A prospective cohort study on progestin-primed ovarian stimulation vs. GnRH antagonist in IVF-ICSI cycles: effects on clinical pregnancy and neonatal outcomes.

This prospective cohort study, conducted from January 2020 to December 2023 at the Assisted Reproductive Medicine Department of the Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine, was approved by the institutional ethics committee. All participants provided informed consent. The study population consisted of infertile patients undergoing IVF or ICSI treatment, all of whom underwent frozen embryo transfer. The inclusion criteria were as follows: (1) age between 20 and 42 years [ 14 ], (2) body mass index (BMI) between 18 and 30 kg/m² [ 15 ], (3) patients undergoing frozen embryo transfer following IVF or ICSI, irrespective of infertility etiology. All patients meeting the inclusion criteria were enrolled in the study. The exclusion criteria were as follows: (1) patients with reproductive tract tumors, acute or chronic reproductive tract inflammation, or other conditions unsuitable for ovarian stimulation; (2) patients who are carriers of abnormal chromosomal traits in either partner; (3) patients who had received hormonal treatment within the past 3 months; (4) patients undergoing fresh embryo transfer (only frozen embryo transfer cycles were included in this study to eliminate potential confounding effects of transfer strategy on outcomes); (5) missing key baseline information. Withdrawal criteria included patients who voluntarily requested to withdraw from the study. This study was approved by the Ethics Committee of the Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine. For patients receiving the PPOS protocol, treatment commenced on day 3 of the menstrual cycle with 10 mg/day of oral medroxyprogesterone acetate (MPA), in combination with human menopausal gonadotropin (HMG) 150–225 IU/day. The dose of HMG was adjusted based on the ovarian response over the subsequent five days to meet the stimulation protocol standards. Additionally, other appropriate medications may be added based on the patient’s condition to optimize the treatment plan and enhance success rates. When the leading follicle reached the maturation standard (follicle diameter ≥ 18 mm), GnRH-a (Decapeptyl, Ferring) 0.1 mg and hCG 1000–2000 IU were administered. Oocyte retrieval was performed 36–38 h later under vaginal ultrasound guidance. For patients receiving the GnRH antagonist protocol, treatment began on day 3 of the menstrual cycle with intramuscular HMG 150 to 225 IU/day. After five days of treatment, the dose of HMG was adjusted based on follicle growth, and a GnRH antagonist was introduced when the leading follicle reached approximately 14 mm. The dose of the GnRH antagonist was determined based on the patient’s BMI and LH level: 0.25 mg/day for patients with a BMI > 25 kg/m² or LH > 2.0 ng/ml, and 0.125 mg/day for patients with lower BMI or LH < 2.0 IU/L. Once the leading follicle reached the maturation standard, GnRH-a (Decapeptyl, Ferring) 0.1 mg and hCG 1000–2000 IU were administered, followed by oocyte retrieval 36–38 h later under vaginal ultrasound. In this study, fertilization was performed using either conventional IVF or ICSI based on sperm quality and historical fertilization data. Fertilization assessment was conducted 18 h post-oocyte retrieval, and embryo cleavage was evaluated at 72 h. Embryo quality was assessed according to the Cummins criteria. Embryos with six or more cells and graded as I or II were deemed suitable and vitrified on day 3. The remaining embryos were cultured to day 5/6, and only those meeting blastocyst quality criteria at that time were cryopreserved. Frozen embryo transfer was performed using either a natural cycle or a letrozole-microstimulation cycle to prepare the endometrium for implantation. For patients with repeated implantation failure or a thin endometrium, hormone replacement therapy (HRT) was administered. When the endometrium was deemed adequately prepared, endometrial transformation was initiated, and embryo transfer was performed either on day 3 (for cleavage-stage embryos) or day 5/6 (for blastocysts). Pregnancy was confirmed by serum β-HCG