approved by the Institutional Review Board of Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan (C202405177), examined all NPP-FET cycles involving single euploid blastocyst transfer performed at two centers: (1) Taipei IVF, Center for Reproduction and Genetics (Hwang Jiann-Loung Gynecology and Obstetrics Clinic), and (2) the Department of Obstetrics and Gynecology, Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan, from January 2023 to October 2024. The NPP-FET protocol was first introduced at the ASRM 2021 Annual Meeting [ 7 ], subsequently validated at the ESHRE 2023 Annual Meeting [ 8 ], and published in a peer-reviewed journal [ 9 ], all by the same research group. Since 2021, this protocol has been routinely implemented in our centers, initially using micronized vaginal progesterone (MVP) with favorable clinical outcomes and greater scheduling flexibility. Due to accumulating clinical evidence supporting DYD as an effective alternative, DYD was later offered as an additional clinical option. All data analyzed in this study were retrospectively collected from routine clinical practice, reflecting standard, non-interventional management strategies based on previously established protocols [ 7 ]. The inclusion criteria consisted of women under 45 years of age with regular menstrual cycles (defined as cycle intervals of 21–35 days), undergoing their first NPP-FET cycle and first single euploid blastocyst transfer, following PGT-A. Participants were required to have undergone prior assessment via ultrasonography, hysterosalpingography, hysteroscopy, and hormonal evaluation, including thyroid-stimulating hormone and prolactin levels. Exclusion criteria included patients who underwent oocyte donation, had non-obstructive azoospermia, Asherman’s syndrome, uterine anomalies, untreated hydrosalpinges, untreated uterine myomas that distorted the uterine cavity, or women with polycystic ovary syndrome (PCOS). Of the initially identified 215 participants, 19 were excluded from the final analysis for the following reasons: luteinized unruptured follicles (LUF) ( n  = 7), occurrence of vaginal bleeding ( n  = 4), endometrial thickness less than 7 mm ( n  = 5), and failure of embryos to survive the thawing process ( n  = 3). LUF was diagnosed in the absence of UDO despite a documented LH surge (LH > 17 IU/L, followed by a > 30% decrease in E2 levels the next day) and an elevated serum P4 level > 1.5 ng/mL one to two days after the LH surge onset [ 12 , 13 ]. The diagnosis was further confirmed by the persistent absence of UDO on TVS monitoring for an additional two to three consecutive days, along with a subsequent rise in serum P4 to > 3 ng/mL [ 13 ]. Consequently, 196 subjects were included in the final analysis. All patients underwent controlled ovarian stimulation (COS) using progestin-primed ovarian stimulation (PPOS), a GnRH antagonist or short agonist protocol, tailored to their ovarian reserve and previous response [ 14 , 15 ]. Final oocyte maturation was induced using hCG, a GnRH agonist, or a dual trigger combining hCG (1500 IU) and a GnRH agonist, with transvaginal oocyte retrieval performed 36 h later [ 14 ]. Fertilization was achieved through intracytoplasmic sperm injection (ICSI). Blastocysts underwent biopsy on day 5 or 6, after laser-assisted hatching of the zona pellucida on day 4. Next-generation sequencing (NGS) was used for PGT-A. All biopsied blastocysts were subsequently vitrified on day 5 or 6. Embryo transfer was conducted 3–4 h after the thawing process, which was carried out in the morning. Embryo morphology was graded on the day of vitrification following the criteria established by Gardner et al. [ 16 ]. High-quality blastocysts were defined as those with expansion and hatching status 3–6, and inner cell mass and trophectoderm grade A or B. In contrast, low-quality blastocysts were defined as those with any grade lower than this [ 10 ]. The vitrification/thawing protocol utilized the Cryotop method (Kitazato, Japan), based on the approach described by Kuwayama [ 17 ]. TVS was performed on menstrual cycle day 2 or 3 as a baseline assessment to document the presence or absence of ovarian or CL cysts, in order to clearly identify the growing follicle in the subsequent monitoring. Serum E2, LH, and P4 levels were also measured concurrently. Subsequently, TVS was conducted 3–4 days before the anticipated ovulation. In routine clinical practice, DYD (10 mg four times daily) supplementation was flexibly initiated prior to ovulation when the following predefined clinical criteria were met: leading follicle ≥ 14 mm, endometrial thickness ≥ 7 mm, serum E2 > 150 pg/mL, and P4 < 1.5 ng/mL. These parameters were retrospectively analyzed in the present study. Although minor adjustments in DYD timing were occasionally required due to a clinic policy limiting FET procedures to a maximum of five per day, this constraint was applied uniformly and independently of patient clinical characteristics. Following the initiation of DYD, daily TVS was performed to monitor follicular growth and endometrial thickness until UDO. UDO was defined as the complete disappearance of the follicle, a reduction in its volume with thickening of the follicular wall, or its replacement by an area with a spongy appearance, as described by Wetzels and Hoogland [ 18 ]. Serial serum hormone assessments were conducted daily and continued until three days after UDO. Ovulation was confirmed by UDO accompanied by a corresponding serum P4 level exceeding 3 ng/mL within three days of UDO [ 19 – 21 ]. Single euploid FET was performed on the sixth day of DYD administration, and DYD