Progestin-primed ovarian stimulation (PPOS) in preimplantation genetic testing for aneuploidy: a retrospective study and meta-analysis.

This retrospective study included a total of 962 ovarian stimulation cycles, comprising 413 cases of the PPOS protocol, 327 cases of the GnRH-ant protocol, and 222 cases of the GnRH-a protocol. Significant differences were observed among the three groups in terms of couples’ age, antral follicle count (AFC), serum levels of anti-Müllerian hormone (AMH) and basal Follicle-stimulating hormone (FSH). Patients with younger age and better ovarian reserve were more likely to receive the GnRH-a protocol. The GnRH-a group had the longest stimulation duration and the highest Gn dosage requirements. Significant differences were found in hormone levels on trigger day, number of retrieved oocytes, mature oocytes, 2PN fertilized oocytes, euploid blastocysts, 2PN rate and euploid rate per biopsied blastocyst among the three groups (Table  1 ). Table 1 The baseline characteristics and ovarian stimulation outcomes PPOS GnRH antagonist GnRH agonist p value Stimulation cycles ( n ) 413 327 222 Female age (years) 40 (38, 42) 41 (39, 43) 39 (35, 41) < 0.001 a,b,c Male age (years) 42 (37, 45) 41 (37, 45) 39 (35, 42) < 0.001 b,c Female BMI (kg/m 2 ) 22.6 (20.8, 24.9) 22.9 (20.8, 25.1) 22.2 (20.7, 24) 0.151 AFC 7 (4, 10) 5 (3, 9) 11 (8, 14) < 0.001 a,b,c AMH 1.68 (0.87, 2.95) 1.12 (0.6, 2.26) 2.64 (1.8, 3.87) < 0.001 a,b,c Basal E2 (pg/mL) 32.53 (23.22, 46.62) 32.53 (23.01, 45.28) 34.14 (25.65, 52.15) 0.155 Basal FSH (IU/L) 6.14 (5.19, 7.77) 6.59 (5.22, 9.04) 5.25 (4.52, 6.22) < 0.001 b,c E2 on trigger day (pg/ml) 1680.78 (873.44, 2751.5) 928.05 (418.41, 1832.86) 2219.19 (1569.12, 3266.92) < 0.001 a,b,c LH on trigger day (IU/L) 2.63 (1.79, 3.82) 2.23 (1.32, 3.48) 0.83 (0.63, 1.14) < 0.001 a,b,c Stimulation duration (days) 8 (7, 9) 8 (7, 9) 11 (10, 12) < 0.001 b,c Gn dosage (IU) 2250 (1950, 2625) 2250 (1575, 2550) 2968.88 (2550, 3300) < 0.001 b,c No. of oocytes retrieved ( n ) 7 (4, 12) 5 (2, 10) 13 (9, 18) < 0.001 a,b,c No. of MII ( n ) 6 (3, 10) 4 (1, 8) 11 (8, 15)  < 0.001 a,b,c MII rate (%) 84.09 (2982/3546) 82.72 (1814/2193) 83.84 (2687/3205) 0.370 No. of fertilized oocytes (2PN) 5 (2, 8) 3 (0, 6) 8 (6, 12) < 0.001 a,b,c 2PN rate (%) * 76.22 (2276/2982) 73.04 (1325/1814) 75.44 (2027/2687) 0.036 b No. of blastocyst biopsied 2 (1, 4) 1 (0, 3) 4 (2, 6) < 0.001 a,b,c No. of euploid blastocysts ( n ) 0 (0, 1) 0 (0, 1) 1 (0, 3) < 0.001 a,b,c Euploid rate per biopsied blastocyst (%)* 29.60 (346/1169) 32.54 (206/633) 36.72 (361/983) 0.002 b a PPOS vs. GnRH antagonist, p  < 0.05 b PPOS vs. GnRH agonist, p  < 0.05 c GnRH antagonist vs. GnRH agonist, p  < 0.05 * p value is considered to be significant when < 0.016 [Bonferroni correction (0.05/3)] The baseline characteristics and ovarian stimulation outcomes a PPOS vs. GnRH antagonist, p  < 0.05 b PPOS vs. GnRH agonist, p  < 0.05 c GnRH antagonist vs. GnRH agonist, p  < 0.05 * p value is considered to be significant when < 0.016 [Bonferroni correction (0.05/3)] To exclude confounding factors, we further performed multiple linear regression to evaluate the effects of ovarian stimulation protocols on blastocyst euploid rate and number of euploid blastocysts. The results indicated that neither the MII rate, 2PN rate, euploid rate per biopsied blastocyst nor the average number of euploid blastocysts was associated with the ovarian stimulation protocol used. However, female age and AMH level significantly influenced the rate of euploid blastocysts. In addition, AFC and E2 level on trigger day was correlated with the number of euploid blastocysts (Table  2 and Supplementary Table 3). Table 2 Multivariate liner regression analysis for number of euploid blastocysts and euploid rate per biopsied blastocyst Independent variable No. of euploid blastocysts Euploid rate per biopsied blastocyst β t p value 95% CI for Exp(B) β t p value 95% CI for Exp(B) Lower Upper Lower Upper PPOS GnRH antagonist 0.089 0.877 0.381 − 0.111 0.289 0.015 0.732 0.464 − 0.026 0.056 GnRH agonist 0.214 1.693 0.091 − 0.034 0.462 − 0.013 − 0.487 0.626 − 0.064 0.038 Female age − 0.116 − 8.439 0.000 − 0.143 − 0.089 − 0.024 − 8.633 0.000 − 0.030 − 0.019 Male age − 0.023 − 2.443 0.015 − 0.041 − 0.005 − 0.003 − 1.442 0.150 − 0.007 0.001 AFC 0.030 