at 14 days after FET, and clinical pregnancy was confirmed by transvaginal ultrasound 28 days after FET. Luteal phase support continued until the 10th week of pregnancy. The primary outcome measure of this study was the live birth rate, defined as the birth of a live baby after 28 weeks of gestation [ 16 ], including preterm births. Secondary outcomes were stratified into three categories: (1) Embryo development parameters, including oocyte retrieval rate (retrieved oocytes per punctured follicle), mature oocyte rate (mature oocytes per retrieved oocytes), normal fertilization rate (normally fertilized oocytes per retrieved oocytes), cleavage rate (cleaved embryos per fertilized oocytes), high-quality embryo rate (high-quality embryos per cleaved embryos), and effective blastocyst formation rate (usable blastocysts per embryos cultured to blastocyst stage) [ 17 ], (2) Pregnancy outcomes, including clinical pregnancy rate, ectopic pregnancy rate, miscarriage rate (encompassing all pregnancy losses at any gestational stage) [ 18 ], and (3) Neonatal outcomes, comprising low birth weight incidence (< 2500 g) [ 19 ] and congenital anomaly rates (diagnosed at birth) [ 20 , 21 ]. These indicators were used to evaluate the efficacy and safety of the PPOS protocol in assisted reproductive technology, comparing it with the GnRH antagonist protocol to assess whether PPOS could serve as an effective alternative to GnRH antagonists, which are already considered safe. Statistical analysis was performed using the R programming language (version 4.2.1; R Foundation for Statistical Computing, Austria). A P-value of < 0.05 was considered statistically significant. The normality of continuous variables was assessed using the Kolmogorov-Smirnov test, and data were presented as mean ± standard deviation or median, as appropriate. Comparisons of continuous variables were conducted using t-tests or Mann-Whitney U tests, while categorical variables were evaluated using chi-square tests. Univariate and multivariate logistic regression analyses were utilized to examine the associations between variables and cycle treatment protocols. Both crude and adjusted odds ratios (OR) with 95% confidence intervals (CI) were calculated. The propensity score (PS) for each individual was first calculated using a logistic regression model. The propensity score calculation incorporated demographic variables, clinical characteristics, and other covariates that could potentially influence treatment assignment in observational studies [ 22 , 23 ]. Following this, propensity score matching (PSM) was performed using a 1:1 nearest-neighbor matching method [ 24 ], with a caliper set at 0.2 times the standard deviation of the propensity score. This method ensured the balance of baseline characteristics between the PPOS and GnRH antagonist groups, while also maximizing sample retention [ 25 ]. In cases where missing data were encountered, these instances were excluded from the analysis due to the minimal occurrence of missing values (less than 1%). This approach was chosen to preserve the integrity and consistency of the dataset, ensuring that the analysis remains robust and accurately reflective of the study population’s characteristics. After matching, inverse probability of treatment weighting (IPTW) was applied for further adjustment. The formula for IPTW in the PPOS group was 1/PS, while in the GnRH antagonist group, the formula was 1/(1-PS) [ 26 ]. This weighting technique adjusted the propensity score distribution of both groups to resemble that of a randomized controlled trial, thereby minimizing potential biases and enhancing the reliability of causal inferences. The IPTW-adjusted data were analyzed to control for baseline differences, reinforcing the robustness of the study findings and improving the accuracy of causal inference. This methodological framework provides an effective tool for similar studies, ensuring the scientific rigor and credibility of the results.