supplementation was continued until the day of serum pregnancy testing, 11 days post-FET. Upon confirmation of pregnancy, the DYD dosage was reduced to 10 mg twice daily until the 10th week of gestation. To investigate the time interval between the LH surge and the day of UDO, we applied two established LH surge criteria to our participant data. The first criterion, defined by Irani et al. [ 12 ], requires an LH level > 17 IU/L with a subsequent E2 drop of > 30% the following day. The second criterion, from Bartels et al. [ 22 ], defines the LH surge as an LH level > 20 IU/L. We then calculated the time interval between the LH surge (based on each criterion) and the day of UDO. Serum LH and E2 levels were measured using the IMMULITE 2000 assay system (Siemens Medical Solutions, Malvern, PA, USA) in our endocrinology laboratory. Serum P4 levels were measured with the Roche Cobas e801 analyzer assay (Roche Diagnostics GmbH). The intra-assay coefficients of variation (CV) were < 7.1% for LH, < 11% for E2, and  0.97. The primary outcome was the spontaneous ovulation rate. Serum hormone profiles, including E2, LH, and P4, were meticulously assessed around the day of UDO. Secondary outcomes included the clinical pregnancy rate (CPR), the ongoing pregnancy rate (OPR), and clinical loss rate (CLR), and the implantation rate (IR). Clinical pregnancy was defined as the presence of an intrauterine gestational sac on TVS at 5–6 weeks of gestation, while ongoing pregnancy was defined as the progression of fetal cardiac activity beyond 12 weeks of gestation. Clinical loss was defined as a loss of clinical pregnancy before 12 weeks of gestation. Statistical analyses were conducted using SPSS version 21.0 (IBM Corp., Armonk, NY, USA). The normality of continuous variables was assessed with the Kolmogorov-Smirnov test. As the data were not normally distributed, results are presented as medians with interquartile ranges (IQR, 25th–75th percentile) and compared using the Mann-Whitney U test. Categorical variables are expressed as numbers (percentages) and were compared using chi-square tests, with Yates’ correction when appropriate, or Fisher’s exact test for small sample sizes. Multivariable logistic regression models were constructed separately for each pregnancy outcome, including CPR, OPR, and CLR to assess associations with relevant clinical parameters. A two-step modeling strategy was employed. First, univariable logistic regression analyses were conducted to explore the relationship between each independent variable and each pregnancy outcome. Variables with a P-value < 0.10 were selected for multivariable analysis. In the second step, multivariable logistic regression models were developed to adjust for potential confounding factors. All clinically relevant variables were included regardless of univariable significance to ensure comprehensive control for confounding. Key variables of interest reflected variation in the timing of P4 initiation, including follicular phase duration, serum E2 levels and follicular diameter on the day of P4 initiation, the interval from P4 initiation to UDO, the interval from UDO to FET, and serum P4 levels on the day of FET. Because the interval from P4 initiation to UDO and the interval from UDO to FET always summed to a fixed value of 5 days, they were mathematically collinear. Both were evaluated in univariable analysis, but to avoid multicollinearity, only the interval from P4 initiation to UDO was included in the multivariable model. Follicular diameter at the time of P4 initiation was analyzed in two separate models: once as a continuous variable and once as a categorical variable (14.0–15.9 mm vs. ≥16.0 mm), reflecting clinical practice in which hCG is typically administered when the dominant follicle reaches 16–20 mm in modified NC-FET protocols [ 1 ]. Similarly, serum P4 levels on the day of FET were analyzed both as a continuous variable and as a categorical variable (< 10 ng/mL vs. ≥10 ng/mL), based on prior evidence that P4 levels below 10 ng/mL are negatively associated with live birth in NC-FET cycles [ 23 ]. Additional covariates known to influence pregnancy outcomes were also incorporated: age at oocyte retrieval, body mass index (BMI), anti-Müllerian hormone (AMH) levels, infertility duration, type of infertility (primary vs. secondary), parity, infertility etiology, follicular diameter one day before UDO, endometrial thickness on the day of FET, embryo morphology (high-quality vs. low-quality blastocysts), and day of blastocyst biopsy (Day 5 vs. Day 6). These final multivariable models, constructed in the second step, were adjusted for all covariates to better estimate the independent effect of variation in P4 initiation timing on pregnancy outcomes. Results were reported as adjusted odds ratios (aORs) with 95% confidence intervals (CIs). In this study, a two-sided P-value of < 0.05 was considered statistically significant.

