Effects of high basal luteinizing hormone levels on in vitro fertilization/intra-cytoplasmic injection outcomes in polycystic ovarian syndrome: a retrospective study Authors: Hongfa Peng, Lu Wang, Jingjing Jiang, Ting Pan, Yang Yang, Na Cui, Yanan Sun, Guimin Hao, Wei Wang Published: 2025 ABSTRACT BackgroundThe clinical implications of elevated luteinizing hormone (LH) levels in patients with polycystic ovary syndrome (PCOS) undergoing in vitro fertilization/intracytoplasmic sperm injection (IVF/ICSI) fresh embryo transfer cycles remain unclear, particularly regarding oocyte maturation and pregnancy outcomes. We aimed to evaluate the impact of basal LH elevation on treatment outcomes and compare the efficacy of protocols involving gonadotropin-releasing hormone antagonist (GnRH-ant) and gonadotropin-releasing hormone agonist (GnRH-a) for ovarian stimulation in patients with PCOS undergoing in vivo IVF/ICSI.MethodsThis single-center, retrospective cohort study included 1,269 patients with PCOS who underwent IVF/ICSI treatment following a flexible GnRH-ant or long GnRH-a protocol with a human chorionic gonadotropin trigger between January 2013 and December 2023. Primary outcomes included the number of retrieved oocytes, metaphase II (MII) oocytes, two pronuclear embryos, high-quality embryos, and the ovarian hyperstimulation syndrome (OHSS) rate. Secondary outcomes included clinical pregnancy, miscarriage, and live birth rates following fresh embryo transfer.ResultsAmong the participants, 492 (38.77%) exhibited basal LH levels of ≥ 10 mIU/mL, and 777 (61.23%) had LH levels of < 10 mIU/mL. Additionally, 465 (36.64%) and 804 (63.36%) participants underwent GnRH-ant and GnRH-a protocols, respectively. Elevated basal LH levels showed no significant associations with MII oocyte yield, fertilization rate, high-quality embryo count, or clinical pregnancy outcomes (all p > 0.05). However, patients with high LH levels demonstrated increased oocyte retrieval (p = 0.008), a high incidence of ovarian hyperstimulation syndrome (OHSS) (p < 0.001), and reduced fresh embryo transfer rates (p = 0.003). The GnRH-ant protocol required shorter stimulation duration (p < 0.001) and lower amounts of recombinant follicle-stimulating hormone (p < 0.001) than did the GnRH-a protocol. The GnRH-ant protocol also yielded significantly higher proportions of oocytes (p = 0.022), MII oocytes (p = 0.003), and high-quality embryos (p < 0.001); however, it exhibited lower fresh embryo transfer rates (p < 0.001). The two protocols showed no significant differences concerning clinical pregnancy success or OHSS risk.ConclusionsElevated basal LH levels did not impair the outcome of fresh embryo transfer in women with PCOS undergoing IVF/ICSI. However, the number of retrieved oocytes and the OHSS rate were significantly higher in patients with high basal LH levels. The GnRH-ant protocol is an efficient alternative for this subpopulation, particularly regarding stimulation efficiency.Clinical trial numberNot applicable.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13048-025-01899-6. Background Polycystic ovary syndrome (PCOS), a major endocrine disorder affecting women of reproductive age, is characterized by the hypersecretion of luteinizing hormone (LH) owing to the hyperactivity of the gonadotropin-releasing hormone (GnRH) neural circuit and steroid hormone feedback defects [1, 2]. LH is a glycoprotein synthesized by the adenohypophysis that critically regulates folliculogenesis, embryo implantation, and pregnancy maintenance. Anovulatory infertility represents a key feature of PCOS. Elevated basal LH levels drive PCOS-related anovulatory infertility, making in vitro fertilization/intra-cytoplasmic sperm injections (IVF/ICSI) the preferred treatment option for refractory cases. The clinical implications of elevated basal LH levels remain unclear. While some studies have associated high LH levels with compromised oocyte competence and pregnancy outcomes [3, 4], others have reported preserved embryo quality and reproductive success [5, 6]. In this study, we aimed to evaluate the impact of high basal serum LH levels (≥ 10 mIU/mL) on IVF/ICSI outcomes in patients with PCOS and compare the efficacy of stimulation approaches based on GnRH agonists (GnRH-a) or antagonists (GnRH-ant) protocol in this specific subpopulation. Materials and methods Ethics statements This study was conducted per the Declaration of Helsinki and approved by the Institutional Review Board of the Second Hospital of Hebei Medical University (2023KY016). The requirement for written informed consent was waived by the board owing to the retrospective nature of the study. Study design and population This retrospective cohort study included patients with PCOS aged < 40 (range: 20–40) years during their first IVF/ICSI cycle, following a GnRH-ant or a GnRH-a protocol and stimulation with recombinant follicle-stimulating hormone (rFSH) at the Reproductive Medicine Center of the Second Hospital of Hebei Medical University between January 2013 and December 2023. The PCOS diagnosis was based on the Rotterdam criteria [7]. Participation required meeting two or more of the following criteria: oligo or anovulation, clinical and/or biochemical signs of hyperandrogenism, and polycystic ovaries on ultrasonography after excluding other endocrine pathologies. The exclusion criteria were abnormal parental karyotypes; recurrent implantation failure or spontaneous