This study included infertile women who underwent IVF/ICSI cycles using either the PPOS or GnRH antagonist protocols between January 2020 and December 2023 ( n  = 14,077). After excluding patients lost to follow-up, those outside the age range of 20–42 years, with a BMI outside the 18–30 kg/m² range, or who underwent fresh embryo transfer, a total of 12,483 patients undergoing frozen embryo transfer (FET) were enrolled (PPOS group: n  = 10,516; GnRH antagonist group: n  = 1,967). Baseline characteristics were balanced using 1:1 propensity score matching and inverse probability of treatment weighting, resulting in 1,950 patients in each group. The cohort selection process is illustrated in Fig.  1 . Fig. 1 Detailed flowchart of the cohort selection process Detailed flowchart of the cohort selection process In this prospective study, 1:1 propensity score matching (PSM) and inverse probability of treatment weighting (IPTW) were employed to reduce bias and assess the effects of PPOS and GnRH antagonist protocols in IVF/ICSI. Baseline characteristics of all patients before and after matching are shown in Table  1 . Before adjustment, significant differences were observed between the PPOS group ( n  = 10,516) and the GnRH antagonist group ( n  = 1,967), including infertility duration (2 [1–4] vs. 2 [1–4] years, P  < 0.001, SMD = 0.161), baseline FSH (6.28 [5.00–7.00] vs. 5.92 [5.00–7.00] IU/ml, P  = 0.060, SMD = 0.059), LH (3.71 [2.00–4.00] vs. 3.48 [2.00–4.00] IU/ml, P  = 0.120, SMD = 0.048), and endometrial thickness (8.39 ± 2.58 vs. 10.18 ± 2.67 mm, P  < 0.001, SMD = 0.658). Table 1 Demographic and cycle characteristics before and after mixed PSM and IPTW Before matching ( n = 12483 ) PPOS GnRH antagonist P -value SMD (n = 10516) (n = 1967) Age (years) 32.67 ± 4.50 32.55 ± 4.35 0.227 0.030 Body mass index (kg/m2) 22.45 (20.0–24.0.0.0) 22.4 1(20.0–24.0.0.0) 0.602 0.010 Duration of infertility (years) 2 (1–4) 2 (1–4) <0.001 0.161 Basal FSH (IU/ml) 6.28 (5.0–7.0.0.0) 5.92 (5.0–7.0.0.0) 0.060 0.059 Basal LH (IU/ml) 3.71 (2.0–4.0.0.0) 3.48 (2.0–4.0.0.0) 0.120 0.048 Basal E2 (pg/ml) 35.71 (25.0–41.0.0.0) 37.71 (27.0–44.0.0.0) 0.058 0.057 Pregnancy history, n (%) 0.063 0.045  0 5801 (55.16%) 1181 (60.04%)  1 2521 (23.98%) 358 (18.20%) ≥ 2 2194 (20.86%) 428 (21.76%) Endometrium thickness (mm) 8.39 ± 2.58 10.18 ± 2.67 <0.001 0.658 Type of infertility, n (%) 0.232 0.029  Primary infertility 5845 (55.58%) 1122 (57.04%)  Secondary infertility 4671 (44.42%) 845 (42.96%) Periodic treatment days 11 (10–12) 11 (10–12) <0.001 0.089 Fertilization method, n (%) 0.726 0.009 IVF 4722 (44.90%) 894 (45.45%) ICSI 3886 (36.95%) 718 (36.50%) IVF + ICSI 1908 (18.15%) 355 (18.05%) Basal antral follicle count 12 (8–17) 12 (8–17) 0.001 0.083 After matching ( n=3900 ) PPOS GnRH antagonist P -value SMD   (n = 1950)   (n = 1950) Age (years) 32.52 ± 4.39 32.54 ± 4.36 0.886 0.004 Body mass index (kg/m2) 22.03 (20.31–24.22.31.22) 21.79 (20.17–24.03.17.03) 0.805 0.007 Duration of infertility (years) 2 (1–5) 2 (1–4) 0.833 0.005 Basal FSH (IU/ml) 5.73 (4.87–6.83.87.83) 5.56 (4.73–6.66.73.66) 0.965 0.001 Basal LH (IU/ml) 3.14 (2.27–4.26.27.26) 3.11 (2.25–4.12.25.12) 0.799 0.006 Basal E2 (pg/ml) 32.00 (25.00–41.00.00.00) 33.00 (27.00–44.00.00.00) 0.551 0.019 Pregnancy history, n (%) 0.977 0.001  0 1111 (57.07%) 1171 (60.05%)  1 437 (22.41%) 381 (19.54%) ≥ 2 402 (20.62%) 398 (20.41%) Endometrium thickness (mm) 10.15 ± 3.00  10.08 ± 2.62 0.342 0.025 Type of infertility, n (%) 0.919 0.003  Primary infertility 1111 (56.97%) 1112 (57.03%)  Secondary infertility 839 (43.03%) 838 (42.97%) Periodic treatment days 11 (10–12) 11 (10–12) 0.697 0.01 Fertilization method, n (%) 0.854 0.005 IVF 898 (46.05%) 885 (45.39%) ICSI 695 (35.64%) 711 (36.46%) IVF + ICSI 357 (18.31%) 354 (18.15%) Basal antral follicle count 11 (7–16) 12 (8–17) 0.625 0.013 Data are given as mean ± SD (normal distribution) and median (25th percentile-75th percentile) (no normal distribution) for continuous variables and n (%) for dichotomous variables. All P values were assessed with the use of χ² or Fisher’s exact test (dichotomous variables) and t test or Mann-Whitney U test (continuous variables) After matching, results were adjusted using a combination of propensity score matching (PSM, 1:1 matching, caliper = 0.2) and inverse probability weighting (IPTW) PPOS Progestin-Primed Ovarian Stimulation, GnRH antagonist Gonadotropin-Releasing