Spontaneous ovulation, confirmed by both UDO and a P4 assessment exceeding 3.0 ng/ml within 3 days post-UDO, was observed in all subjects (196/196 = 100%). TVS demonstrated that complete disappearance of the follicle was the predominant sign of UDO, observed in 74.5% (146/196) of cases. In 25.5% (50/196), ovulation was confirmed by either a reduction in follicular volume with thickening of the follicular wall or its replacement by a spongy-appearing area. The proportion of subjects with serum P4 > 3.0 ng/mL increased progressively post-UDO: 1.5% (3/196) on the day of UDO, 66.8% (131/196) one day after, 96.4% (189/196) two days after, and 100% by the third day. Together with UDO, serum P4 assessment consistently confirmed spontaneous ovulation in all participants, validating the robustness of our dual-confirmation strategy. Serum E2, LH, and P4 levels across the peri-ovulatory period are presented in Table  1 . As shown in Fig.  1 , serum E2 and LH peaked two days before UDO and declined thereafter, while P4 levels rose progressively following UDO, reflecting the hormonal dynamics of natural cycles. The median diameter of the dominant follicle one day before UDO was 18.6 mm (IQR: 17.8–19.6 mm) (Table  2 ), consistent with expected ovulatory development. Table 1 Summary of median concentrations and the 5th, 25th, 75th, and 95th percentiles of E2, LH, and P4 levels around the day of ultrasound-documented ovulation (UDO) in all subjects (N = 196) E2 (pg/ml) LH (mIU/ml) P4 (ng/ml) Median Percentile Median Percentile Median Percentile 5th 25th 75th 95th 5th 25th 75th 95th 5th 25th 75th 95th UDO-2 days 320.0 190.3 247.3 414.5 504.9 29.7 11.2 18.8 40.6 61.5 0.31 0.07 0.16 0.55 0.84 UDO-1 day 228.0 85.7 164.0 298.0 430.0 24.9 10.4 18.2 46.3 64.9 0.71 0.32 0.53 0.93 1.27 UDO 78.3 40.3 53.0 99.4 184.0 11.7 3.9 8.9 16.6 21.3 1.59 0.61 0.84 2.18 2.92 UDO + 1 day 90.2 46.6 72.0 114.0 187.4 9.1 4.7 6.2 13.2 16.4 3.68 1.53 2.26 5.03 7.70 UDO + 2 days 129.5 59.2 95.2 168.0 245.4 7.2 3.3 5.0 9.7 12.9 6.38 2.96 4.45 8.66 13.87 UDO + 3 days 153.0 75.0 112.3 200.0 289.0 5.9 2.6 3.6 7.7 9.9 9.24 4.11 6.09 11.90 15.10 E2, estradiol; LH, luteinizing hormone; P4, progesterone Summary of median concentrations and the 5th, 25th, 75th, and 95th percentiles of E2, LH, and P4 levels around the day of ultrasound-documented ovulation (UDO) in all subjects (N = 196) E2, estradiol; LH, luteinizing hormone; P4, progesterone Fig. 1 Boxplots showing median serum hormone levels (E2, LH, and P4) relative to the day of ultrasound-documented ovulation (UDO) in natural proliferative phase frozen embryo transfer (NPP-FET) cycles ( N  = 196). E2, estradiol; LH, luteinizing hormone; P4, progesterone Boxplots showing median serum hormone levels (E2, LH, and P4) relative to the day of ultrasound-documented ovulation (UDO) in natural proliferative phase frozen embryo transfer (NPP-FET) cycles ( N  = 196). E2, estradiol; LH, luteinizing hormone; P4, progesterone Table 2 Patient demographics and baseline characteristics stratified by ongoing pregnancy outcome All subjects ( N  = 196) Ongoing pregnancy positive ( N  = 115) Ongoing pregnancy negative ( N  = 81) P value Female age at oocyte retrieval (years) 39.0 (36.0–40.0) 39.0 (36.0–40.0) 39.0 (36.0–41.0) 0.633 BMI (kg/m 2 ) 21.0 (19.5–23.8) 20.9 (19.5–23.9) 21.2 (19.6–23.7) 0.892 AMH (ng/mL) 2.89 (2.27–3.52) 2.88 (2.19–3.56) 2.98 (2.32–3.49) 0.826 Duration of infertility (years) 3.0 (2.0–5.0) 3.0 (2.0–5.0) 3.0 (2.0–5.0) 0.904 Type of infertility 0.807  -Primary 119 (60.7%) 69 (60.0%) 50 (61.7%)  -Secondary 77 (39.3%) 46 (40.0%) 31 (38.3%) Parity 0.859  -P0 139 (70.9%) 81 (70.4%) 58 (71.6%)  - P  ≥ 1 57 (29.1%) 34 (29.6%) 23 (28.4%) Etiology of infertility  -Male factor 75 (38.3%) 42 (36.5%) 33 (40.7%) 0.550  -Ovulatory dysfunction 24 (12.2%) 15 (13.0%) 9 (11.1%) 0.684  -Tubal factor 17 (8.7%) 9 (7.8%) 8 (9.9%) 0.615  -Advanced maternal age 44 (22.4%) 28 (24.3%) 16 (19.8%) 0.448  -Diminished ovarian reserve 16 (8.2%) 8 (7.0%) 8 (9.9%) 0.462  -Endometriosis 11 (5.6%) 7 (6.1%) 4 (4.9%) 1.0  -Unexplained 6 (3.1%) 4 (3.5%) 2 (2.5%) 1.0  -Others 3 (1.5%) 2 (1.7%) 1 (1.2%) 1.0 Follicular diameter one day prior UDO (mm) 18.6 (17.6–19.6) 18.6 (17.8–19.6) 18.6 (17.6–19.9) 0.527 Endometrial thickness on the day of FET (mm) 11.5 (9.7–12.9) 11.5 (9.6–12.9) 11.5 (9.8–13.0) 0.701 Embryo morphology < 0.001  -High-quality blastocysts 170 (86.7%) 110 (95.6%) 60 (74.1%)  -Low-quality blastocysts 26 (13.3%) 5 (4.3%) 21 (25.9%) Day of embryo biopsy < 0.001  -Day 5 141 (71.9%) 98 (85.2%) 43 (53.1%)  -Day 6 55 (28.1%) 17 (14.8%) 38 (46.9%) Values are expressed as median (IQR, 25th-75th percentiles), or n (%) BMI, body mass index; AMH, anti- Müllerian hormone; UDO, ultrasound-documented ovulation; FET, frozen embryo transfer Patient demographics and baseline characteristics stratified by ongoing pregnancy outcome Values are expressed as median (IQR, 25th-75th percentiles), or n (%) BMI, body mass index; AMH, anti- Müllerian hormone; UDO, ultrasound-documented ovulation; FET, frozen embryo transfer The interval between the LH surge and UDO is summarized in Table  3 based on two established criteria. According to Irani et al. [ 18 ] (LH > 17 IU/L with a ≥ 30% E2 drop the following day), 96.4% ovulated within 1–2 days of the surge. Using Bartels’ criterion [ 19 ] (LH > 20 IU/L), 92.2% ovulated within 1–2 days. Notably, four subjects ovulated before reaching an LH level of 20 IU/L. Table 3 Interval between the luteinizing hormone (LH) surge, defined by two distinct criteria, and the day of ultrasound-documented ovulation (UDO) LH criterion UDO occurred 1 day after LH surge UDO occurred 2 days after LH surge UDO occurred 3 days after LH surge UDO occurred 4 days after LH surge LH > 17 IU/L with E2 drop > 30% next day* 97 (49.5%) 92 (46.9%) 7 (3.6%) 0 (0%) LH > 20 IU/L** 54 (28.1%) 123 (64.1%) 13 (6.8%) 2 (1.0%) Data are presented as n (%) * N  = 196 ** N  = 192, UDO occurred before LH > 20IU/L in 4 subjects, reducing the final analysis cohort to 192 subjects Interval between the luteinizing hormone (LH) surge, defined by two distinct criteria, and the day of ultrasound-documented ovulation (UDO) Data are presented as n (%) * N  = 196 ** N  = 192, UDO occurred before LH > 20IU/L in 4 subjects, reducing the final analysis cohort to 192 subjects Pregnancy outcomes were as follows: CPR 