abortions; submucosal fibroids or intrauterine adhesions and endometrial polyps; thyroid dysfunction, elevated prolactin levels or other endocrine diseases; autoimmune diseases; congenital or acquired uterine anomalies; severe male infertility factors; a history of ovarian surgery; and inadequate data for analysis. To eliminate the potential bias resulting from an increase in the rate of mature oocyte retrieval in normal and high responders in response to GnRH-a triggering, only human chorionic gonadotropin (hCG) trigger cycles were included. The patients were divided into high- and low-LH groups, using basal LH levels of 10 mIU/mL as the cut-off. The threshold was determined based on the cut-offs established in prior studies [3, 8]. At this threshold, the model exhibited a sensitivity of 0.953 and a 1-specificity of 0.049. The receiver operating characteristic curve analysis yielded an area under the curve of 0.942. The distribution of basal LH concentrations exhibited marked right-skewness. Concentrations were predominantly clustered at lower levels (median, 8.18 IU/L; interquartile range, 5.26–12.73 IU/L), while a few individuals exhibited substantially higher values. In the long GnRH-a protocol, a long-acting GnRH-a (triptorelin acetate, 3.75 mg; Ferring GmbH, Saint-Prex, Switzerland) or leuprorelin acetate microspheres (3.75 mg; Takeda, Tokyo, Japan) was administered in the early follicular phase. Estradiol (E2), follicle-stimulating hormone (FSH), LH, and progesterone (P) levels were measured after 35–40 days. Follicle diameter was monitored via vaginal ultrasonography. The standards of downregulation included FSH levels < 5 mIU/mL, LH levels < 5 mIU/mL, E2 levels < 30 pg/mL, and follicle diameter 4.5–5 mm. Controlled ovarian stimulation (COS) was initiated upon meeting the aforementioned criteria. In the GnRH-ant protocol, rFSH was administered, starting on cycle Days 2–3. All patients underwent a GnRH-a or a GnRH-ant (Cetrotide, Serono, Geneva, Switzerland) protocol for COS, with daily injections of rFSH (Gonal-F, Serono or Puregon; Organon, Oss, the Netherlands). The starting rFSH dose depended on the patient’s age, body mass index (BMI), and anti-Mullerian hormone (AMH) levels [9]. The dosage was adjusted according to the follicle size and serum levels of FSH, LH, E2, and P [9]. In the GnRH-ant protocol, when the diameter of the dominant follicle reached 12–14 mm or the E2 levels were ≥ 300 ng/mL, the antagonists ganirelix (Merck Sharp & Dohme) or cetrorelix (Merck Serono) were injected subcutaneously at 0.25–0.50 mg/day, according to the weight and serum LH levels of the patients until the trigger day. When the diameter of the dominant follicle reached ≥18 mm, final oocyte maturation was triggered using hCG (Lizhu Pharmaceutical Factory, Zhuhai, China). Transvaginal ultrasonography was used to guide ovum pickup at 36–38 h after injection. Additionally, when the number of follicles on trigger day was ≥20, and the E2 levels were ≥ 5,000 pg/mL, a freeze-all strategy was adopted to avoid ovarian hyperstimulation syndrome (OHSS). Conventional IVF or ICSI was conducted approximately 4–6 h after follicular aspiration, depending on the results of the sperm motility measurements. Morphological criteria were used for embryo scoring. The quality of Day-3 embryos was scored by evaluating the specific number, size, shape, symmetry, the proportion of fragments in the embryo, and the thickness of the zona pellucida. High-quality embryos were those of Grade I or II [10]. Except for patients who refused transfer owing to various factors (including the prevention of OHSS, unavailability of transferable embryos, and abnormal increase in progesterone levels), fresh embryos were transferred—guided by abdominal ultrasonography—on Day 3 after oocyte retrieval. The strategy for embryo transfer was based primarily on embryo quality and maternal age, with single-embryo transfer recommended for patients aged < 35 years who had high-quality embryos, whereas double-embryo transfer was considered when embryo quality was suboptimal or for women aged ≥ 35 years. For luteal support, we administered intramuscular progesterone (40 mg; Xianju Pharmaceutical Factory, Taizhou, China) and oral Duphaston (30 mg; Abbott Healthcare Products B.V., Hoofddorp, the Netherlands) once daily. To assess IVF treatment outcomes, blood tests were performed between 12 and 14 days after embryo transfer. Serum β-HCG levels of >200 IU/L indicated biochemical pregnancy [11]. If the pregnancy test was positive, progesterone was continuously administered until 8–10 weeks of gestation, and an ultrasound examination was performed 30 days after embryo transfer [11]. Clinical pregnancy was defined as the presence of a gestational sac with a fetal heart on ultrasonography, including cases of ectopic pregnancy. All pregnant patients were monitored until delivery. Data collection Age, BMI, serum levels of FSH, LH, E2, P, testosterone (T), and AMH on Day 2/3, and variables related to infertility treatment were collected from the patients’ records. Additionally, we retrieved data on IVF/ICSI cycle characteristics and the number of oocytes retrieved, metaphase II (MII) oocytes, two pronuclear embryos, high-quality embryos, and fresh embryos transferred. The fertilization rate was defined as the total number of embryos formed divided by the total number of oocytes inseminated or injected. Fresh embryo transfer outcomes were analyzed according to clinical