Hormone Antagonist, IVF , in vitro fertilization, ICSI Intracytoplasmic sperm injection Demographic and cycle characteristics before and after mixed PSM and IPTW Data are given as mean ± SD (normal distribution) and median (25th percentile-75th percentile) (no normal distribution) for continuous variables and n (%) for dichotomous variables. All P values were assessed with the use of χ² or Fisher’s exact test (dichotomous variables) and t test or Mann-Whitney U test (continuous variables) After matching, results were adjusted using a combination of propensity score matching (PSM, 1:1 matching, caliper = 0.2) and inverse probability weighting (IPTW) PPOS Progestin-Primed Ovarian Stimulation, GnRH antagonist Gonadotropin-Releasing Hormone Antagonist, IVF , in vitro fertilization, ICSI Intracytoplasmic sperm injection After adjustment, these differences were minimized. The groups were balanced for age (32.52 ± 4.39 vs. 32.54 ± 4.36, P  = 0.886, SMD = 0.004), BMI (22.03 [20.31–24.22] vs. 21.79 [20.17–24.03] kg/m², P  = 0.805, SMD = 0.007), and infertility duration (2 [1–5] vs. 2 [1–4] years, P  = 0.833, SMD = 0.005). Endometrial thickness (10.15 ± 3.00 vs. 10.08 ± 2.62 mm, P  = 0.342, SMD = 0.025) and other characteristics were also balanced after matching. Figure 2 (A) presents the kernel density plots of the propensity scores before and after matching for both groups. Prior to matching, significant standardized mean differences (SMD) were observed for variables such as endometrial thickness, infertility duration, and baseline FSH (e.g., SMD for endometrial thickness = 0.658, SMD for infertility duration = 0.161, SMD for baseline FSH = 0.059). Fig. 2 A Kernel density plots of propensity scores before and after matching.  B  Standardized mean difference (SMD) plot after matching A Kernel density plots of propensity scores before and after matching.  B  Standardized mean difference (SMD) plot after matching Following matching, the SMD for all variables decreased to nearly 0 (e.g., SMD for endometrial thickness = 0.025, SMD for infertility duration = 0.005, SMD for baseline FSH = 0.001), indicating that baseline characteristics were effectively balanced. Figure 2 (B) further demonstrates the effectiveness of matching by presenting an SMD plot. After matching, the propensity score density curves for both groups overlapped significantly, illustrating a reduction in selection bias. The oocyte and embryo development outcomes were compared between the PPOS and GnRH antagonist protocols, as presented in Table  2 . After matching, the outcomes for oocyte and embryo development were evaluated between the PPOS and GnRH antagonist protocols. The number of available embryos was significantly higher in the PPOS group (3.05 ± 1.72 vs. 2.93 ± 1.81, P  = 0.010), and the number of mature oocytes was also higher in the PPOS group (10.47 ± 6.59 vs. 10.14 ± 6.09, P  = 0.037). Similarly, the number of high-quality embryos was significantly greater in the PPOS group (4.60 ± 3.41) compared to the GnRH antagonist group (4.43 ± 3.04, P  = 0.037). In terms of developmental rates, the retrieved oocyte rate was statistically significantly higher in the GnRH antagonist group (0.81 ± 0.18) compared to the PPOS group (0.80 ± 0.18, P  < 0.001), although the absolute difference was minimal. The blastocyst formation rate was significantly higher in the PPOS group (0.32 ± 0.32) compared to the GnRH antagonist group (0.30 ± 0.31, P  = 0.020). Table 2 Comparison of oocyte and embryo development between PPOS and GnRH antagonist protocols Before matching After matching ( n = 12483 ) (n = 3900) PPOS GnRH antagonist P -value PPOS GnRH antagonist P -value   (n = 10516) (n = 1967)  (n = 1950)  (n = 1950) Absolute Numbers of Oocytes and Embryos  Number of mature oocytes 9.86 ± 6.50 10.14 ± 6.09 0.078 10.47 ± 6.59 10.14 ± 6.09 0.037  Number of fertilized oocytes 8.36 ± 5.56 8.68 ± 5.68 0.016 8.84 ± 5.67 8.68 ± 5.68 0.290  Number of available embryos 3.03 ± 1.75 2.94 ± 1.81 0.030 3.05 ± 1.72 2.93 ± 1.81 0.010  Number of high-quality embryos 4.44 ± 3.34 4.43 ± 3.04 0.949 4.60 ± 3.41 4.43 ± 3.04 0.037 Rates of Oocyte and Embryo Development  Retrieved oocytes rate 0.80 ± 0.18 0.82 ± 0.18 <0.001 