66.3% (130/196), OPR 58.7% (115/196), and CLR 11.5% (15/130). Patient demographics and baseline characteristics for the entire cohort, stratified by ongoing pregnancy status, are presented in Table  3 . Patients with ongoing pregnancies had significantly higher proportions of high-quality blastocysts and Day 5 biopsied embryos, whereas other characteristics did not differ significantly. Clinical parameters reflecting variation in P4 initiation timing are presented in Table  4 . No significant differences were observed in any P4-related variable between the ongoing pregnancy and non-ongoing pregnancy groups. Table 4 Clinical parameters reflecting variation in the timing of progesterone (P4) initiation, stratified by ongoing pregnancy outcome All subjects ( N  = 196) Ongoing pregnancy positive ( N  = 115) Ongoing pregnancy negative ( N  = 81) P value Follicular phase duration (days) 12.0 (8.0–15.0) 12.0 (8.0–15.0) 12.0 (8.0–14.0) 0.470 Serum E2 on day of P4 initiation (pg/mL) 231.9 (195.3–284.0) 231.8 (191.0-288.0) 232.0 (199.8–283.0) 0.626 Follicular diameter on day of P4 initiation (mm) 16.6 (15.1–17.8) 16.6 (15.0-17.8) 16.7 (15.5–17.8) 0.609 Interval from P4 initiation to UDO (days) 2.0 (2.0–3.0) 2.0 (2.0–3.0) 2.0 (2.0–3.0) 0.946 Interval from UDO to FET (days) 3.0 (2.0–3.0) 3.0 (2.0–3.0) 3.0 (2.0–3.0) 0.946 Serum P4 on day of FET (ng/mL) 8.1 (6.4–11.2) 7.9 (6.4–11.2) 8.3 (6.4–11.4) 0.711 Values are presented as median (IQR, 25th-75th percentiles) E2, estradiol; P4, progesterone; UDO, ultrasound-documented ovulation; FET, frozen embryo transfer Clinical parameters reflecting variation in the timing of progesterone (P4) initiation, stratified by ongoing pregnancy outcome Values are presented as median (IQR, 25th-75th percentiles) E2, estradiol; P4, progesterone; UDO, ultrasound-documented ovulation; FET, frozen embryo transfer The analyses shown in Tables  3 and 4 were repeated using CPR and CLR as alternative outcome measures in place of OPR. Similar findings were observed, with embryo morphology and biopsy day remaining the only variables showing significant associations (data not shown). In univariable logistic regression, only embryo morphology and biopsy day were significantly associated with all four outcomes ( P  < 0.10) and were included in the multivariable analysis. Although P4 timing–related variables were not significant in univariable analyses, they were retained in multivariable models to ensure comprehensive adjustment. As shown in Table  5 , after adjusting for relevant covariates, none of the P4 timing–related parameters—including follicular phase duration, serum E2 levels and follicular diameter on the day of P4 initiation, the interval from P4 initiation to UDO, the interval from UDO to FET, and serum P4 levels on the day of FET—were significantly associated with any pregnancy outcome. Adjusted odds ratios for these variables were consistently near 1.0, with confidence intervals including unity and P values exceeding 0.25 across all pregnancy outcomes. While the intervals for serum E2 were relatively narrow, they still included 1.0, indicating no significant relationship with outcomes. By contrast, embryo morphology and biopsy day remained independently associated with all pregnancy outcomes. Supplementary analysis of follicular size and serum P4 categories . Table 5 Multivariable logistic regression analysis of factors associated with pregnancy outcomes Ongoing pregnancy rate Clinical pregnancy rate Clinical loss rate aOR (95% CI) P value aOR (95% CI) P value aOR (95% CI) P value Embryo morphology* 5.165 (1.708, 15.614) 0.004 2.771 (1.075, 7.143) 0.035 9.179 (1.326, 63.516) 0.025 Day of embryo biopsy** 4.203 (2.004, 8.814) < 0.001 2.357 (1.151, 4.826) 0.019 8.051 (1.937, 33.460) 0.004 Follicular phase duration (days) 1.040 (0.950, 1.139) 0.398 1.024 (0.937, 1.119) 0.601 1.070 (0.883, 1.295) 0.491 Serum E2 on day of P4 initiation (pg/mL) 1.000 (0.993, 1.006) 0.887 0.998 (0.992, 1.004) 0.542 1.006 (0.993, 1.020) 0.367 Follicular diameter on day of P4 initiation (mm) 0.977 (0.700, 1.363) 0.891 0.932 (0.675, 1.287) 0.670 0.914 (0.415, 2.010) 0.822 Interval from P4 initiation to UDO (days) 0.998 (0.578, 1.724) 0.995 1.086 (0.630, 1.871) 0.767 0.747 (0.234, 2.388) 0.623 Serum P4 on day of FET (ng/mL) 1.065 (0.923, 1.229) 0.387 1.085 (0.940, 1.252) 0.268 1.009 (0.730, 1.395) 0.957 aOR, adjusted odds ratio; CI, confidence interval; E2, estradiol; P4, progesterone; FET, frozen embryo transfer Analysis adjusted for age at oocyte retrieval, body mass index (BMI), anti-Müllerian hormone (AMH) levels, infertility duration, type of infertility (primary vs. secondary), parity, infertility etiology, follicular diameter one day before UDO, endometrial thickness on the day of FET, embryo morphology (high-quality vs. low-quality blastocysts), and day of blastocyst biopsy (Day 5 vs. Day 6) * High-quality blastocysts vs. low-quality blastocysts (reference) ** D5-biopsy vs. D6-biopsy (reference) Multivariable logistic regression analysis of factors associated with pregnancy outcomes aOR, adjusted odds ratio; CI, confidence interval; E2, estradiol; P4, progesterone; FET, frozen embryo transfer Analysis adjusted for age at oocyte retrieval, body mass index (BMI), anti-Müllerian hormone (AMH) levels, infertility duration, type of infertility (primary vs. secondary), parity, infertility etiology, follicular diameter one day before UDO, endometrial thickness on the day of FET, embryo morphology (high-quality vs. low-quality blastocysts), and day of blastocyst biopsy (Day 5 vs. Day 6) * High-quality blastocysts vs. low-quality blastocysts (reference) ** D5-biopsy vs. D6-biopsy (reference) To further test the robustness of this finding, follicular diameter at P4 initiation was recoded as a categorical variable (14.0–15.9 mm vs. ≥16.0 mm) and reanalyzed in a separate multivariable model. Results remained non-significant across all outcomes, confirming that follicular diameter at the time of P4 initiation did not influence outcomes. Similarly, serum P4 levels on the day of FET were assessed as a categorical variable (< 10 ng/mL vs. ≥10 ng/mL). This analysis also showed no significant associations with any pregnancy outcome, further supporting the lack of impact of serum P4 levels at transfer on clinical results.