pregnancy, miscarriage, and live birth rate. The clinical pregnancy, miscarriage, and live birth rates were calculated. The clinical pregnancy rate was calculated as the number of clinical pregnancy cycles per fresh embryo transfer cycle. The miscarriage rate was defined as the proportion of clinical pregnancies that ended spontaneously before 28 weeks of gestation. Based on the timing of the loss, miscarriages were categorized as early (occurring before 12 weeks) or late (occurring between 12 and 28 weeks). The live birth rate was considered as the number of cycles resulting in at least one live birth per fresh embryo transfer cycle. The OHSS rate was the incidence of severe and moderate OHSS defined according to Golan et al.’s criteria [12]. Statistical analysis Clinical and demographic characteristics, IVF/ICSI cycle characteristics, embryological data, and pregnancy outcomes were analyzed and compared between groups. Statistical analysis was performed using SPSS 26.0 (IBM Corp., Armonk, NY, USA), with normally distributed variables expressed as mean ± standard deviation and those without a normal distribution presented as median (25th and 75th percentiles). Quantitative data were compared between two groups using the rank-sum test or independent samples t-test, and count data were expressed as frequency (%) using the chi-square test. A p-value < 0.05 was considered statistically significant. Results Among the 20,653 patients treated within the study period, 1,269 met the inclusion criteria. There were 492 and 777 patients in the high- and low-LH groups, respectively. Of the 1,269 patients included, 465 and 804 were treated with the GnRH-ant and GnRH-a protocols, respectively. At the beginning of the COS, the mean age of the patients was 28.71 ± 3.53 (range: 24–40) years, and their mean BMI was 25.59 ± 3.82 (range: 16.2–42.24.2.24) kg/m2. The median basal LH levels were 9.77 ± 6.29 mIU/mL. Furthermore, 72.97% of the patients were nulliparous at diagnosis. Of the 1,269 patients, 1,260 underwent egg retrieval, nine had their cycles canceled owing to hyperresponse or hypoergia of the ovary, and 87 (6.86%) developed OHSS. Among the 1,260 patients who underwent egg retrieval, 990 and 228 corresponded to IVF and ICSI, respectively; 538 underwent fresh embryo transfers, whereas 51 had no embryos available for transfer. The clinical pregnancy, miscarriage, and live birth rates were 60.41%,18.62%, and 48.14%, respectively. The clinical and demographic characteristics of the patients are presented in Table 1. Table 1Patients’ baseline characteristicsclinical parametersvalues (mean ± SD)Age (yr)28.71 ± 3.53BMI (kg/m2)25.59 ± 3.82Baseline FSH (mIU/mL)6.44 ± 1.85Baseline LH (mIU/mL)9.77 ± 6.29Baseline T (ng/mL)0.87 ± 3.17Baseline E2 (pg/mL)46.65 ± 33.42Baseline p (mg/L)0.88 ± 1.94Baseline AMH (ng/mL)8.24 ± 4.52Infertility diagnosis, n (%) primary926(72.97%) secondary343(27.03%)P on hCG day (mg/L)1.28 ± 0.65LH on hCG day (mIU/mL)2.49 ± 3.39E2 on hCG day (pg/mL)3962.42 ± 1654.42No. of total retrieved oocytes18.57 ± 9.60Total of Gn days11.52 ± 3.15Total rFSH dose (IU)2354.25 ± 1091.11No. of total retrieved oocytes18.50 ± 9.32No. of IVF cases n (%)1027 (81.38%)No. of ICSI cases n (%)235 (18.62%)No. of embryos transferred n (%)538 (42.40%)Clinical pregnancy rate n (%)325 (60.41%)Live birth rate n (%)259 (48.14%)No. of OHSS n (%)87 (6.86%)BMI Body mass index, FSH Follicle-stimulating hormone, LH Luteinizing hormone, T Testosterone, E2 Estradiol, P Progesterone, AMH Anti-Mullerian hormone, hCG Human chorionic gonadotropin, Gn Gonadotropins, rFSH Recombinant follicle-stimulating hormone, IVF In Vitro Fertilization, ICSI Intra-Cytoplasmic Injections, OHSS Ovarian hyperstimulation syndrome Table 1Patients’ baseline characteristicsclinical parametersvalues (mean ± SD)Age (yr)28.71 ± 3.53BMI (kg/m2)25.59 ± 3.82Baseline FSH (mIU/mL)6.44 ± 1.85Baseline LH (mIU/mL)9.77 ± 6.29Baseline T (ng/mL)0.87 ± 3.17Baseline E2 (pg/mL)46.65 ± 33.42Baseline p (mg/L)0.88 ± 1.94Baseline AMH (ng/mL)8.24 ± 4.52Infertility diagnosis, n (%) primary926(72.97%) secondary343(27.03%)P on hCG day (mg/L)1.28 ± 0.65LH on hCG day (mIU/mL)2.49 ± 3.39E2 on hCG day (pg/mL)3962.42 ± 1654.42No. of total retrieved oocytes18.57 ± 9.60Total of Gn days11.52 ± 3.15Total rFSH dose (IU)2354.25 ± 1091.11No. of total retrieved oocytes18.50 ± 9.32No. of IVF cases n (%)1027 (81.38%)No. of ICSI cases n (%)235 (18.62%)No. of embryos transferred n (%)538 (42.40%)Clinical pregnancy rate n (%)325 (60.41%)Live birth rate n (%)259 (48.14%)No. of OHSS n (%)87 (6.86%)BMI Body mass index, FSH Follicle-stimulating hormone, LH Luteinizing hormone, T Testosterone, E2 Estradiol, P Progesterone, AMH Anti-Mullerian hormone, hCG Human chorionic gonadotropin, Gn Gonadotropins, rFSH Recombinant follicle-stimulating hormone, IVF In Vitro Fertilization, ICSI Intra-Cytoplasmic Injections, OHSS Ovarian hyperstimulation syndrome Patients’ baseline characteristics BMI Body mass index, FSH Follicle-stimulating hormone, LH Luteinizing hormone, T Testosterone, E2 Estradiol, P Progesterone, AMH Anti-Mullerian hormone, hCG Human chorionic gonadotropin, Gn Gonadotropins, rFSH Recombinant follicle-stimulating hormone, IVF In Vitro Fertilization, ICSI Intra-Cytoplasmic Injections, OHSS Ovarian hyperstimulation syndrome Baseline characteristics (age, infertility type, P levels, and T levels) were similar between high- and low-LH groups (all p > 0.05). The high-LH group showed higher