0.80 ± 0.18 0.81 ± 0.18 <0.001  Mature oocytes rate 0.86 ± 0.15 0.85 ± 0.15 0.170 0.85 ± 0.15 0.85 ± 0.15 0.970  Fertilization rate 0.74 ± 0.18 0.74 ± 0.18 0.760 0.73 ± 0.18 0.74 ± 0.18 0.061  Top-quality embryo rate 0.58 ± 0.26 0.57 ± 0.27 0.005 0.57 ± 0.26 0.57 ± 0.27 0.577  Blastocyst formation rate 0.32 ± 0.32 0.30 ± 0.31 0.051 0.32 ± 0.32 0.30 ± 0.31 0.020 Data are presented as mean ± SD for continuous variables. All P values were assessed with the use of χ² test (dichotomous variables) and t test (continuous variables) After matching, results were adjusted using a combination of propensity score matching (PSM, 1:1 matching, caliper = 0.2) and inverse probability weighting (IPTW). Patients with Live Births, Refers to the number of patients who achieved a live birth PPOS Progestin-Primed Ovarian Stimulation, GnRH antagonist Gonadotropin-Releasing Hormone Antagonist, IVF in vitro fertilization, ICSI Intracytoplasmic sperm injection Comparison of oocyte and embryo development between PPOS and GnRH antagonist protocols Data are presented as mean ± SD for continuous variables. All P values were assessed with the use of χ² test (dichotomous variables) and t test (continuous variables) After matching, results were adjusted using a combination of propensity score matching (PSM, 1:1 matching, caliper = 0.2) and inverse probability weighting (IPTW). Patients with Live Births, Refers to the number of patients who achieved a live birth PPOS Progestin-Primed Ovarian Stimulation, GnRH antagonist Gonadotropin-Releasing Hormone Antagonist, IVF in vitro fertilization, ICSI Intracytoplasmic sperm injection Table  3 presents the pregnancy and neonatal outcomes in the PPOS and GnRH antagonist protocols, both before and after matching. After matching, the clinical pregnancy rate was comparable between the PPOS and GnRH antagonist groups (46.26% vs. 46.21%, P  = 0.979), as were the rates of live birth (36.26% vs. 35.49%, P  = 0.550), miscarriage (10.05% vs. 10.72%, P  = 0.375), and ectopic pregnancy (0.82% vs. 0.82%, P  = 0.880), indicating similar pregnancy outcomes. In singleton births, the incidence of low birth weight (< 2500 g) was 4.86% in the PPOS group and 3.16% in the GnRH antagonist group ( P  = 0.056). The rate of congenital anomalies was 2.35% vs. 1.58% ( P  = 0.250), and sex distribution also showed no significant difference ( P  = 0.417). Regarding twin births, the proportion of twins was 30.26% in the PPOS group and 26.26% in the GnRH antagonist group. No significant differences were found in low birth weight (47.49% vs. 66.01%, P  = 0.949), congenital anomalies (2.70% vs. 4.93%, P  = 0.182), or sex distribution ( P  = 0.345). Table 3 Pregnancy and neonatal outcomes in PPOS vs. GnRH antagonist protocols Before matching After matching　 PPOS GnRH antagonist P -value PPOS GnRH antagonist P -value   (n = 10516) (n = 1967)  (n = 1950)  (n = 1950) Pregnancy Outcomes in Patients  Clinical pregnancy 4752 (45.19%) 909 (46.21%) 0.402 902 (46.26%) 901 (46.21%) 0.979  Ectopic pregnancy 101 (0.96%) 16 (0.83%) 0.535 16 (0.82%) 16 (0.82%) 0.880  Miscarriage 1068 (10.16%) 212 (10.78%) 0.404 196 (10.05%) 209 (10.72%) 0.375  Live birth 3684 (35.03%) 697 (35.43%) 0.732 707 (36.26%) 692 (35.49%) 0.550 PPOS GnRH antagonist P -value PPOS GnRH antagonist  P -value  (n = 4356)  (n = 818) (n = 856) (n = 773) Singleton Number of neonates 3184 (73.09%) 576 (70.42%) 597 (69.74%) 570 (73.74%) Sex distribution of neonates 0.425 0.417 Male 1592 (50.00%) 307 (51.42%) 307 (51.42%) 304 (53.33%) Female 1592 (50.00%) 290 (48.58%) 290 (48.58%) 266 (46.67%) Low birth weight (<2500 g), n (%) 168 (5.28%) 29 (4.86%) 0.027 29 (4.86%) 18 (3.16%) 0.056 Congenital anomalies, n (%) 69 (2.17%) 14 (2.35%) 0.390 14 (2.35%) 9 (1.58%) 0.250 Twins　 Number of neonates 1172 (26.91%) 242 (29.58%) 259 (30.26%) 203 (26.26%) Sex distribution of neonates 0.515 0.345 Male 624 (53.24%) 135 (55.79%) 119 (45.95%) 103 (50.74%) Female 548 (46.76%) 107 (44.21%) 140 (54.05%) 100 (49.26%) Low birth weight (<2500 g), n (%) 624 (53.24%) 135 (55.79%) 0.515 123 (47.49%) 134 (66.01%) 0.949 Congenital anomalies, n (%) 44 (3.75%) 8 (3.31%) 0.881 7 (2.70%) 10 (4.93%) 0.182 Data are given as n (%) for dichotomous