Letrozole-stimulated cycles versus hormone replacement treatment cycles for frozen embryo transfer in women with polycystic ovary syndrome: a prospective randomized controlled trial.

Purpose This study aimed to compare the pregnancy outcomes of letrozole (LE)-induced ovulation and hormone replacement treatment (HRT) in endometrial preparation for frozen embryo transfer (FET) in women with polycystic ovary syndrome (PCOS).

A randomized controlled trial involved 200 patients with PCOS from December 2017 to December 2022. Participants, who underwent FET with one or two good-quality blastocysts or cleavage-stage embryos, were randomly assigned to the LE group or HRT group in a 1:1 ratio. The primary outcome was the clinical pregnancy rate. Secondary outcomes included biochemical pregnancy, implantation, ectopic pregnancy, miscarriage, multiple pregnancy, and live birth rates.

The clinical pregnancy rate was 66.0% in the LE group compared to 53.0% in the HRT group (absolute difference, 13.0% [95% CI, − 0.5 to 26.5%]; RR, 1.25 [95% CI, 0.98 to 1.57]; P = 0.061). The biochemical pregnancy rate was higher in the LE group (71.0% vs 57.0%; absolute difference, 14.0% [95% CI, 0.8 to 27.2%]; RR, 1.25 [95% CI, 1.01 to 1.54]; P = 0.039). No significant differences were observed for the other secondary outcomes. The LE group showed higher biochemical pregnancy and live birth rates in the normal weight and normal androgen subgroups. Pregnancy outcomes were similar in the overweight and hyperandrogenic subgroups.

There were no statistically significant differences in clinical pregnancy rates between the LE and HRT cycles for FET in women with PCOS. However, the LE protocol may be a preferable option to HRT for women with normal weight or normal androgen levels. Clinical trial registration This study was registered at Chinese Clinical Trial Registry (http://www.chictr.org.cn/; ChiCTR-IOR-17014124). Supplementary Information The online version contains supplementary material available at 10.1007/s10815-025-03500-x.

Letrozole, Hormone replacement treatment, Frozen embryo transfer, Polycystic ovarian syndrome, Pregnancy outcomes

Polycystic ovarian syndrome (PCOS) is a prevalent endocrine disorder affecting approximately 10% of reproductive-age women, and is a major cause of anovulatory infertility. Due to the high ovarian reserve, women with PCOS are at higher cancellation rates of fresh embryo transfer cycles and higher rates of frozen embryo transfer cycles (FET) [1]. Advances in embryo cryopreservation and thawing techniques have demonstrated significantly higher live birth rates with FET compared to fresh embryo transfer in PCOS women [2]. Therefore, FET is recommended for PCOS women by International Evidence-based Guideline [3]. Selecting an appropriate endometrial preparation protocol is crucial for successful FET in PCOS women. Endometrial preparation protocols for FET include natural cycle (NC), hormone replacement treatment (HRT), and ovulation induction (OI). While current studies comparing these protocols have focused on women with regular ovulation, showing no significant differences in pregnancy outcomes [4, 5], research on women with PCOS is limited. Due to ovulation dysfunction in PCOS patients, HRT is commonly used for endometrial preparation. HRT requires prolonged estrogen use and a large amount of progesterone for luteal support. Additionally, HRT may increase pregnancy loss, obstetric complications, and perinatal outcomes due to the lack of corpus luteum [6, 7]. Therefore, it is essential to explore the optimal endometrial preparation protocol for FET in PCOS women. Letrozole (LE), a third-generation aromatase inhibitor, is the first-line treatment for ovulation induction in women with PCOS [3, 8]. LE has been shown to increase the expression of integrin αvβ3 and HOXA10 in the endometrium, which are essential genes for endometrial receptivity and implantation [9]. Additionally, LE induces a hypo-estrogenic and hyper-progestogenic state, potentially improving pregnancy outcomes [10]. A large retrospective cohort study including 110,722 single FET cycles found higher clinical pregnancy rate (CPR) and live birth rate (LBR), along with lower miscarriage rates in the LE group compared to the natural and HRT groups [11]. However, the use of LE for FET in PCOS women remains infrequent. A systematic review and meta-analysis, including eight retrospective studies and one randomized clinical trial (RCT) study, reported higher CPRs in the LE group compared to the HRT group [12]. However, two RCTs have shown no significant difference in CPRs and LBRs [13, 14]. Given these inconsistencies, this randomized controlled trial was conducted to investigate whether using LE for ovulation induction results in different clinical pregnancy rates compared to HRT for endometrial preparation in PCOS patients.