basal levels of FSH (mean difference [MD] = 0.920, 95% confidence interval (CI) [0.718, 1.122] mIU/mL; p < 0.001), E2 (MD = 16.355, 95%CI [11.056, 21.654] pg/mL; p < 0.001), and AMH (MD = 2.011, 95%CI [1.405, 2.617] ng/mL; p < 0.001) and a lower BMI (MD=−0.791, 95%CI [−1.227, −0.356] kg/m2; p < 0.001) than did the low-LH group (Table 2). No differences were observed in stimulation duration, P levels at trigger day, cycle cancellation rates, oocyte/embryo parameters (MII count, two-pronuclei fertilization rate, or number of high-quality embryos), or clinical outcomes (pregnancy, miscarriage, or live birth rate) (all p > 0.05) between the high- and low-LH groups. The high-LH group required a lower rFSH dose (MD=−90.646, 95%CI [−169.341, −11.952] IU; p = 0.024) and had higher E2 (MD = 215.470, 95%CI [24.36, 406.576] pg/mL; p = 0.027) and LH (MD = 0.932, 95%CI [0.545, 1.320] mIU/mL; p < 0.001) levels on trigger day, a higher number of retrieved oocytes (MD = 1.468, 95%CI [0.384, 2.552] oocytes; p = 0.008), a higher OHSS rate (odds ratio [OR] = 2.262, 95%CI [1.456, 3.157]; p < 0.001), and a lower fresh embryo transfer rate (OR = 0.704, 95%CI [0.559, 0.887]; p = 0.003) than did the low-LH group (Table 3). Table 2Baseline characteristics of patients in the LH ≥ 10 IU/L and LH < 10IU/L groupsParametersLH ≥ 10 (n = 492)LH<10 (n = 777)p valueAge (years)28.55 ± 3.4228.82 ± 3.600.19BMI (kg/m2)25.12 ± 3.6825.88 ± 3.870.001Infertility diagnosis, n (%) primary365(74.19%)560(72.52%)0.43 secondary127(25.81%)216(27.48%)Baseline FSH (mIU/mL)7.01 ± 1.916.09 ± 1.70<0.001Baseline LH (mIU/mL)15.70 ± 5.675.88 ± 2.24<0.001Baseline T (ng/mL)0.83 ± 0.560.93 ± 3.330.433Baseline E2 (pg/mL)55.72 ± 41.6440.50 ± 24.70<0.001Baseline P (mg/L)0.79 ± 1.140.94 ± 2.350.743Baseline AMH (ng/mL)9.40 ± 4.827.43 ± 4.13<0.001BMI Body mass index, FSH Follicle-stimulating hormone, LH Luteinizing hormone, T Testosterone, E2 Estradiol, P Progesterone, AMH Anti-Mullerian hormone Table 2Baseline characteristics of patients in the LH ≥ 10 IU/L and LH < 10IU/L groupsParametersLH ≥ 10 (n = 492)LH<10 (n = 777)p valueAge (years)28.55 ± 3.4228.82 ± 3.600.19BMI (kg/m2)25.12 ± 3.6825.88 ± 3.870.001Infertility diagnosis, n (%) primary365(74.19%)560(72.52%)0.43 secondary127(25.81%)216(27.48%)Baseline FSH (mIU/mL)7.01 ± 1.916.09 ± 1.70<0.001Baseline LH (mIU/mL)15.70 ± 5.675.88 ± 2.24<0.001Baseline T (ng/mL)0.83 ± 0.560.93 ± 3.330.433Baseline E2 (pg/mL)55.72 ± 41.6440.50 ± 24.70<0.001Baseline P (mg/L)0.79 ± 1.140.94 ± 2.350.743Baseline AMH (ng/mL)9.40 ± 4.827.43 ± 4.13<0.001BMI Body mass index, FSH Follicle-stimulating hormone, LH Luteinizing hormone, T Testosterone, E2 Estradiol, P Progesterone, AMH Anti-Mullerian hormone Baseline characteristics of patients in the LH ≥ 10 IU/L and LH < 10IU/L groups BMI Body mass index, FSH Follicle-stimulating hormone, LH Luteinizing hormone, T Testosterone, E2 Estradiol, P Progesterone, AMH Anti-Mullerian hormone Table 3IVF/ICSI cycle characteristics and fresh embryo transfer outcomes of patients in the LH ≥ 10 IU/L and LH < 10IU/L groupsParametersLH>10 (n = 492)LH<10 (n = 777)p valueTotal of Gn days11.51 ± 3.1511.62 ± 3.280.585Total rFSH dose (IU)1870.31 ± 685.811960.95 ± 701.010.024hormone profile at trigger day E2 (pg/ml)4084.30 ± 1696.183868.83 ± 1622.210.027 P (ng/ml)1.31 ± 0.641.26 ± 0.660.199 LH (mIU/mL)3.06 ± 4.462.12 ± 2.39<0.001No. of retrieved oocytes19.55 ± 10.5617.98 ± 8.900.008No. of Metaphase II oocytes13.00 ± 8.4312.35 ± 8.000.117Mode of fertilization, n (%) IVF402(82.38%)595(76.87%)0.294 ICSI86(17.62%)149(23.13%)No. of 2PN11.37 ± 7.3210.92 ± 6.740.263No. of high-quality embryo2.44 ± 3.112.35 ± 3.160.612No. of available embryo5.24 ± 3.484.98 ± 3.480.183No. of OHSS n (%)50 (10.16%)37 (4.76%)<0.001No. of cycle Cancellation5 (1.02%)3 (0.39%)0.3092PN Fertilized rate (% of total retrieved oocytes)59.12 ± 26.0860.33 ± 21.070.386Metaphase II oocytes rate (% of total oocytes)67.78 ± 28.2065.87 ± 28.480.243High-quality embryo rate (% of total oocytes)13.41 ± 17.5212.90 ± 15.760.598Available embryo rate (% of total oocytes)30.39 ± 22.0229.23 ± 16.980.293No. of FET cycles, n (% of total cycles)183(37.20%)355(45.69%)0.003Clinical pregnancy, n (% of FET cycles)113(61.75%)212(59.72%)0.648Miscarriages, n (% of clinical pregnancy)20(17.70%)40(18.87%)0.796Early miscarriages, n (% of clinical pregnancy)10(8.85%)24(11.32%)0.488Late miscarriages, n (% of clinical pregnancy)10(8.85%)16(7.55%)0.68Live births, n (% of FET cycles)89(48.63%)170(47.89%)0.87Gn Gonadotropins, rFSH Recombinant follicle-stimulating hormone, E2 Estradiol, P Progesterone, LH Luteinizing hormone, IVF In Vitro Fertilization, ICSI Intra-Cytoplasmic Injections, PN Pronuclei, OHSS Ovarian hyperstimulation syndrome, FET Fresh embryo transfer Table 3IVF/ICSI cycle characteristics and fresh embryo transfer outcomes of patients in the LH ≥ 10 IU/L and LH < 10IU/L groupsParametersLH>10 (n = 492)LH<10 (n = 777)p valueTotal of Gn days11.51 ± 3.1511.62 ± 3.280.585Total rFSH dose (IU)1870.31 ± 685.811960.95 ± 701.010.024hormone profile at trigger day E2 (pg/ml)4084.30 ± 1696.183868.83 ± 1622.210.027 P (ng/ml)1.31 ± 0.641.26 ± 0.660.199 LH (mIU/mL)3.06 ± 4.462.12 ± 2.39<0.001No. of retrieved oocytes19.55 ± 10.5617.98 ± 8.900.008No. of Metaphase II oocytes13.00 ± 8.4312.35 ± 8.000.117Mode of fertilization, n (%) IVF402(82.38%)595(76.87%)0.294 ICSI86(17.62%)149(23.13%)No. of 2PN11.37 ± 7.3210.92 ± 6.740.263No. of high-quality embryo2.44 ± 3.112.35 ± 3.160.612No. of