variables. All p values were assessed with the use of χ² or Fisher’s exact test After matching, results were adjusted using a combination of propensity score matching (PSM, 1:1 matching, caliper = 0.2) and inverse probability weighting (IPTW) PPOS Progestin-Primed Ovarian Stimulation, GnRH antagonist  Gonadotropin-Releasing Hormone Antagonist Pregnancy and neonatal outcomes in PPOS vs. GnRH antagonist protocols Data are given as n (%) for dichotomous variables. All p values were assessed with the use of χ² or Fisher’s exact test After matching, results were adjusted using a combination of propensity score matching (PSM, 1:1 matching, caliper = 0.2) and inverse probability weighting (IPTW) PPOS Progestin-Primed Ovarian Stimulation, GnRH antagonist  Gonadotropin-Releasing Hormone Antagonist To more accurately evaluate the effects of ovarian stimulation protocols, this study restricted the analysis to singleton live births, excluding twin pregnancies due to their inherently higher risks of low birth weight, preterm birth, and other complications. The crude and adjusted odds ratios for pregnancy and neonatal outcomes (singleton births only) are shown in Table  4 . After matching, pregnancy outcomes were comparable between the PPOS and GnRH antagonist protocols. There were no significant differences in pregnancy rates (adjusted OR = 0.94, P  = 0.279) or miscarriage rates (adjusted OR = 1.08, P  = 0.338), indicating similar efficacy between the two stimulation strategies. The pregnancy rate was 46.26% in the PPOS group and 46.21% in the GnRH antagonist group, while the miscarriage rates were 10.05% and 10.72%, respectively. Table 4 Crude and adjusted odds ratios (OR) for pregnancy and neonatal outcomes (Singletons Only) Before matching Crude OR P-value Adjusted OR P-value (95% CI) (95% CI) Pregnancy 1.04 (0.95–1.15.95.15) 0.402 0.98 (0.89–1.08.89.08) 0.677 Miscarriage 1.07 (0.91–1.25.91.25) 0.404 1.08 (0.92–1.27.92.27) 0.338 Live births 1.02 (0.92–1.13.92.13) 0.732 0.94 (0.85–1.05.85.05) 0.279 Male 1.08 (0.90–1.29.90.29) 0.400 1.10 (0.92–1.32.92.32) 0.303 Low birth weight (<2500 g) 1.78 (1.11–3.01.11.01) 0.023 1.67 (1.04–2.86.04.86) 0.047 Congenital anomalies 0.70 (0.32–1.33.32.33) 0.312 0.70 (0.32–1.37.32.37) 0.336 After matching Crude OR P-value Adjusted OR P-value  (95% CI) (95% CI) Pregnancy 0.99 (0.90–1.11.90.11) 0.979 0.99 (0.90–1.11.90.11) 0.963 Miscarriage 1.08 (0.91–1.27.91.27) 0.367 1.08 (0.91–1.27.91.27) 0.373 Live births 0.97 (0.87–1.08.87.08) 0.551 0.97 (0.87–1.08.87.08) 0.586 Male 1.08 (0.90–1.30.90.30) 0.417 1.08 (0.90–1.30.90.30) 0.418 Low birth weight (<2500 g) 1.64 (0.98–2.72.98.72) 0.058 1.62 (0.97–2.71.97.71) 0.063 Congenital anomalies 0.65 (0.31–1.36.31.36) 0.254 0.67 (0.32–1.39.32.39) 0.278 RRs and 95%Cl were based on the univariate analysis. Adjusted RRs and 95% Cl werebased on the multiple logistic regression model after adjusting for Female Age, Body mass index, Endometrial thickness, Number of viable embryos After matching, results were adjusted using a combination of propensity score matching (PSM, 1:1 matching, caliper = 0.2) and inverse probability weighting (IPTW) Crude and adjusted odds ratios (OR) for pregnancy and neonatal outcomes (Singletons Only) RRs and 95%Cl were based on the univariate analysis. Adjusted RRs and 95% Cl werebased on the multiple logistic regression model after adjusting for Female Age, Body mass index, Endometrial thickness, Number of viable embryos After matching, results were adjusted using a combination of propensity score matching (PSM, 1:1 matching, caliper = 0.2) and inverse probability weighting (IPTW) In neonatal outcomes, the proportion of male infants was similar in both groups, with 51.23% male infants in the PPOS group and 50.97% male infants in the GnRH antagonist group ( P  = 0.417). The risk of low birth weight (< 2500 g) was significantly higher in the PPOS group, both before (OR = 1.78, P  = 0.023) and after matching (adjusted OR = 1.67, P  = 0.047). The incidence of low birth weight was 4.86% in the PPOS group compared to 3.16% in the GnRH antagonist group. The rate of congenital anomalies was 2.35% in the PPOS group compared to 1.58% in the GnRH antagonist group ( P  = 0.250).