Patients This trial involved infertile women with PCOS who underwent IVF/ICSI and FET at Sun Yat-sen Memorial Hospital between December 2017 and December 2022. The study received approval from the Sun Yat-sen Memorial Hospital Ethics Committee (No. 2017[14]), and the trial was registered with the Chinese Clinical Trial Registry (http://www.chictr.org.cn/; ChiCTR-IOR-17014124). Written informed consent was obtained from all participants. The study followed the CONSORT (Consolidated Standards of Reporting Trials) and IMPRINT (Improving the Reporting of Clinical Trials of Infertility Treatments) guidelines [15]. Eligible participants were those undergoing FET with one or two good-quality blastocysts or cleavage embryos, aged 20 to 40 years, and diagnosed with PCOS based on the Rotterdam criteria [16]. Exclusion criteria included uterine factors, history of endometrium thickness lower than 7 mm, severe endometriosis, hydrosalpinx, immunologic disorders, candidates for preimplantation genetic diagnosis, history of embryo implantation failure at least twice, history with recurrent miscarriage, and severe disease not suitable for embryo transfer. Each participant was limited to a single FET cycle throughout the study period. Patients were classified into different PCOS sub-phenotypes according to their body mass index (BMI) and the status of hyperandrogenism (HA). Overweight was defined for Chinese adults by a BMI ≥ 24 kg/m2 according to the established guidelines [17]. Biologic HA was evaluated through androgen assay, with total testosterone (TT) level > 2.6 nmol/L. Randomization Participants meeting the criteria were randomly assigned to the intervention group (LE group) or the control group (HRT group) in a 1:1 ratio according to a computer-generated randomization list. Clinical procedures In the LE group, letrozole (Jiangsu Hengrui Medicine Co., China) was administered orally at a dose of 5 mg/day from the third to fifth day of the menstrual cycle for 5 days. Then ultrasound was done in the 10th to 12th day to check the follicle development and endometrial thickness. If the dominant follicle with ≥ 12 mm diameter was observed, no other medicine was given. If there was no dominant follicle, then human menopausal gonadotropin (HMG) (Lizhu Pharmaceutical Trading Co., China) was injected at dose of 37.5-75 IU/day. When the dominant follicle reached 14 mm diameter or more, the urine LH was measured daily. When the LH surge happened or the dominant follicle ≥ 20 mm with endometrium thickness ≥ 8 mm but no spontaneous LH surge was detected, recombinant human chorionic gonadotropin (hCG) 250 µg (Merck Serono, Germany) was injected. Micronized vaginal progesterone (Utrogestan, BESINS, Belgium) was administered 400 mg once daily plus oral dydrogesterone supplementation (Duphaston, Abbott Healthcare, Netherlands) 10 mg three times per day from the day of ovulation. The cycle was cancelled if the endometrial thickness was less than 8 mm on the day of spontaneous LH surge or the dominant follicle ≥ 20 mm. Additionally, if no dominant follicle was observed after 10 days of HMG injection, or if more than three dominant follicles developed, the cycle was also cancelled. In the HRT group, on the third day of menstrual cycle, patients received oral estradiol valerate (Progynova, Delpharm Lille, France) at a dose of 2 mg twice daily for 7 days, then ultrasound monitor was used to check the endometrium thickness and follicle development. The dose of estradiol valerate was adjusted depending on the endometrium thickness until to 9 mg daily. After using estradiol at least 10 days, when