available embryo5.24 ± 3.484.98 ± 3.480.183No. of OHSS n (%)50 (10.16%)37 (4.76%)<0.001No. of cycle Cancellation5 (1.02%)3 (0.39%)0.3092PN Fertilized rate (% of total retrieved oocytes)59.12 ± 26.0860.33 ± 21.070.386Metaphase II oocytes rate (% of total oocytes)67.78 ± 28.2065.87 ± 28.480.243High-quality embryo rate (% of total oocytes)13.41 ± 17.5212.90 ± 15.760.598Available embryo rate (% of total oocytes)30.39 ± 22.0229.23 ± 16.980.293No. of FET cycles, n (% of total cycles)183(37.20%)355(45.69%)0.003Clinical pregnancy, n (% of FET cycles)113(61.75%)212(59.72%)0.648Miscarriages, n (% of clinical pregnancy)20(17.70%)40(18.87%)0.796Early miscarriages, n (% of clinical pregnancy)10(8.85%)24(11.32%)0.488Late miscarriages, n (% of clinical pregnancy)10(8.85%)16(7.55%)0.68Live births, n (% of FET cycles)89(48.63%)170(47.89%)0.87Gn Gonadotropins, rFSH Recombinant follicle-stimulating hormone, E2 Estradiol, P Progesterone, LH Luteinizing hormone, IVF In Vitro Fertilization, ICSI Intra-Cytoplasmic Injections, PN Pronuclei, OHSS Ovarian hyperstimulation syndrome, FET Fresh embryo transfer IVF/ICSI cycle characteristics and fresh embryo transfer outcomes of patients in the LH ≥ 10 IU/L and LH < 10IU/L groups Gn Gonadotropins, rFSH Recombinant follicle-stimulating hormone, E2 Estradiol, P Progesterone, LH Luteinizing hormone, IVF In Vitro Fertilization, ICSI Intra-Cytoplasmic Injections, PN Pronuclei, OHSS Ovarian hyperstimulation syndrome, FET Fresh embryo transfer Baseline characteristics, including age, infertility type, and LH, E2, P, T, and AMH levels, were comparable between the GnRH-ant and GnRH-a groups (Table 4). The GnRH-ant group exhibited a higher mean BMI (MD = 0.603, 95%CI [0.160, 1.046] kg/m², p = 0.004) and required shorter gonadotropin stimulation (MD=−1.567, 95%CI [-1.859, -1.267] days; p < 0.001) and lower amounts of rFSH (MD=-581.104, 95%CI [-693.355, -468.854] IU; p < 0.001) than did the GnRH-a group. On the trigger day, patients treated with a GnRH-ant protocol showed elevated E2 (MD = 460.271, 95%CI [269.674, 650.868] pg/mL; p < 0.001), P (MD = 0.272, 95%CI [0.197, 0.346] mg/mL; p < 0.001), and LH (MD = 3.702, 95%CI [3.370, 4.035] mIU/mL; p < 0.001) levels compared with those in individuals treated with a GnRH-a protocol. The GnRH-ant protocol group also had a higher proportion of retrieved oocytes (MD = 2.712, 95%CI [1.624, 3.801] oocytes; p = 0.022), more MII oocytes (MD = 1.739, 95%CI [0.791, 2.687] oocytes; p = 0.003), and a higher rate of MII oocytes (OR = 0.052, 95%CI [0.018, 0.086]; p = 0.012). Despite improved embryo quality (high-quality embryos: MD = 0.892, 95%CI [0.534, 1.250] embryos; p < 0.001), the GnRH-ant group had a lower fresh embryo transfer rate (OR = 0.550, 95%CI [0.433, 0.697]; p < 0.001). No significant differences were observed in clinical outcomes (live birth, clinical pregnancy, miscarriage, OHSS, or cycle cancellation rates) (Table 5). Table 4Baseline characteristics of patients in GnRH-ant and GnRH-a groupsParametersGnRH-ant (n = 465)GnRH-a (n = 804)p valueAge (years)28.8 ± 3.4628.67 ± 3.570.517BMI (kg/m2)26.00 ± 3.6825.35 ± 3.870.004Infertility diagnosis, n (%) primary335(72.04%)590(73.38%)0.605 secondary130(27.96%)214(26.62%)Baseline FSH (mIU/mL)6.46 ± 1.836.45 ± 1.840.973Baseline LH (mIU/mL)9.65 ± 6.939.95 ± 6.690.215Baseline T (ng/mL)0.93 ± 3.770.87 ± 1.630.694Baseline E2 (pg/mL)50.14 ± 40.054.23 ± 50.7480.14Baseline P (mg/mL)0.95 ± 1.631.02 ± 1.990.577Baseline AMH7.94 ± 4.478.40 ± 4.530.158BMI Body mass index, FSH Follicle-stimulating hormone, LH Luteinizing hormone, T Testosterone, E2 Estradiol, P Progesterone, AMH Anti-Mullerian hormone Table 4Baseline characteristics of patients in GnRH-ant and GnRH-a groupsParametersGnRH-ant (n = 465)GnRH-a (n = 804)p valueAge (years)28.8 ± 3.4628.67 ± 3.570.517BMI (kg/m2)26.00 ± 3.6825.35 ± 3.870.004Infertility diagnosis, n (%) primary335(72.04%)590(73.38%)0.605 secondary130(27.96%)214(26.62%)Baseline FSH (mIU/mL)6.46 ± 1.836.45 ± 1.840.973Baseline LH (mIU/mL)9.65 ± 6.939.95 ± 6.690.215Baseline T (ng/mL)0.93 ± 3.770.87 ± 1.630.694Baseline E2 (pg/mL)50.14 ± 40.054.23 ± 50.7480.14Baseline P (mg/mL)0.95 ± 1.631.02 ± 1.990.577Baseline AMH7.94 ± 4.478.40 ± 4.530.158BMI Body mass index, FSH Follicle-stimulating hormone, LH Luteinizing hormone, T Testosterone, E2 Estradiol, P Progesterone, AMH Anti-Mullerian hormone Baseline characteristics of patients in GnRH-ant and GnRH-a groups BMI Body mass index, FSH Follicle-stimulating hormone, LH Luteinizing hormone, T Testosterone, E2 Estradiol, P Progesterone, AMH Anti-Mullerian hormone Table 5IVF/ICSI cycle characteristics and fresh embryo transfer outcomes of patients in GnRH-ant and GnRH-a groupsParametersGnRH-ant (n = 465)GnRH-a (n = 804)p valueTotal of Gn days9.81 ± 2.2211.38 ± 2.72<0.001Total rFSH dose (IU)1986.12 ± 869.082567.23 ± 1148.77<0.001hormone profile at trigger day E2 (pg/ml)4238.48 ± 2001.863778.20 ± 1373.21<0.001 P (ng/ml)1.44 ± 0.771.17 ± 0.55<0.001 LH (mIU/mL)4.78 ± 4.401.08 ± 1.17<0.001No. of retrieved oocytes19.32 ± 9.9418.04 ± 8.940.022No, of Metaphase II oocytes13.49 ± 8.0412.07 ± 8.210.003Mode of fertilization, n (%) IVF380(82.25%)647(81.08%)0.605 ICSI82(17.75%)151(18.92%)No. of 2PN11.10 ± 7.1111.08 ± 6.900.969No. of high-quality embryo2.95 ± 3.502.06 ± 2.86<0.001No. of available embryo5.32 ± 3.524.94 ± 3.450.058No. of OHSS, n (%)26(5.59%)61(7.59%)0.175No. of cycle Cancellation, n (% of clinical ET cycles)4(0.86%)5(0.62%)0.8882PN