The use of frozen embryo transfer (FET) cycles has increased substantially worldwide over the past decade, driven by advancements in vitrification techniques, the adoption of ”freeze-all” strategies, the widespread implementation of preimplantation genetic testing for aneuploidies (PGT-A), and the growing utilization of surplus embryos [ 1 ]. The two primary approaches for preparing the endometrium for FET are hormone replacement therapy FET (HRT-FET) and natural cycle FET (NC-FET) [ 1 ]. HRT-FET offers scheduling flexibility but has been associated with increased risks of hypertensive disorders of pregnancy, including preeclampsia, as well as higher rates of placenta previa, postpartum hemorrhage, cesarean delivery, and adverse neonatal outcomes such as macrosomia, compared to NC-FET [ 2 ]. One potential explanation for these outcomes is the absence of factors derived from the corpus luteum (CL), such as relaxin and other vasoactive substances, in HRT-FET cycles. This deficiency may impair the maternal cardiovascular adaptations necessary for a healthy pregnancy [ 3 , 4 ]. Accumulating evidence supports the use of NC-FET, which allows for physiological follicle development and CL formation, potentially mitigating the risks observed with HRT-FET [ 5 ]. Challenges associated with NC-FET, especially true NC-FET cycles, include the need for precise ovulation tracking and limited scheduling flexibility [ 5 ]. A proof-of-concept study demonstrated the feasibility of initiating progesterone (P4) supplementation during the follicular phase to guide the timing of embryo transfer [ 6 ]. In this approach, P4 supplementation was initiated without human chorionic gonadotropin (hCG) administration once the following criteria were met: endometrial thickness ≥ 7 mm, a dominant follicle ≥ 12 mm, and estradiol (E2) levels > 80 pg/mL. Transfer of day 2 cleavage-stage embryos was subsequently performed on the third day of P4 supplementation. While ovulation was not directly assessed, the absence of luteal phase estrogen supplementation and the favorable clinical outcomes suggested sufficient CL activity [ 6 ]. A similar approach, referred to as “natural proliferative phase frozen embryo transfer” (NPP-FET), was introduced by Mendes Godinho et al. [ 7 – 9 ]. In this protocol, P4 supplementation was initiated during the proliferative phase prior to ovulation, without hCG triggering, and with flexible timing. P4 was commenced once the following criteria were met: endometrial thickness ≥ 7 mm, P4 ≤ 1.5 ng/mL, and dominant follicle diameter ≥ 14 mm, regardless of serum luteinizing hormone (LH) levels. Blastocyst transfer was performed on the sixth day of P4 exposure. Although the presence of a CL was reportedly observed on early gestational ultrasounds in some pregnancies following NPP-FET, the incidence and characteristics of CL formation were not systematically assessed [ 9 ]. Another similar method, the progesterone-modified natural cycle (P4mNC) for blastocyst transfer, initiated P4 supplementation when the dominant follicle diameter exceeded 16 mm and the endometrial thickness was ≥ 7 mm [ 10 ]. However, the occurrence of spontaneous ovulation was not formally documented [ 10 ]. To date, only three peer-reviewed studies have investigated NPP-FET or comparable protocols involving flexible P4 initiation during the follicular phase [ 6 , 9 , 10 ]. These approaches aim to combine the scheduling flexibility of HRT-FET with the physiologic benefits of preserving spontaneous ovulation, as seen in natural cycles. They are based on the premise that initiating P4 during the follicular phase can transform the endometrium into a receptive luteal phase while allowing spontaneous ovulation and corpus luteum formation to occur. However, none of these studies objectively confirmed ovulation using imaging or serum P4 measurement [ 6 , 9 , 10 ], leaving key physiologic assumptions unverified. This gap underscores the need for further research to rigorously assess whether spontaneous ovulation is consistently preserved in such protocols. Eggersmann et al. [ 11 ] presented preliminary findings at the 2024 ESHRE Annual Meeting, reporting a 94.4% spontaneous ovulation rate in patients undergoing an NPP-FET cycle; however, peer-reviewed publication of these findings is still pending. Their work represented the first attempt to specifically assess spontaneous ovulation under flexible P4 initiation strategies during the follicular phase. In their protocol, dydrogesterone (DYD; Duphaston, Abbott Biologicals, Netherlands) was initiated flexibly once the following criteria were met: a dominant follicle > 16 mm, an endometrial thickness > 6.0 mm, and serum E2 levels > 180 pg/mL. Embryo transfer was performed between day 2 and day 5 of development. Notably, ovulation verification relied on serum P4 measurements, without explicitly stating a threshold for confirmation, assessed either on the day of FET or during the late luteal phase [ 11 ]. Since DYD is chemically