2.436 0.015 0.006 0.055 0.001 0.388 0.698 − 0.004 0.006 AMH 0.100 2.907 0.004 0.033 0.168 0.019 2.649 0.008 0.005 0.033 Basal FSH 0.003 0.237 0.812 − 0.020 0.026 0.000 − 0.189 0.850 − 0.005 0.004 E2 on trigger day 0.000 4.315 0.000 0.000 0.000 0.000 1.418 0.157 0.000 0.000 LH on trigger day − 0.002 − 0.157 0.875 − 0.025 0.021 0.003 1.351 0.177 − 0.001 0.008 BMI 0.004 0.223 0.824 − 0.028 0.035 0.000 − 0.060 0.952 − 0.007 0.006 Stimulation duration − 0.016 − 0.456 0.648 − 0.086 0.054 − 0.004 − 0.506 0.613 − 0.018 0.011 Gn dosage 0.000 − 0.077 0.938 0.000 0.000 0.000 1.915 0.056 0.000 0.000 Multivariate liner regression analysis for number of euploid blastocysts and euploid rate per biopsied blastocyst A total of 382 FET cycles had been completed by December 2023. There were no significant differences in the rates of biochemical pregnancy, clinical pregnancy, premature delivery, live birth, or miscarriage among the three groups. There was one ectopic pregnancy and one twin pregnancy in the PPOS group, as well as in the GnRH-a group. The twin pregnancy in the PPOS group was naturally reduced in the early pregnancy and a single live birth was obtained. The twin pregnancy in the GnRH-a group obtained twin live births. There was one case of induced labor at 28 weeks of gestation due to intrauterine growth retardation in the GnRH-a group (Table  3 ). Table 3 Clinical outcomes after euploid blastocyst transfer PPOS GnRH antagonist GnRH agonist p value Cycles ( n ) 156 91 135 Endometrial preparation ( n ) 0.437 Natural cycles 30 21 39 Ovulation induction cycles 57 32 43 Hormone replacement cycles 69 38 53 Endometrial thickness (mm) 9.20 (8.20, 11.00) 9.40 (8.50, 10.80) 9.00 (8.20, 11.00) 0.981 Biochemical pregnancy rate (%) 70.51 (110/156) 70.33 (64/91) 66.67 (90/135) 0.746 Clinical pregnancy rate (%) 65.38 (102/156) 65.93 (60/91) 63.70 (86/135) 0.931 Premature birth rate (%) 7.84 (8/102) 11.67 (7/60) 12.8 (11/86) 0.513 Live birth rate (%) 54.49 (85/156) 57.14 (52/91) 49.63 (67/135) 0.507 Miscarriage rate (%) 15.69 (16/102) 13.33 (8/60) 19.77 (17/86) 0.563 Clinical outcomes after euploid blastocyst transfer We identified 2111 records from databases and clinical trial registration websites, and finally included nine studies through screening, of which seven were retrospective cohort studies [ 12 – 18 ], 1 was a prospective matched case–control study [ 19 ], and 1 was a prospective self-control trial [ 20 ]. We did not find any published RCTs on this topic. All studies were published in English and involved comparisons between the PPOS protocol and the antagonist protocol, with 3 studies also comparing these protocols to others. Trophoblast cell biopsy and next-generation sequencing technology were performed in all studies. The screening process and characteristics of the included studies are presented in Fig.  1 and Supplementary Table 2. Fig. 1 Flow chart of study selection Flow chart of study selection In the ROBINS-I assessments, confounding bias was rated as moderate to serious risk in most studies due to their retrospective nature. In contrast, the intervention and post-intervention biases were rated as low to moderate risk due to the relative well-developed medical records and follow-up monitoring system in most reproductive centers (Supplementary Fig. 1). Results with balanced baseline characteristics were pooled, all of which were derived from comparisons between PPOS and antagonist protocols. A narrative analysis was performed on data with imbalanced baseline characteristics (such as significant differences in age and ovarian function). The nine included studies reported the main outcomes in terms of the number of euploid blastocysts and/or blastocyst euploid rate. A meta-analysis of the data with balanced baseline characteristics showed that there were no significant differences in the number of euploid blastocysts (MD = − 0.19, 95% CI: − 0.48–0.11, I 2  = 88%, P  = 0.22, 6 studies, 7346 cycles), the euploid rate per biopsied blastocyst (OR = 0.91, 95% CI: 0.75–1.12, I 2  = 34%, P  = 0.39, 5 studies, 1123 cycles), the euploid rate per MII (OR = 0.95, 95% CI: 0.79–1.15, I 2  = 45%, P  = 0.62, 4 studies, 1035 cycles) and the euploid rate per oocyte retrieved (OR = 0.98, 95% CI: 0.83–1.15, I 2  = 27%, P  = 0.80, 4 studies, 1035 cycles) between PPOS and