Fertilized rate (% of total retrieved oocytes)56.8 ± 21.2561.64 ± 23.99<0.001Metaphase II oocytes rate (% of total oocytes)68.95 ± 20.1765.25 ± 32.130.012High-quality embryo rate (% of total oocytes)15.4 ± 16.6211.77 ± 16.22<0.001Available embryo rate (% of total oocytes)29.68 ± 15.9329.97 ± 20.600.7785No. of FET cycles, n (% of total cycles)155(33.41%)383(47.76%)<0.001Clinical pregnancy, n (% of FET cycles)94(60.65%)231(60.31%)0.943Miscarriages, n (% of clinical pregnancy)17(18.09%)44(19.05%)0.84Early miscarriages, n (% of clinical pregnancy)11(11.70%)23(9.96%)0.641Late miscarriages, n (% of clinical pregnancy)5(5.32%)21(9.09%)0.256Live births, n (% of FET cycles)73(40.10%)186(48.56%)0.758Gn Gonadotropins, rFSH Recombinant follicle-stimulating hormone, E2 Estradiol, P Progesterone, LH Luteinizing hormone, IVF In Vitro Fertilization, ICSI Intra-Cytoplasmic Injections, PN Pronuclei, OHSS Ovarian hyperstimulation syndrome, FET Fresh embryos transfer Table 5IVF/ICSI cycle characteristics and fresh embryo transfer outcomes of patients in GnRH-ant and GnRH-a groupsParametersGnRH-ant (n = 465)GnRH-a (n = 804)p valueTotal of Gn days9.81 ± 2.2211.38 ± 2.72<0.001Total rFSH dose (IU)1986.12 ± 869.082567.23 ± 1148.77<0.001hormone profile at trigger day E2 (pg/ml)4238.48 ± 2001.863778.20 ± 1373.21<0.001 P (ng/ml)1.44 ± 0.771.17 ± 0.55<0.001 LH (mIU/mL)4.78 ± 4.401.08 ± 1.17<0.001No. of retrieved oocytes19.32 ± 9.9418.04 ± 8.940.022No, of Metaphase II oocytes13.49 ± 8.0412.07 ± 8.210.003Mode of fertilization, n (%) IVF380(82.25%)647(81.08%)0.605 ICSI82(17.75%)151(18.92%)No. of 2PN11.10 ± 7.1111.08 ± 6.900.969No. of high-quality embryo2.95 ± 3.502.06 ± 2.86<0.001No. of available embryo5.32 ± 3.524.94 ± 3.450.058No. of OHSS, n (%)26(5.59%)61(7.59%)0.175No. of cycle Cancellation, n (% of clinical ET cycles)4(0.86%)5(0.62%)0.8882PN Fertilized rate (% of total retrieved oocytes)56.8 ± 21.2561.64 ± 23.99<0.001Metaphase II oocytes rate (% of total oocytes)68.95 ± 20.1765.25 ± 32.130.012High-quality embryo rate (% of total oocytes)15.4 ± 16.6211.77 ± 16.22<0.001Available embryo rate (% of total oocytes)29.68 ± 15.9329.97 ± 20.600.7785No. of FET cycles, n (% of total cycles)155(33.41%)383(47.76%)<0.001Clinical pregnancy, n (% of FET cycles)94(60.65%)231(60.31%)0.943Miscarriages, n (% of clinical pregnancy)17(18.09%)44(19.05%)0.84Early miscarriages, n (% of clinical pregnancy)11(11.70%)23(9.96%)0.641Late miscarriages, n (% of clinical pregnancy)5(5.32%)21(9.09%)0.256Live births, n (% of FET cycles)73(40.10%)186(48.56%)0.758Gn Gonadotropins, rFSH Recombinant follicle-stimulating hormone, E2 Estradiol, P Progesterone, LH Luteinizing hormone, IVF In Vitro Fertilization, ICSI Intra-Cytoplasmic Injections, PN Pronuclei, OHSS Ovarian hyperstimulation syndrome, FET Fresh embryos transfer IVF/ICSI cycle characteristics and fresh embryo transfer outcomes of patients in GnRH-ant and GnRH-a groups Gn Gonadotropins, rFSH Recombinant follicle-stimulating hormone, E2 Estradiol, P Progesterone, LH Luteinizing hormone, IVF In Vitro Fertilization, ICSI Intra-Cytoplasmic Injections, PN Pronuclei, OHSS Ovarian hyperstimulation syndrome, FET Fresh embryos transfer Discussion This study showed that in patients with PCOS, elevated basal LH levels (≥ 10 mIU/mL) were significantly associated with a higher proportion of retrieved oocytes, an increased OHSS risk, and a lower fresh embryo transfer rate; however, the high basal LH levels exerted no negative effects on clinical pregnancy outcomes after transfer. Furthermore, the GnRH-ant protocol demonstrated superior stimulation efficiency despite achieving pregnancy outcomes comparable to those of the GnRH-a protocol. LH critically regulates folliculogenesis, ovulation, and steroidogenesis, while modulating embryo implantation and luteal function. This endocrine axis is particularly relevant in patients with PCOS exhibiting LH hypersecretion, where elevated follicular-phase LH may trigger early oocyte maturation and ovulation, potentially compromising implantation efficacy [13]. Our findings revealed that elevated basal LH levels (≥ 10 mIU/mL) were associated with increased oocyte yield and OHSS risk. This result is consistent with those of previous studies [8]. However, embryological outcomes were similar in patients with high and low basal LH levels. These results contradict with prior reports linking elevated LH to impaired oocyte competence and poor embryological outcomes [8, 14]. The discrepancy may stem from methodological variations, including heterogeneous LH cut-offs (9–12.45 mIU/mL across studies), population-specific confounders such as advanced maternal age or diminished ovarian reserve, and the timing of LH measurement. Notably, our cohort showed a balanced age distribution between the two LH groups, minimizing potential age-related biases in the interpretation of the outcomes. Ovarian stimulation and pregnancy outcomes are related to mid-follicular LH levels in women undergoing assisted reproduction with GnRH analog downregulation and rFSH stimulation [15]. A significant increase or decrease in LH levels during the mid-follicular phase affects follicular development and endometrial receptivity, influencing pregnancy outcomes. Thus, an adequate serum LH concentration is crucial for FSH-induced follicular development [16]. However, the appropriate LH levels for COS lack consensus. Some studies have shown that LH levels of < 0.5–1.2 mIU/mL affect the clinical outcomes of