the endometrium thickness ≥ 8 mm, with estradiol ≥ 150 pg/L and P ≤ 1.5 pg/mL, then micronized vaginal progesterone was administered 400 mg twice daily plus oral dydrogesterone supplementation 20 mg two times per day. After using estradiol valerate for 20 days, if the endometrial thickness was less than 8 mm, the cycle was cancelled. Embryos at the cleavage stage were evaluated using the Istanbul consensus criteria [18]. The quality of blastocyst was assessed following the Gardner and School craft grading system [19]. Single blastocyst transfer was prioritized. Embryo transfer was performed for the blastocyst stage on the sixth day of progesterone administration or for the cleavage stage on the fourth day of progesterone administration. About 14 days post-embryo transfer, serum β-hCG levels were checked to confirm pregnancy. In cases of pregnancy, luteal phase support was maintained until the 10 th week of gestation in both groups. Outcomes measures The primary outcome was clinical pregnancy rate, identified by the presence of a gestational sac confirmed by ultrasound at 5 weeks after embryo transfer. The secondary outcomes consist of biochemical pregnancy, implantation, ectopic pregnancy, miscarriage, multiple pregnancy, and live birth. Biochemical pregnancy was defined as serum β-hCG > 25 mIU/mL. Live birth was followed up by telephone, defined as the delivery of a live neonate after 28 weeks of gestation. Statistical analysis Based on the clinical pregnancy rates observed in a previous retrospective study (65.0% in the LE group and 40.8% in the HRT group) [20], we conducted a power analysis using SAS 9.4 to determine the appropriate sample size for comparing these rates. Assuming a statistical power of 80% and a two-sided significance level of 5%, we calculated a required sample size of 63 patients per group. Accounting for a 20% dropout rate, we planned to recruit a total of 158 women. To further mitigate potential bias and account for variability, we expanded the sample sizes to 100 patients per group. Demographics were calculated as mean ± standard deviation (SD), median (interquartile range [IQR]), or number (%) of all patients at baseline and according to treatment group. Recruitment numbers, participants lost to follow-up, and other relevant data were reported. Comparisons between groups were made using suitable statistical tests, including the independent sample t-test and Wilcoxon rank-sum test for continuous variables, and the Pearson chi-square test or Fisher’s exact test for categorical variables. Absolute rate differences and relative risks (RR) with their respective 95% confidence intervals (CIs) were calculated. Statistical analyses were performed using SPSS (version 25.0, IBM Corp., Armonk, NY, USA), with statistical significance defined as P < 0.05 for two-sided tests. Primary and secondary outcomes were analyzed using both intention-to-treat and per protocol approaches, and exploratory subgroup analyses were performed using per protocol approaches.

Study participants The flow chart of study subjects, according to the CONSORT guidelines, is shown in Fig. 1. A total of 220 infertile PCOS patients were recruited. Of these, 20 patients were excluded based on the criteria, and 200 patients were randomly assigned to either the LE group (n = 100) or HRT group (n = 100). In the LE group, 3 patients canceled ET for personal reasons and 1 due to thin endometrium thickness. In the HRT group, 5 patients canceled ET for personal reasons and 2 due to thin endometrium thickness. No patients were lost to follow-up, and there were no missing outcome data. Baseline