GnRH-ant protocols (Fig.  2 ). Fig. 2 Forest plots of primary outcomes Forest plots of primary outcomes Three studies with significant differences in baseline characteristics were subjected to regression analysis to evaluate the relationship between ovarian stimulation protocols and primary outcomes. Huang’s results showed that after adjusting for confounding factors, compared with the PPOS protocol, the GnRH-ant protocol was positively correlated with the average number of euploid blastocysts and the euploid rate of blastocysts per biopsy ( β  = 0.652, P  = 0.001), while the GnRH-a protocol had no such relationship [( β  = 0.330, P  = 0.174] [ 13 ]. The other two studies reported different conclusions, with one study showing no significant association between ovarian stimulation protocols ( β  = 0.027, 95% CI: − 18.883–22.839, P  = 0.852) and blastocyst euploidy rate [ 16 ] and the other suggesting that GnRH-ant protocol was associated with a lower embryo euploidy rate per biopsy ( β  = − 0.088, 95% CI: − 0.157 to − 0.02, P  = 0.011), while the euploidy rate per oocyte retrieved was similar across all protocols [ 17 ]. Four retrospective studies reported the outcome of embryo transfer. The meta-analysis showed that the biochemical pregnancy rate (OR = 1.07, 95% CI: 0.91–1.25, I 2  = 0%, P  = 0.42, 4 studies, 6097 cycles), clinical pregnancy rate (OR = 1.01, 95% CI: 0.87–1.18, I 2  = 0%, P  = 0.89, 4 studies, 6097 cycles), ongoing pregnancy rate (OR = 1.12, 95% CI: 0.95–1.31, I 2  = 0%, P  = 0.16, 2 studies, 5849 cycles) and live birth rate were similar (OR = 1.06, 95% CI: 0.70–1.60, I 2  = 0%, P  = 0.80, 3 studies) between PPOS and GnRH-ant protocols, while the miscarriage rate showed a decreasing trend (OR = 0.67, 95% CI: 0.48–0.93, I 2  = 0%, P  = 0.02, 4 studies) (Fig.  3 ). Fig. 3 Forest plots of clinical outcomes. △ Miscarriage rate here was defined as clinical pregnancy loss divided by the number of clinical pregnancies Forest plots of clinical outcomes. △ Miscarriage rate here was defined as clinical pregnancy loss divided by the number of clinical pregnancies One study with unbalanced baseline characteristics compared the PPOS and GnRH-a protocols. The results showed that there was no significant difference in biochemical pregnancy, clinical pregnancy, ongoing pregnancy rate, or live birth rate per ET between the two groups [ 16 ]. No additional data were available on the clinical outcomes of PPOS protocol compared with other ovarian stimulation protocols. The results of meta-analysis showed that the ovarian stimulation duration (MD = 0.06, 95% CI: −0.46–0.58, I 2  = 94%, P  = 0.83, 7 studies, 7564 cycles), Gn dosage (MD = 28.85, 95% CI: −263.94–321.64, I 2  = 97%, P  = 0.85, 6 studies, 7297 cycles) and cycle cancelation rate of PPOS protocol (OR = 1.06, 95% CI: 0.85–1.32, I 2  = 0%, P  = 0.60, 3 studies, 1495 cycles) were similar to those of GnRH-ant protocol. There was no significant difference in the number of retrieved oocytes (MD = 0.28, 95% CI: − 1.02–1.58, I 2  = 85%, P  = 0.67, 7 studies, 7564 cycles) and MII (MD = 0.24, 95% CI: − 0.08–0.56, I 2  = 0%, P  = 0.14, 6 studies, 7436 cycles) between PPOS protocol and GnRH-ant protocol (Fig.  4 ). Fig. 4 Forest plots of stimulation parameters and embryo characteristics Forest plots of stimulation parameters and embryo characteristics Among the three studies with unbalanced baseline characteristics, only Wang’s study provided the results of further regression analysis, suggesting that after adjusting for confounding factors, ovarian stimulation protocol was not related to MII rate or fertilization rate. No significant difference in cycle cancelation rate was found between PPOS protocol and GnRH-a protocol ( β  = 0.287, P  = 0.132, adjust OR = 1.333), while the cycle cancelation rate of GnRH-ant protocol, mild stimulation protocol and other protocols was significantly increased [ 17 ]. The sensitivity analysis for the number of euploid blastocysts, Gn dosage, ovarian stimulation durations, and the number of retrieved oocytes were robust. Subgroup analysis showed that the type of progesterone had no significant effects on the number of euploid blastocysts or stimulation duration. However, the Gn dosage and the number of oocytes in the dydrogesterone group were significantly lower than those in the GnRH-ant group (Supplementary Figs. 2, 3).