FSH-induced COS cycles [15, 17]. Some reports indicate that the optimal LH levels for follicular development are 1.2–5 IU/L [18, 19]. Notably, one study proposed that LH levels ≥ 4 IU/L should be considered the critical value and recommended antagonists when this threshold is exceeded [20]. These findings indicate that LH levels may have an individual optimal value window during follicular development. LH levels should not be excessively high or low during COS. In addition, abnormal high-amplitude LH pulse secretion characterizes neuroendocrine dysfunction in PCOS and is closely associated with hyperandrogenism [21, 22]. Studies on PCOS etiology show that elevated LH levels stimulate theca cells directly or indirectly. This stimulation increases androgen production, including testosterone. Hyperandrogenism impairs folliculogenesis and oocyte development via several mechanisms: high androgen levels arrest antral follicle growth [23] and disrupt the follicular microenvironment—for example, by altering cytokine secretion and other molecular factors—inhibiting oocyte maturation [24]. In this study, the T levels exhibited no significant differences between the high- and low-LH groups. This may be influenced by inherent biases within the study. Reasonable LH levels during ovulation induction are essential for positive pregnancy outcomes. Tonic hypersecretion of LH impairs embryo quality, embryo implantation, and pregnancy persistence [25, 26]. However, severe LH suppression is detrimental in patients receiving cycles of gonadotropin therapy [27]. In this study, no statistically significant differences in clinical pregnancy, miscarriage, or live birth rates were observed between the high- and low-basal LH groups after fresh embryo transfer. Similarly, some studies have shown that elevated basal LH levels do not increase miscarriage risk in women with PCOS [6, 8, 28]. These results suggest that high basal LH levels do not affect pregnancy outcomes in patients with PCOS undergoing fresh IVF cycles. This may be related to the suppression of basal LH levels by gonadotropins during stimulation. Mid-follicular LH levels may affect pregnancy outcomes than do basal LH levels. The latest American Society for Reproductive Medicine consensus indicates that adjusting the gonadotropin dose via the dynamic monitoring of LH may increase the live birth rate in patients with PCOS to 38.7%, which is 9.2% higher than that observed in patients with a fixed dose [29]. In the present study, although the high-LH group had significantly higher LH levels on hCG administration day than did the low-LH group, the LH concentrations in both groups were within the recommended range. Patients with high LH levels (basal LH >10 IU/L) are at risk of premature follicular luteinization. The premature activation of the LH receptor in granulosa cells may disrupt meiotic resumption, impairing oocyte maturation and resulting in a 12–18% reduction in MII rate [6]. Given this risk, the stratification of ovulation induction regimens based on basal LH levels could be considered for patients with PCOS. Nonetheless, the optimal ovarian stimulation protocol for this population remains controversial, with conflicting evidence regarding the ovarian stimulation outcomes of GnRH-a and GnRH-ant protocols. While multiple studies imply that the GnRH-ant protocol requires a shorter stimulation period and lower amounts of gonadotropin than does the GnRH-a protocol, discrepancies exist regarding oocyte yield outcomes [30, 31]. Some reports indicated reduced oocyte retrieval in patients treated with the GnRH-ant protocol [30, 32], inconsistent with our findings and those reported by Hosseini et al. [33], which showed superior oocyte yield and maturation rates with this protocol. This inconsistency may stem from population heterogeneity and uncontrolled confounders such as metabolic parameters (glucose control and lipid profiles), lifestyle interventions, and variations in BMI. Notably, the patients who received GnRH-ant in our study had a higher BMI despite demonstrating better oocyte outcomes than did those who underwent the GnRH-a protocol. The observed outcome disparities underscore the need for prospective studies controlling for metabolic factors and lifestyle interventions in patients with PCOS. In addition, these differences possibly stemmed from protocol-specific gonadotropin dynamics. The GnRH-a and GnRH-ant protocols differ significantly in their regulation of LH. GnRH-a induces pituitary desensitization after the initial hyperstimulation, maintaining stable low LH throughout stimulation. Conversely, GnRH-ant permits early-phase LH elevation, followed by rapid post-antagonist suppression and subsequent gradual recovery. LH level management is critical for ovulation induction in patients with PCOS owing to their characteristically elevated basal LH levels and heightened risks of premature LH surges. Excessive and insufficient follicular-phase LH adversely affect folliculogenesis, oocyte competence, and pregnancy outcomes, potentially driving early pregnancy loss [16]. Optimal follicular development requires balanced LH activity (1.2–5.0 IU/L), as subthreshold levels (< 1.2 IU/L) during FSH stimulation may compromise oocyte maturation and embryogenesis, increasing the risk of miscarriage [34, 35]. In this study, patients treated with the GnRH-a protocol had their basal LH levels reduced to < 5 mIU/mL before the initiation of ovarian