The PPOS protocol is comparable to the GnRH antagonist protocol in terms of live birth rates, pregnancy outcomes, and neonatal health in IVF/ICSI treatments. The lack of significant differences between the two protocols suggests that PPOS could be considered a safe and effective alternative for ovarian stimulation in assisted reproduction, particularly for patients seeking a more patient-friendly and cost-effective approach. Future studies with larger sample sizes and extended follow-up are needed to confirm these findings and explore the long-term impacts on offspring health.

This study aimed to compare the safety and efficacy of the PPOS protocol with the GnRH antagonist protocol in IVF/ICSI treatments using frozen embryo transfer. The results indicate that the PPOS protocol provides comparable outcomes to the GnRH antagonist protocol in terms of clinical pregnancy rates, live birth rates, and neonatal health. Advanced statistical techniques, such as propensity score matching and inverse probability treatment weighting, were employed to ensure comparability between the two groups, minimizing potential biases from baseline differences. After matching, the baseline characteristics—including age, BMI, and infertility duration—were well balanced, thereby enhancing the reliability of the results. Regarding embryo development, the PPOS group demonstrated a significantly higher number of high-quality embryos compared to the GnRH antagonist group (4.60 ± 3.41 vs. 4.43 ± 3.04, P = 0.037). The number of available embryos (3.05 ± 1.72 vs. 2.93 ± 1.81, P = 0.010) and mature oocytes (10.47 ± 6.59 vs. 10.14 ± 6.09, P = 0.037) were also significantly higher in the PPOS group. Both protocols showed similar results in the number of retrieved oocytes, fertilization rates, and top-quality embryos [ 27 ]. The consistency of fertilization rates and top-quality embryo rates further supports the view that the PPOS protocol does not introduce new embryological safety concerns. In terms of clinical pregnancy outcomes, no significant differences were observed between the two protocols. Specifically, the clinical pregnancy rate in the PPOS group was 46.26%, compared to 46.21% in the GnRH antagonist group ( P = 0.979), indicating similar efficacy in achieving clinical pregnancies with both protocols. Tohma et al. reported similar clinical pregnancy rates between the PPOS and GnRH antagonist protocols in patients with endometriosis [ 28 ]. A pilot study by Le et al. also demonstrated similar sustained pregnancy rates between the two protocols, with rates of 46.7% and 53.3%, respectively [ 29 ]. Miscarriage rates were also similar, with the PPOS group showing a rate of 10.05%, while the GnRH antagonist group had a miscarriage rate of 10.72% ( P = 0.375). These findings suggest that the PPOS protocol does not increase the risk of pregnancy loss compared to the GnRH antagonist protocol, supporting its safety in IVF/ICSI treatments. The equivalence in pregnancy outcomes indicates that exogenous progesterone, used in the PPOS protocol to suppress premature LH surges, does not significantly compromise oocyte quality, fertilization, or endometrial receptivity. This finding aligns with the growing consensus that PPOS is not inferior to GnRH antagonist protocols in routine IVF/ICSI practice. Neonatal outcomes further corroborated the safety of the PPOS protocol. After matching, the difference in low birth weight rates between the groups decreased, with the PPOS group showing a rate of 4.86% and the GnRH antagonist group at 3.16% ( P = 0.056). The incidence of congenital anomalies in neonates was similar across both groups. In the PPOS group, 2.35% of neonates exhibited congenital anomalies, compared to 1.58% in the GnRH antagonist group ( P = 0.250). These results align with previous studies suggesting that PPOS does not increase the risk of congenital anomalies compared to the GnRH antagonist protocol [ 30 , 31 ]. In a singleton-only analysis after matching, no significant differences were found in clinical pregnancy, live birth, or miscarriage rates between the two protocols. The incidence of low birth weight (< 2500 g) was not significantly different (adjusted OR = 1.62, 95% CI: 0.97–2.71, P = 0.063). Taken together, these findings provide further evidence that the PPOS protocol does not elevate the risk of adverse neonatal outcomes compared to the GnRH antagonist protocol, supporting the safety of PPOS for both maternal and neonatal health in IVF/ICSI with frozen embryo transfer. In conclusion, this study demonstrates that the PPOS protocol provides a comparable profile of embryological, pregnancy, and neonatal outcomes to the GnRH antagonist protocol, with no apparent increase in the risk of adverse events. These findings support the use of PPOS as a safe and effective alternative for ovarian stimulation in IVF/ICSI cycles with frozen embryo transfer. However, since this study was conducted at a single center with limited neonatal