We reviewed patients who underwent PGT-A at our reproductive center between August 2020 and December 2023. Indications of PGT-A included advanced maternal age, recurrent spontaneous abortion, and repeated implantation failure. Only patients receiving ovarian stimulation with PPOS, GnRH antagonist (GnRH-ant), and GnRH agonist (GnRH-a) protocols were included in the study. We excluded patients with the following criteria: (1) monogenic diseases or chromosomal abnormalities; (2) oocyte donation cycles; and (3) cycles using thawed embryos for genetic testing. The retrospective study was approved by the Ethics Committee of Guangzhou Women and Children’s Hospital Medical Center Liuzhou Hospital (2024-172). In the PPOS protocol, gonadotropin (Gn) and medroxyprogesterone acetate (MPA, Zhejiang Xianju Pharmaceutical Co., Ltd., China) were administered from day 2–3 of the menstrual cycle. MPA was taken orally at a dose of 8–10 mg once a day until the trigger day. In the GnRH-ant protocol, Gn was injected from the second or third day of the menstrual cycle, and cetrorelix (Baxter Oncology GmbH, Germany) or ganirelix (Chia-Tai Tianqing Pharmaceutical Co., Chian) was added with daily doses of 0.25 mg from day 5–6 of ovarian stimulation or when the maximum follicle diameter reached 14 mm until the trigger day. In the GnRH-a protocol, long-acting GnRH-a (Triptorelin Acetate, Ipsen Pharma Biotech, France or Leuprorelin Acetate, Lizhu Pharmaceutical Trading Co., China) was injected subcutaneously in the mid-luteal phase in the menstrual cycle. After 14–20 days, an ultrasound and serum hormone levels were examined, and Gn was given after meeting the down-regulation standard. In the above three protocols, the initial dose of Gn was determined according to female age, ovarian reserve and body mass index (BMI), with subsequent adjustments made according to ovarian response. When one dominant follicle reached ≥ 18 mm or two dominant follicles reached ≥ 17 mm, human chorionic gonadotropin (HCG, Lizhu Pharmaceutical Trading Co., China) 5000 ~ 10,000 IU or HCG 2000 IU combined with triptorelin (Decapeptyl, Ferring Pharmaceuticals, Netherlands) 0.1 mg were administered to trigger oocytes maturation, and oocytes were retrieved 34 ~ 36 h later. The oocytes were fertilized by Intracytoplasmic Sperm Injection (ICSI), and fertilization was assessed 16–18 h later. The embryos that reached the blastocyst stage were evaluated morphologically on day 5–7 of embryonic development according to the Gardner scoring criteria [ 8 ]. On the day of blastocyst formation, 5–8 trophoblastic ectodermal cells were obtained by laser drilling. Trophectoderm biopsy was performed on blastocysts graded 3BC or higher, and occasionally on 3CC blastocysts. The biopsied trophectoderm cells transferred to lysis buffer were subjected to whole-genome amplification (WGA) using the MDA approach (REPLI-g Single Cell Kit, QIAGEN Inc.). The MDA products were performed using the preimplantation embryo chromosome aneuploidy detection kit (Jabrehoo, China), and sequencing was performed using the MiSeq high-throughput sequencing platform (Illumina, USA). Bioinformatics analysis of raw sequencing data was performed using the PGXCloud (Jabrehoo, China). The PGT-A results were reported as euploid, aneuploid, mosaic, or testing failure. Aneuploid and mosaic embryos were classified as aberrant. If PGT-A results showed at least one transferable embryo, the endometrium was prepared for frozen–thawed embryo transfer (FET). Depending on the patient’s menstrual cycle, FET was carried out either during a natural cycle, an ovulation induction cycle, or a hormone replacement cycle. [ 9 ]. A single blastocyst was transferred into the uterine cavity on the 5th day after luteal transformation. Serum HCG levels were detected 12 days after ET. Once clinical pregnancy was confirmed, luteal support would continue until 10 weeks of gestation. The primary outcome was the euploid embryo rate per biopsied embryo. Secondary outcome measures included euploid embryo rate per MII and clinical outcomes after euploid blastocyst transfer. Clinical pregnancy referred to the presence of a gestational sac detected by ultrasound examination 30–35 days after ET. Live birth rate was defined as the delivery of a least