participant characteristics using intention-to-treat analysis are summarized in Table 1. Of the 100 total cycles in the LE group, 25 (25%) required adjunct HMG due to insufficient ovarian response. Table 1. | Characteristics | LE group (n = 100) | HRT group (n = 100) | |---|---|---| | Age at OS (y) | 29.42 ± 3.41 | 29.72 ± 3.38 | | BMI (kg/m2) | 22.06 ± 2.92 | 22.08 ± 3.07 | | AMH (ng/mL) | 11.07 ± 4.88 | 10.87 ± 5.18 | | AFC | 30.97 ± 9.27 | 31.36 ± 11.1 | | FSH (IU/L) | 7.00 ± 1.85 | 7.31 ± 2.51 | | LH (IU/L) | 8.74 ± 5.03 | 8.57 ± 6.22 | | TT (nmol/L) | 1.83 ± 0.84 | 2.11 ± 1.31 | | Fasting glucose (mmol/L) | 5.15 ± 0.42 | 5.19 ± 1.02 | | Fasting insulin (mu/L) | 11.82 ± 6.64 | 11.42 ± 7.80 | | Infertility years (y) | 4.48 ± 2.65 | 4.38 ± 2.50 | | Type of infertility | || | Primary, No. (%) | 71 (71.0) | 74 (74.0) | | Secondary, No. (%) | 29 (29.0) | 26 (26.0) | | Os protocol | || | Agonist, No. (%) | 29 (29.0) | 22 (22.0) | | Antagonist, No. (%) | 71 (71.0) | 78 (78.0) | | Type of ART | || | IVF, No. (%) | 67 (67.0) | 62 (62.0) | | ICSI, No. (%) | 8 (8.0) | 16 (16.0) | | IVF + ICSI, No. (%) | 25 (25.0) | 22 (22.0) | | No. of oocyte retrieved | 18.83 ± 7.20 | 20.06 ± 7.31 | | Oocyte maturation rate | 85.30 ± 12.27 | 83.77 ± 12.11 | | Fertilization rate | 60.96 ± 17.11 | 61.12 ± 15.51 | | Viable embryos | 6.76 ± 3.36 | 6.83 ± 2.89 | LE letrozole, HRT hormone replacement therapy, OS ovarian stimulation, BMI body mass index, AMH anti-müllerian, AFC antral follicles, FSH follicle-stimulating hormone, LH luteinizing hormone, TT total testosterone, IVF in vitro fertilization, ICSI intracytoplasmic sperm injection. Data are mean ± SD or No. % Primary and secondary outcomes In the intention-to-treat analysis (Table 2), the primary outcome, clinical pregnancy rate, was 66.0% (66/100) in the LE group and 53.0% (53/100) in the HRT group (absolute difference, 13.0% [95% CI, − 0.5 to 26.5%]; RR, 1.25 [95% CI, 0.98 to 1.57]; P = 0.061). The secondary outcome, biochemical pregnancy rate, was higher in the LE group than in the HRT group (71.0% vs 57.0%; absolute difference, 14.0% [95% CI, 0.8 to 27.2%]; RR, 1.25 [95% CI, 1.01 to 1.54]; P = 0.039). No significant differences were observed for the other secondary outcomes. The live birth rate was 54.0% (54/100) in the LE group, and 41.0% (41/100) in the HRT group (absolute difference, 13.0% [95% CI, − 0.7 to 26.7%]; RR, 1.32 [95% CI, 0.98 to 1.77]; P = 0.066). Table 2. | Outcomes | LE group (n = 100) | HRT group (n = 100) | Absolute difference (95%CI) % | Relative risk (95% CI) | P value | |---|---|---|---|---|---| | Primary outcome | ||||| | Clinical pregnancy rate, No. (%) | 66 (66.0) | 53 (53.0) | 13.0 (− 0.5 to 26.5) | 1.25 (0.98 to 1.57) | 0.061 | | Secondary outcomes | ||||| | Biochemical pregnancy rate, No. (%) | 71 (71.0) | 57 (57.0) | 14.0 (0.8 to 27.2) | 1.25 (1.01 to 1.54) | 0.039 | | Implantation rate, No. (%) | 74 (48.7) | 67 (45.6) | 3.1 (-8.2 to 14.4) | 1.07 (0.84 to 1.36) | 0.591 | | Ectopic pregnancy rate, No. (%) | 1 (1.5) | 1 (1.9) | − 0.4 (− 5.1 to 4.3) | 0.80 (0.05 to 12.54) | 1.000 | | Miscarriage rate, No. (%) | 11 (16.7) | 11 (20.8) | − 4.1 (− 18.3 to 10.1) | 0.80 (0.34 to 1.71) | 0.568 | | Multiple pregnancy rate, No. (%) | 7 (10.6) | 9 (17.0) | − 6.4 (− 18.9 to 6.1) | 0.63 (0.25 to 1.57) | 0.311 | | Live birth rate, No. (%) | 54 (54.0) | 41 (41.0) | 13.0 (-0.7 to 26.7) | 1.32 (0.98 to 1.77) | 0.066 | FET frozen embryo transfer, LE letrozole, HRT hormone replacement therapy, CI confidence interval. Data are mean ± SD or No. % The per-protocol analysis results (Supplemental Table 1) were similar to