distinct from endogenous P4, serum P4 levels attributable to CL function could still be accurately measured during DYD administration, facilitating ovulation assessment [ 11 ]. Building on this early work [ 11 ], our study applied a more robust dual-confirmation strategy to verify spontaneous ovulation, using transvaginal ultrasonography (TVS) and subsequent serum P4 measurements > 3.0 ng/mL following ultrasound-documented ovulation (UDO). Furthermore, P4 supplementation was initiated once the leading follicle reached a diameter ≥ 14 mm, in conjunction with defined hormonal and sonographic criteria. Additionally, we characterized peri-ovulatory hormonal dynamics in NPP-FET cycles to further define the endocrine profile associated with this protocol. By examining changes in serum E2, LH, and P4 levels surrounding the ovulatory period, we sought to evaluate whether early P4 supplementation could preserve the hormonal milieu typically observed in a natural cycle. Recognizing that flexible P4 initiation introduces variability in cycle parameters, we also investigated whether variation in P4 initiation timing influences clinical outcomes by evaluating several potentially predictive factors related to P4 timing. To minimize embryo-related confounding, the study was restricted to first NPP-FET cycles involving single euploid blastocyst transfers. Given that the defining advantage of NPP-FET lies in its scheduling flexibility, this third objective aimed to test whether clinical outcomes remain consistent across a spectrum of P4 initiation timings. Therefore, to our knowledge, this is the first study to comprehensively evaluate: (1) the occurrence of spontaneous ovulation in NPP-FET cycles where P4 was initiated at a follicular diameter < 16 mm; (2) the peri-ovulatory hormonal dynamics in these cycles; and (3) the impact of variation in P4 initiation timing on pregnancy outcomes. By addressing these three key knowledge gaps, we aimed to assess whether NPP-FET represents a physiologically grounded protocol that offers greater scheduling flexibility while maintaining consistent clinical outcomes across a range of P4 initiation timings.

In this study, all patients achieved spontaneous ovulation when DYD was flexibly initiated upon meeting predefined physiologic criteria, including a leading follicle diameter ≥ 14 mm, endometrial thickness ≥ 7 mm, serum E2 > 150 pg/mL, and P4 < 1.5 ng/mL. Peri-ovulatory hormone profiles also closely resembled those of natural cycles, further supporting the physiologic integrity of this protocol. To our knowledge, this is the first study to specifically evaluate spontaneous ovulation in NPP-FET cycles with P4 initiation at a follicular diameter < 16 mm. Importantly, we also found that variation in progesterone initiation timing was not associated with pregnancy outcomes, suggesting that NPP-FET can be flexibly scheduled without compromising clinical success. Together, these results position NPP-FET as a clinically adaptable, physiologically sound, and scheduling-friendly protocol for FET. In our study, DYD was chosen to allow accurate measurement of serum P4 from the CL. DYD has previously demonstrated a favorable safety profile, supported by a scoping review and meta-analysis showing no association between its first-trimester use and fetal abnormalities [ 24 ]. However, a recent disproportionality analysis of the World Health Organization’s global pharmacovigilance database (VigiBase) reported a higher frequency of certain congenital anomalies—most notably hypospadias and congenital heart defects (CHDs)—in pregnancies exposed to DYD, particularly when compared to progesterone [ 25 ]. While this finding does not imply causality, it underscores the need for continued safety monitoring. Previous studies in HRT-FET and NC-FET cycles have used DYD doses of 30–40 mg/day [ 26 ], with some suggesting 30 mg/day may be insufficient for certain patients [ 27 ]. Given initial uncertainty regarding the preservation of spontaneous ovulation with earlier P4 initiation, we adopted a conservative approach using 40 mg/day. Since all patients in our study ovulated spontaneously, a lower dose of 30 mg/day—similar to that used by Eggersmann et al. [ 11 ]—may be sufficient and warrants further evaluation in future protocol optimization. In our study, TVS was employed to confirm ovulation due to its well-established accuracy. We also applied a serum P4 threshold > 3 ng/mL as a hormonal marker for ovulation, following recommendations from the American Society for Reproductive Medicine (ASRM) [ 20 ] and prior studies [ 19 , 21 ]. This criterion was met in all subjects within three days after UDO. The 5.6% non-ovulation rate reported by Eggersmann et al. [ 11 ] contrasts with our 100% ovulation rate, possibly attributable to more frequent peri-ovulatory monitoring, smaller sample size, or differences in patient population. Taken together, our findings and those of Eggersmann et al. [ 11 ] suggest that NPP-FET may preserve spontaneous ovulation in a substantial proportion of patients undergoing DYD supplementation when the leading follicle measures 14–16 mm or greater. Weiss et al. [ 6 ] initiated P4 supplementation at follicle sizes as small as 12 mm; the impact of this threshold on spontaneous ovulation warrants further investigation. In our study, the median follicular diameter one day prior to UDO was 18.6 mm, consistent with the 17–24 mm range typically observed in natural ovulatory cycles [ 28 ]. Additionally, the peri-ovulatory hormone levels (Table  1 ) closely matched those reported in previous studies of natural menstrual cycles [ 29 – 31 ]. Figure  