one live newborn after 24 weeks of gestation. Premature birth was defined as delivery between 24 and 36 +6  weeks of gestation. The miscarriage rate was defined as the number of clinical pregnancy losses before 24 weeks of gestation divided by the number of clinical pregnancies. Statistical calculations were performed using SPSS software (version 26.0, Chicago, USA). The Kruskal–Wallis test was used for continuous variables and Chi-square test was used for categorical variables. A bilateral P value of < 0.05 was considered statistically significant. Linear regression analysis was used to evaluate the relationship between factors of interest and the number of euploid blastocysts. The Bonferroni correction method was used for post hoc testing. This review was conducted under the guidance of the Cochrane Handbook for systematic reviews of interventions and was registered in PROSPERO (registration number: CRD42024554149). This study evaluated whether the PPOS protocol affects the therapeutic efficacy of PGT-A. We searched Chinese and English databases, including PubMed, Embase, Cochrane Library, Web of Science, Sinomed, CNKI, Wanfang and VIP database. Since the PPOS proposal was proposed in 2015, we set the search period from January 2015 to May 2024. We also reviewed the websites of the clinical trial registry. The search protocol for PubMed is provided in Supplementary Table 1. Studies that met the following criteria were included: (1) randomized controlled trials (RCT), non-randomized controlled trials (non-RCT), or observational studies with complete data; (2) patients treated with PGT-A; (3) the intervention group received PPOS protocol without restricted progesterone types, while the control group received traditional ovarian stimulation protocols including GnRH analogues and other protocols. Two investigators independently extracted information from the studies, including the title, authors, year of publication, journal name, country, study design, research span, inclusion and exclusion criteria, intervention and control protocols, patient age, sample size, and outcome data. Any disagreements were resolved through discussion with a third author. In cases where results were not clearly reported but could be inferred from existing raw data, manual calculations would be performed to ensure the comprehensiveness of the data. The primary outcomes were the euploid embryo rate and the number of euploid blastocysts per cycle. Secondary outcomes included: the number of oocytes retrieved; the number of mature oocytes (metaphase II, MII) and 2 pronuclear (2PN) fertilization; ovarian stimulation duration and dosage of gonadotropin (Gn); biochemical pregnancy rate (BPR); clinical pregnancy rate (CPR); ongoing pregnancy rate (OPR); live birth rate (LBR); miscarriage rate and cycle cancelation rate (no transferable euploid embryos due to various reasons). As no RCT had been retrieved, only the ROBINS‐I tool was used to assess the quality of non-RCT studies of interventions or exposure effects [ 10 ]. The ROBINS-I tool covers seven domains of bias in three parts, including pre-intervention (confounding and selection bias), at intervention (bias in measurement classification of interventions), and post-intervention (bias due to deviations from intended interventions, missing data, measurement of outcomes, and selection of the reported results). Overall risk of bias is classified as low, moderate, severe, critical, and no information. Two researchers independently assessed each included study, and any disagreements would be resolved by discussion with a third researcher. The Review Manager software 5.4 (RevMan) (Cochrane Collaboration, Oxford, UK) and Stata 14 software were adopted to perform meta-analyses. Continuous variables were measured using mean difference (MD) with 95% confidence intervals (CIs). Dichotomous results were expressed as the relative risks (RRs) or Odds ratios (ORs) with 95% CI. A P value < 0.05 was considered statistically significant. Heterogeneity among the studies was reported as I-squared ( I 2 ). The random effect model was applied [ 11 ]. Sensitivity and subgroup analyses were performed for results with significant heterogeneity ( I 2  > 50%) to examine potential sources of heterogeneity and validate the robustness of the combined results.