the intention-to-treat analysis, except that the biochemical pregnancy rates showed no significant differences between the two groups. Sixty-six of 96 women (68.8%) in the LE group and 53 of 93 women (57.0%) in the HRT group achieved clinical pregnancy (absolute difference, 11.8% [95% CI, − 1.9 to 25.5%]; RR, 1.21 [95% CI, 0.97 to 1.51]; P = 0.094). Exploratory subgroup analyses Exploratory subgroup analyses were performed based on BMI (Table 3) and androgen levels (Table 4). In the normal weight subgroup, the peak endometrium was thicker in the LE group than in the HRT group (9.92 mm vs 9.25 mm, P = 0.007). The rates of biochemical pregnancy and live birth were higher in the LE group (n = 72) than in the HRT group (n = 73) (77.8% vs 61.6%, P = 0.035; 62.5% vs 43.8%, P = 0.024). The rates of clinical pregnancy, implantation, ectopic pregnancy, miscarriage, and multiple pregnancy did not show significant differences. In the overweight subgroup, the pregnancy outcomes were similar between the groups. Table 3. | Characteristics and outcomes | Normal Weight (n=145) | Overweight (n=44) | |||| |---|---|---|---|---|---|---| | LE Group (n=72) | HRT Group (n=73) | P 1 value | LE Group (n=24) | HRT Group (n=20) | P 2 value | | | Age at FET (y) | 29.92 ± 3.72 | 30.16 ± 3.08 | 0.663 | 31.17 ± 4.01 | 30.45 ± 4.42 | 0.576 | | Peak endometrium thickness (mm) | 9.92 ± 1.79 | 9.25 ± 1.41 | 0.007 | 11.23 ± 2.46 | 9.78 ± 1.44 | 0.019 | | E2 of luteal transform day (pg/mL) | 296.76 ± 245.50 | 290.83 ± 132.94 | 0.858 | 205.74 ± 164.79 | 202.17 ± 128.46 | 0.940 | | No. of transferred embryos | |||||| | 1, No. (%) | 32 (44.4) | 28 (38.4) | 0.457 | 8 (33.3) | 11 (55.0) | 0.149 | | 2, No. (%) | 40 (55.6) | 45 (61.6) | 16 (66.7) | 9 (45.0) | || | Embryo Stage | |||||| | Cleavage, No. (%) | 37 (51.4) | 39 (53.4) | 0.806 | 17 (70.8) | 7 (35.0) | 0.017 | | Blastocyst, No. (%) | 35 (48.6) | 34 (46.6) | 7 (29.2) | 13 (65.0) | || | Clinical pregnancy rate, No. (%) | 52 (72.2) | 42 (57.5) | 0.064 | 14 (58.3) | 11 (55.0) | 0.824 | | Biochemical pregnancy rate, No. (%) | 56 (77.8) | 45 (61.6) | 0.035 | 15 (62.5) | 12 (60.0) | 0.865 | | Implantation rate, No. (%) | 57 (50.9) | 52 (44.1) | 0.300 | 17 (42.5) | 15 (51.7) | 0.448 | | Ectopic pregnancy rate, No. (%) | 1 (1.9) | 1 (2.4) | 1.000 | 0 | 0 | / | | Miscarriage rate, No. (%) | 6 (11.5) | 9 (21.4) | 0.193 | 5 (35.7) | 2 (18.2) | 0.407 | | Multiple pregnancy rate, No. (%) | 5 (9.6) | 7 (16.7) | 0.308 | 2 (14.3) | 2 (18.2) | 1.000 | | Live birth rate, No. (%) | 45 (62.5) | 32 (43.8) | 0.024 | 9 (37.5) | 9 (45.0) | 0.614 | BMI, body mass index; FET, frozen embryo transfer; E2, estradiol. Data are mean ± SD or No. % Table 4. | Characteristics and outcomes | Normal androgen (n = 150) | P 1 value | HA (n = 39) | P 2 value | || |---|---|---|---|---|---|---| | LE group (n = 78) | HRT group (n = 72) | LE group (n = 18) | HRT group (n = 21) | ||| | Age at FET (y) | 30.51 ± 3.79 | 30.35 ± 3.45 | 0.390 | 29.00 ± 3.76 | 29.81 ± 3.22 | 0.237 | | Peak endometrium thickness (mm) | 10.35 ± 2.05 | 9.35 ± 1.49 | < 0.001 | 9.79 ± 2.00 | 9.39 ± 1.21 | 0.222 | | E2 of luteal transform day (pg/mL) | 286.49 ± 240.64 | 265.66 ± 134.68 | 0.259 | 223.83 ± 181.55 | 295.14 ± 141.69 | 0.088 | | No. of transferred embryos | |||||| | 1, No. (%) | 36 (46.2) | 29 (40.3) | 0.468 | 4 (22.2) | 10 (47.6) | 0.099 | | 2, No. (%) | 42 (53.9) | 43 (59.7) | 14 (77.8) | 11 (52.4) | || | Embryo stage | |||||| | Cleavage, No. (%) | 40 (51.3) | 36 (50.0) | 0.875 | 14 (77.8) | 10 (47.6) | 0.054 | | Blastocyst, No. (%) | 38 (48.7) | 36 (50.0) | 4 (22.2) | 11 (52.4) | || | Clinical pregnancy