1 illustrates dynamic serum hormonal profiles relative to the day of UDO, demonstrating a pattern characteristic of spontaneous ovulatory cycles. Our study also revealed that the interval between the LH surge and UDO was generally 1–2 days (Table  2 ). This aligns with a recent systematic review and meta-analysis by Erden et al. [ 32 ], which reported a mean interval of 33.9 h (range: 22–56 h) between the LH surge and ovulation in natural cycles. These findings confirm that flexible DYD initiation preserves the endocrine and follicular dynamics characteristic of natural ovulation without disrupting ovulation timing, thereby supporting the physiologic integrity of the NPP-FET protocol. Adequate estrogen priming—both in terms of threshold and duration—is essential for optimal endometrial proliferation, glandular development, and progesterone receptor expression, all of which are critical for successful transition into the secretory phase [ 33 , 34 ]. Serum E2 levels of 50–100 pg/mL are generally considered sufficient to initiate this process [ 35 , 36 ]. However, higher levels—typically exceeding 100 pg/mL—are commonly observed during dominant follicle development, particularly once the follicular diameter surpasses 12 mm [ 36 , 37 ]. Notably, recent studies have shown that even follicles as small as 13 mm can yield high-quality, euploid blastocysts [ 38 ]. Based on this evidence, P4 supplementation in our protocol was initiated once the leading follicle reached ≥ 14 mm and serum E2 exceeded 150 pg/mL, thereby ensuring adequate estrogen priming and endometrial preparation. At the time of P4 initiation, the median E2 level was 231.9 pg/mL and the median follicular diameter was 16.6 mm, supporting the appropriateness of the hormonal environment for successful luteal transition (Table  4 ). NC-FET and NPP-FET differ fundamentally in the timing and source of P4 exposure. In NC-FET, the luteal phase begins with spontaneous ovulation, while in NPP-FET, it is initiated by exogenous P4 prior to ovulation. This leads to a shorter and more variable proliferative phase, resulting in individual differences in E2 exposure during the proliferative phase and cumulative exogenous and endogenous P4 exposure by the day of embryo transfer. To assess whether these physiological differences impact clinical outcomes, we conducted multivariable logistic regression analysis. Notably, none of the P4 timing–related parameters—follicular phase duration, serum E2 levels and follicular diameter on the day of P4 initiation, the interval from P4 initiation to UDO, the interval from UDO to FET, or serum P4 levels on the day of FET—were significantly associated with any pregnancy outcomes, including OPR, CPR, or CLR. Instead, embryo morphology and Day 5 biopsy emerged as the only independent predictors (Table  5 ), consistent with a large cohort study of 1,037 single euploid transfers, which reported that both parameters were independently associated with improved pregnancy outcomes [ 39 ]. These findings suggest that, when adequate E2 priming and luteal support are provided, pregnancy outcomes in NPP-FET remain similar across a range of P4 initiation timings. This supports the clinical feasibility of scheduling flexibility within this protocol. Importantly, these conclusions are drawn from comparisons within NPP-FET cycles and do not reflect comparisons to NC-FET or HRT-FET protocols. As summarized in Table  6 , prior studies investigating flexible P4 initiation in NPP-FET have used heterogeneous criteria for initiating P4, including follicular diameter thresholds (12–16 mm), serum E2 levels (ranging from 80 to 180 pg/mL or not reported), and different P4 formulations. Weiss et al. explored feasibility in cleavage-stage transfers in a prospective case series proof-of-concept study; Mendes Godinho et al. and Kornilov et al. assessed retrospective studies comparing blastocyst outcomes versus NC- or HRT-FET; and Eggersmann et al., though only in abstract form, was the first to attempt ovulation confirmation using dydrogesterone [ 6 , 9 – 11 ]. Despite methodological differences, all studies reported favorable reproductive outcomes. Collectively, these findings support the feasibility of flexible P4 initiation across diverse physiologic thresholds, consistent with our own data. Table 6 Summary of published studies on natural proliferative phase frozen embryo transfer (NPP-FET) Authors &Ref Study design N Criteria of progesterone initiation Progesterone Supplement Stage of transferred embryos Verification of ovulation Weiss et al., 2021 (Ref [ 6 ]) Prospective 42 Leading follicle ≧ 12 mm EM ≧ 7.0 mm E2 > 80 pg/ml Endometrin 100 mg TID P4 continued till 12 weeks of gestation Day2 cleavage stage embryos Nil Mendes Godinho et al., 2024 (Ref [ 9 ]) Retrospective 369 Leading follicle ≧ 14 mm EM ≧ 7.0 mm P4 ≦ 1.5ng/ml irrespective of LH level Vaginal micronized progesterone 400 mg Q12H, 100 mg P4 IM Q2 day, if needed on the day of FET P4 continued till 8 weeks of gestation Blastocyst Nil, Corpus luteum frequently observed during early pregnancy Kornilov et al., 2024 (Ref [ 10 ]) Retrospective 327 Leading follicle > 16 mm EM ≧ 7.0 mm Vaginal micronized progesterone 200 mg bid, decreasing to 200 mg qd on 9 days after FET. No P4 support upon pregnancy Euploid blastocyst Nil Eggersmann et al., 2024 (Ref [ 11 ]) Prospective 599 Leading follicle > 16 mm EM > 6.0 mm E2 > 180 pg/ml Oral dydrogesterone 10 mg tid Day2/3 cleavage stage; Day4/5 morula/blastocyst Serum P4 above a certain threshold on day of FET or late luteal phase Ref, reference; EM, endometrial thickness; E2, estradiol; P4, progesterone; LH, luteinizing