Progestin-primed ovarian stimulation (PPOS) is an emerging approach that uses progestin to prevent the early rise of luteinizing hormone (LH) induced by multiple follicles development during controlled ovarian stimulation [ 1 ], in order to minimize the detrimental effect of premature LH surge on oocyte yield and quality. Compared with the traditional pituitary inhibitor-GnRH analog, oral progestin is more affordable and reduces the need for repeated injections, thereby enhancing patient compliance and satisfaction. However, early exposure to progestin will lead to asynchrony between endometrial and embryo development, making delayed transfer of embryo inevitable, which limits the application of the PPOS protocol in oocyte retrieval-fresh embryo transfer (ET) sequence. Nevertheless, combined with advancing embryo cryopreservation techniques, PPOS protocol may have broader applications in fertility preservation, oocyte donation, embryo accumulation, and prevention of ovarian hyperstimulation syndrome (OHSS) [ 2 ]. PGT-A is an active area in assisted reproductive technology that screens for euploid embryos to improve pregnancy success rates and reduce miscarriage risk. In PGT-A cycles, the purpose of ovarian stimulation has shifted from generating embryos available for transfer to producing an adequate number of embryos for biopsy to increase the opportunities of obtaining euploid embryos [ 3 ]. The application of PPOS protocol in PGT-A cycles lags behind other conventional ovarian stimulation protocols. Concerns about long-term exposure of follicles to progestin remain, and controversy over whether progestin impairs oocyte quality and embryo development potential still exists [ 4 – 7 ]. Careful consideration is still needed regarding the impact of the PPOS protocol on embryos and clinical outcomes. The rate of euploid embryo, as a critical indicator of embryo quality, can serve as a basis for assessing the effectiveness and safety of the PPOS protocol. Our study consists of two parts: (1) a retrospective analysis of embryo results and clinical outcomes under different ovarian stimulation protocols in the PGT-A cycles, and (2) a meta-analysis to evaluate the effectiveness of PPOS protocol in the PGT-A cycles.