follow-up, further research with multicenter cohorts and long-term offspring surveillance is necessary to fully assess the long-term safety of PPOS. This study offers several strengths. By employing a prospective cohort design, it effectively minimized biases such as recall and selection bias through real-time tracking of participants and data collection. This approach ensured robust causal inference and increased the clinical relevance of the results, supported by the large sample size, which enhances generalizability to various clinical contexts. The use of propensity score matching (PSM) and inverse probability treatment weighting (IPTW) ensured balance in baseline characteristics, reducing confounding and selection bias. While these methods do not fully replicate randomized controlled trials (RCTs), they bring observational studies closer to the rigor of RCTs, strengthening the reliability of the findings. The study also stands out for its comprehensive assessment, including embryological parameters, offering novel insights into how ovarian stimulation protocols influence embryo development. The incorporation of neonatal outcomes provides a more complete evaluation of maternal and neonatal health, a rare feature in similar studies, thus adding significant clinical value. By combining PSM and IPTW, the study reduced biases and ensured the robustness of data analysis, making it a valuable reference for similar research. In assisted reproduction, controlling for confounders and enhancing result precision is critical. Despite its strengths, the study has some limitations. The single-center design may restrict the external validity of the findings. Conducted at a high-volume assisted reproductive center in Shanghai, the results may not fully represent other centers, as variations in ovarian response, laboratory practices, and treatment protocols could affect the outcomes. Although advanced statistical methods were used to balance baseline characteristics, residual confounding factors, such as smoking, psychological status, and socioeconomic variables, may still influence pregnancy and neonatal outcomes. Additionally, selection bias could have impacted the comparison between protocols since the GnRH antagonist group had the option of both fresh and frozen embryo transfers. While neonatal outcomes were assessed, the evaluation was limited to short-term health data. Without long-term follow-up, especially regarding childhood development and neurodevelopment, the study could not determine the prolonged effects of different protocols. Future research with extended follow-up is necessary to assess the long-term impact on children’s health.

In the field of human assisted reproductive technology (ART), the prevention of premature ovulation remains a critical challenge for achieving optimal reproductive outcomes. For several years, GnRH analogs, including long, short, and antagonist protocols, have been extensively utilized to prevent premature LH surges and ovulation [ 1 ]. In 2014, Professor Yanping Kuang from the Ninth People’s Hospital, Shanghai Jiao Tong University, introduced the progesterone-primed ovarian stimulation (PPOS) protocol. This protocol effectively mitigates premature LH surges by combining progesterone with ovarian stimulation, thereby inhibiting the positive feedback loop of GnRH neurons, which prevents premature ovulation [ 2 – 5 ]. The PPOS protocol is characterized by its straightforward operational process and reduced risk profile, positioning it as a key area of research in recent years. PPOS is associated with lower medication costs, averaging 898.3 euros per cycle, in contrast to 1196.4 euros for the GnRH antagonist protocol [ 6 ]. Previous studies have demonstrated that PPOS effectively suppresses early LH surges [ 7 ], reduce the incidence of moderate and severe ovarian hyperstimulation syndrome (OHSS) [ 8 ], and yields favorable therapeutic outcomes in patients with polycystic ovary syndrome [ 9 ] and high body mass index [ 10 ]. In comparison to the GnRH protocol, PPOS yields higher implantation rates (43.4% vs. 31.9%) and clinical pregnancy rates (61.8% vs. 47.4%) [ 11 ]. Although previous studies comparing the PPOS and GnRH antagonist protocols in assisted reproductive technology have employed traditional regression models to control for confounders (e.g., age, BMI, ovarian reserve) [ 12 ], residual selection bias remains in non-randomized designs. Additionally, an overreliance on short-term outcomes (e.g., clinical pregnancy rate) neglects critical endpoints, such as neonatal health [ 13 ]. Therefore, this study aimed to accurately balance baseline characteristics between the PPOS and GnRH antagonist protocol groups, address the limitations of traditional embryological parameters, and prospectively track neonatal outcomes. This approach provides new insights into the safety of using PPOS as a potential alternative to GnRH antagonist protocols in assisted reproductive technology.