hormone; FET, frozen embryo transfer Summary of published studies on natural proliferative phase frozen embryo transfer (NPP-FET) Leading follicle ≧ 12 mm EM ≧ 7.0 mm E2 > 80 pg/ml Endometrin 100 mg TID P4 continued till 12 weeks of gestation Mendes Godinho et al., 2024 (Ref [ 9 ]) Leading follicle ≧ 14 mm EM ≧ 7.0 mm P4 ≦ 1.5ng/ml irrespective of LH level Vaginal micronized progesterone 400 mg Q12H, 100 mg P4 IM Q2 day, if needed on the day of FET P4 continued till 8 weeks of gestation Nil, Corpus luteum frequently observed during early pregnancy Leading follicle > 16 mm EM ≧ 7.0 mm Vaginal micronized progesterone 200 mg bid, decreasing to 200 mg qd on 9 days after FET. No P4 support upon pregnancy Eggersmann et al., 2024 (Ref [ 11 ]) Leading follicle > 16 mm EM > 6.0 mm E2 > 180 pg/ml Day2/3 cleavage stage; Day4/5 morula/blastocyst Ref, reference; EM, endometrial thickness; E2, estradiol; P4, progesterone; LH, luteinizing hormone; FET, frozen embryo transfer Further supporting this notion, Alonso-Mayo et al. [ 40 ] reported comparable clinical outcomes in modified NC-FET cycles triggered with hCG when the dominant follicle measured 13–22 mm, suggesting that some degree of variation in E2 exposure may be acceptable. This further supports the concept that a strictly defined proliferative phase followed by natural ovulation may not be essential for effective luteal transition and implantation [ 40 ]. However, a retrospective analysis of true NC-FET cycles [ 36 ] found that when E2 levels exceeded 100 pg/mL for ≤ 4 days, including the LH surge day, clinical outcomes were negatively affected. In contrast, our study found that follicular phase duration was not significantly associated with pregnancy outcomes. These discrepancy highlights the need for further studies to clarify the optimal E2 priming duration and intensity, as well as the formulation, dosage, and timing of P4 initiation in NPP-FET. A deeper understanding of how exogenous P4 interacts with endogenous hormone will be essential for refining NPP-FET protocols and optimizing outcomes. However, it remains unclear whether the CL formed from smaller follicles (12–15 mm) in NPP-FET cycles is functionally equivalent to those in natural ovulation. It is not yet known whether these CLs produce sufficient levels of relaxin, vasoactive substances, and angiogenic factors to provide the same protective effects observed in natural cycles [ 5 ]. Further studies are needed to determine whether NPP-FET, similar to NC-FET, can reduce the risk of pregnancy complications associated with HRT-FET. Although true NC-FET requires precise ovulation timing, limiting scheduling flexibility, modifications have been proposed to enhance flexibility without compromising outcomes. These include triggering ovulation at follicular diameters between 13 and 22 mm [ 40 ], and employing GnRH antagonists combined with low-dose gonadotropins at follicular sizes of 14–17 mm to delay dominant follicle development, thereby facilitating flexible FET scheduling [ 41 ]. Such strategies, including NPP-FET, represent alternative protocols designed to achieve ovulation while enhancing clinical scheduling flexibility. A major strength of this study is the use of a robust dual-confirmation strategy, combining UDO and serum P4 assessment, which consistently verified spontaneous ovulation in all participants, along with the exclusive use of single euploid embryos and a relatively short study duration that minimized temporal bias. We also demonstrated that hormonal dynamics around UDO in NPP-FET closely resembled those observed in natural menstrual cycles (Table  1 ; Fig.  1 ). These findings enhance our understanding of NPP-FET physiology and support its broader application in clinical practice. Limitations include the retrospective nature of the study and the potential for selection bias. To mitigate these concerns, the analysis was restricted to first-time NPP-FET cycles involving a single euploid blastocyst transfers. Although FET timing was not randomized, minor scheduling adjustments due to daily procedural limits were implemented uniformly and without reference to patient clinical characteristics, lessening the risk of selection bias associated with transfer timing. Importantly, women with PCOS were excluded to minimize physiological heterogeneity, as PCOS is associated with impaired oocyte competence [ 42 ] and endometrial dysfunction [ 43 ], which could confound hormonal and clinical outcomes. Despite the small sample size, this study provides valuable preliminary data and practical insights to guide future research.

This study demonstrates that spontaneous ovulation can be consistently preserved in NPP-FET cycles when DYD is flexibly initiated based on specific follicular and hormonal criteria, confirming its physiologic alignment with natural cycles. Peri-ovulatory hormone profiles closely resembled those of natural cycles. Importantly, variation in the timing of P4 initiation showed no association with pregnancy outcomes, supporting the scheduling flexibility of NPP-FET while maintaining consistent clinical success across a range of P4 initiation timings. These findings establish NPP-FET as a clinically viable, physiologically grounded, and scheduling-friendly alternative to existing protocols. This approach might also be associated with improved obstetric outcomes compared to HRT-FET, although further validation is needed. Additional studies are warranted to confirm these findings, optimize estrogen priming and progesterone supplementation strategies, and compare both reproductive and obstetric outcomes across different FET regimens.