For patients undergoing PGT-A, an ideal ovarian stimulation protocol can accumulate enough embryos as quickly as possible without affecting their euploid status while ensuring safety. PPOS protocol has been showed to be effective in inhibiting premature LH surge and reducing the risk of OHSS, and these benefits are related to the dose of progesterone rather than the type used [ 21 ]. In non-PGT-A cycles, PPOS has demonstrated comparable ovarian stimulation efficacy and clinical outcomes to traditional methods [ 22 , 23 ]. Particularly in populations with diminished ovarian reserve, the PPOS protocol has resulted in higher numbers of retrieved oocytes, available embryos, and clinical pregnancy rates compared to control protocols, accompanied by significantly reduced cycle cancelation rate [ 24 , 25 ]. These findings support the application of the PPOS protocol in PGT-A cycles. Furthermore, PPOS protocol can reduce the cost of ovarian stimulation in cases where Gn dosage and stimulation days are similar or lower, alleviating the economic burden of patients [ 26 ]. This may prompt patients with adverse pregnancy outcomes due to advanced age and decreased ovarian function to choose PPOS protocol, as shown in our research and studies by Wang and Huang [ 13 , 16 ]. Nevertheless, the key to determining the choice of ovarian stimulation protocol hinges on the quality of embryos and subsequent clinical outcomes. Despite encouraging data from most studies, debates on the benefits of PPOS still exist. A RCT in 2019 indicated that the use of PPOS protocol in PCOS population resulted in a lower count of MII oocytes and 2PN oocytes [ 5 ]. Another RCT study on oocyte donation cycles indicated unexpectedly lower reproductive outcomes in PPOS group compared to GnRH-ant group [ 27 ]. In PGT-A cycles, Pai’s small-sample study suggested a reduction in euploid blastocyst rate in patients aged ≥ 38 years compared to conventional GnRH-ant protocols [ 14 ]. Huang’s study also demonstrated that the GnRH-ant protocol, rather than the PPOS protocol, was positively correlated with the euploid blastocyst rate and the average number of euploid blastocysts [ 13 ].Conversely, Vaiarelli found no significant differences in embryo parameters between these two protocols in the 35–42 age group [ 15 ]. Wang’s retrospective analysis showed that in the comparison of antagonists, agonists and PPOS protocols, only GnRH-ant protocol was associated with a lower euploidy rate per embryo biopsy [ 17 ]. In our study, patient baseline characteristics were unbalanced, which may have contributed to the significant differences in embryo outcomes among the three groups. In order to reduce the influence of confounding factors, we conducted multiple linear regression analysis, and the results showed that neither MII rate, 2PN rate, blastocyst euploid rate nor average euploid blastocyst number were correlated with ovarian stimulation protocols. Retrospective studies may have the problem of unbalanced baseline characteristics of patients, and the choice of different confounding factors may affect the results of regression analysis [ 13 , 16 , 17 ]. Due to the controversy, in the absence of large-sample prospective RCT studies, we further conducted a meta-analysis to aggregate data from multiple studies to assess the application of the PPOS protocol in PGT-A cycles. Baseline characteristic balance is an important factor for the comparability of observed outcomes. Our meta-analysis showed that when only baseline characteristic balance studies were included, there were no significant differences in embryo outcomes, including the number of oocytes retrieved, MII oocytes, euploid blastocysts, and the rate of euploid blastocyst between PPOS and GnRH-ant protocols. It is worth noting that our subgroup analysis found that the use of dydrogesterone may lead to a decrease in the number of oocytes retrieved, a finding that warrants further investigation to understand the underlying mechanisms. In terms of reproductive outcomes, including our retrospective study, two studies involved the comparison of PPOS protocol with GnRHa protocol, and five studies involved the comparison with GnRH-ant protocol. Regardless of whether the baseline characteristics were balanced, these studies showed similar biochemical pregnancy, clinical pregnancy, ongoing pregnancy or live birth rate between different stimulation protocols after euploid blastocyst transfer. Meta-analysis results also indicated a trend toward lower miscarriage rates with the PPOS protocol compared to the GnRH-ant protocol. Some basic studies suggest that progesterone can regulate the growth and development of follicles, and normalize the early function of the corpus luteum [ 28 – 31 ]. In addition, progesterone may inhibit potential endometriosis through anti-angiogenesis, anti-inflammatory and immunomodulatory effects, thereby improving reproductive outcomes [ 32 ]. There is currently a lack of high-quality RCT studies on the application of the PPOS protocol in the field of PGT-A, and some important confounding factors cannot be fully controlled, which may adversely affect the quality of the meta-analysis. Secondly, existing studies mainly focus on comparisons between the PPOS protocol and GnRH-a protocols with the utilization of trophectoderm biopsy and next-generation sequencing, which limits the generalizability of assessment results. Third, the relatively recent application of the PPOS protocol in PGT-A cycles has not allowed for sufficient long-term follow-up on clinical outcomes, leaving gaps in our understanding of the its potential effects on offspring. Nevertheless, meta-analysis still provides encouraging evidence. Current data indicate that the PPOS protocol is comparable to other ovarian stimulation protocols in embryo euploid status or pregnancy outcomes. Given its cost-effectiveness and operational advantages, the PPOS protocol represents an attractive option for patients undergoing PGT-A treatment. However, due to the retrospective nature of most studies, our findings should be interpreted with caution, and further high-quality RCTs and long-term follow-up are needed to validate the safety and efficacy of the PPOS protocol.