stimulation. Conversely, patients who underwent the GnRH-ant protocol maintained basal LH levels above 10 mIU/mL at the start of ovarian stimulation. Based on this distinction, we hypothesized that if high basal LH levels negatively impacted ovarian stimulation outcomes, the GnRH-a protocol should theoretically mitigate this effect. However, in patients with PCOS with basal LH >10 mIU/mL, the GnRH-a and GnRH-ant protocols yielded comparable results in most outcomes, including ovarian stimulation and pregnancy rates. The only exception was a significant difference in the number of high-quality embryos (Supplemental Table 1). This result does not support the earlier hypothesis. The effect of high basal LH levels on ovarian stimulation outcomes in patients with PCOS remains controversial [5, 6, 36]. Our findings, which suggest no significant effect, are consistent with the evidence indicating that elevated LH levels may not be detrimental. Nevertheless, this conclusion warrants caution owing to potential inherent biases in the study design, such as factors related to sample selection or specific protocol implementation. The potential ability of GnRH-a to counteract negative effects would require confirmation through a more rigorously controlled investigation. Regarding pregnancy outcomes in patients with PCOS, current evidence indicates that GnRH antagonist and agonist long protocols yield comparable results. Although some studies have reported associations between GnRH-ant protocol and reduced clinical pregnancy and live birth rates following fresh embryo transfers [37–39], others have described improved pregnancy outcomes [33, 40]. Nonetheless, high-quality meta-analyses and randomized controlled trials consistently demonstrate no statistically significant differences between these protocols in key efficacy endpoints, including clinical pregnancy and live birth rates [30, 31, 40–42]. Aligned with these observations, our study revealed no statistically significant difference in clinical pregnancy or live birth rates between the GnRH agonist and antagonist groups. Notably, although the GnRH-ant group exhibited higher LH levels on the day of hCG trigger—which correlated with improved laboratory parameters such as oocyte maturation rate and embryo quality—these advantages did not translate into superior pregnancy outcomes compared with those in the GnRH-a group. These findings suggest that LH threshold requirements may differ between folliculogenesis and post-fertilization stages, warranting further investigation into protocol-specific LH dynamics and their clinical implications. Concerning safety, the GnRH-ant protocol demonstrated a superior profile, particularly through a marked reduction in the incidence of OHSS—a critical advantage in high-risk patients with PCOS. Meta-analyses indicate that the GnRH-ant protocol is associated with a 40–60% reduction in OHSS incidence compared with the effect of the GnRH-a protocol in this population, particularly when incorporating GnRH-a triggering and freeze-all strategies [30, 43]. This enhanced safety profile may support an increased proportion of fresh embryo transfers, potentially reducing the time to pregnancy. Nonetheless, some studies reported conflicting results, noting comparable OHSS rates between protocols [30, 44]. In our study, no significant differences in OHSS incidence were observed, possibly owing to the consistent application of standardized preventive measures. This discrepancy underscores the significance of risk mitigation strategies when comparing treatment protocols. For patients with PCOS undergoing IVF, optimizing outcomes involves balancing the characteristic hyper-response with effective OHSS prevention. Individualized stimulation protocols should consider key patient-specific factors, such as basal LH levels, LH/FSH ratio, ovarian phenotype, and metabolic profile. The GnRH-ant protocol emerges as a preferable strategy in this population, offering an optimal balance between oocyte yield and significantly reduced OHSS risk [45]. Additional benefits include reduced stimulation duration, reduced gonadotropin requirements, and flexible dosing, which enhance treatment efficiency and tolerability. Strengths and limitations To our knowledge, this is the largest retrospective study designed to assess two key conditions in patients with PCOS undergoing IVF/ICSI with fresh embryo transfer: the impact of elevated basal LH levels and the comparative efficacy of GnRH-ant versus GnRH-a protocols. While providing novel insights, this study has some limitations inherent to its single-center design, including potential selection bias and unaddressed confounders such as BMI variability, heterogeneous PCOS phenotypes, and dynamic endocrine profiles. Evolving laboratory practices over the 10-year study period may also have introduced heterogeneity. Additionally, the study did not consider frozen embryo transfers, which could provide further insight into long-term success rates. Conclusions In patients with PCOS, elevated basal LH levels showed no adverse effects on oocyte quality, fertilization rates, embryo quality, or fresh transfer outcomes; however, they were linked to increased oocyte yields and OHSS risk. While GnRH-a and GnRH-ant protocols achieved comparable success rates for fresh embryo transfer, the GnRH-ant protocol demonstrated superior efficacy at ovulation induction. Supplementary Information Supplementary Material 1. Supplementary Material 1.