rate, No. (%) | 57 (73.1) | 40 (55.6) | 0.025 | 9 (50.0) | 13 (61.9) | 0.455 | | Biochemical pregnancy rate, No. (%) | 60 (76.9) | 44 (61.1) | 0.036 | 11 (61.1) | 13 (61.9) | 0.959 | | Implantation rate, No. (%) | 64 (53.3) | 50 (43.5) | 0.131 | 10 (31.3) | 17 (53.1) | 0.076 | | Ectopic pregnancy rate, No. (%) | 1 (1.8) | 1 (2.5) | 1.000 | 0 | 0 | / | | Miscarriage rate, No. (%) | 10 (17.5) | 10 (25.0) | 0.372 | 1 (11.1) | 1 (7.7) | 1.000 | | Multiple pregnancy rate, No. (%) | 7 (12.3) | 6 (15.0) | 0.699 | 0 | 3 (23.1) | 0.240 | | Live birth rate, No. (%) | 46 (59.0) | 29 (40.3) | 0.022 | 8 (44.4) | 12 (57.1) | 0.429 | HA hyperandrogenism, FET frozen embryo transfer; E2 estradiol. Data are mean ± SD or No. % In the normal androgen subgroup, the LE group showed higher peak endometrium thickness than the HRT group (10.35 mm vs 9.35 mm, P < 0.001). The rates of clinical pregnancy, biochemical pregnancy, and live birth were higher in the LE group than in the HRT group (73.1% vs 55.6%, P = 0.025; 76.9% vs 61.1%, P = 0.036; 59.0% vs 40.3%, P = 0.022). However, no significant differences were found in the pregnancy outcomes between groups among the HA patients.

This single center randomized controlled trial suggested that there was no statistically significant difference in the clinical pregnancy rates between LE application cycles and HRT cycles in FET for patients with PCOS. However, in the normal weight subgroup, the LE application cycles showed higher rates of biochemical pregnancy and live birth. Additionally, in the normal androgen subgroup, clinical pregnancy rates, biochemical pregnancy rates, and live birth rates were higher in the LE cycles. LE reduces the production of estrogen, which increases FSH secretion, and leads to mono-ovulatory cycles [21]. Furthermore, LE has a short half-life of about 48 h, potentially sparing estrogen-target tissues from adverse effects [22]. Therefore, several studies have evaluated the efficacy of using LE for endometrial preparation before FET in patients with PCOS. Four retrospective studies [20, 23–25] and two randomized clinical trials [13, 14] have investigated this topic. The retrospective studies all reported higher pregnancy rates in cycles stimulated with LE compared to HRT cycles. However, a RCT involving 116 PCOS patients found no significant differences in clinical pregnancy rates, miscarriage rates, and ongoing pregnancy rates between LE combined with HMG and HRT for FET [14]. This trial had a smaller sample size and did not report potential confounding factors such as embryo quality and freezing practices. Another recent RCT study compared live birth rates between LE application and HRT cycle for endometrium preparation during FET in a large sample of PCOS patients [13]. The findings showed no difference in pregnancy outcomes between the two groups. However, the inclusion criteria required at least 1 day 3 embryos or 1 day 5 blastocyst, with good-quality embryo ratios of 63.7% and 67.8% in the two groups. Our study controlled for embryo quality by requiring at least one good-quality embryo for FET. Additionally, the previous RCT study only included patients with PCOS with menstrual cycles longer than 35 days, limiting the generalizability of the results to PCOS patients with oligomenorrhea. To address this limitation, we included PCOS patients identified using Rotterdam criteria with various phenotypes. In our RCT study, we found that the biochemical pregnancy rates were higher in the LE group than in the HRT group (71.0% vs 57.0%, P = 0.039). Although there was a trend towards higher clinical pregnancy rates in the LE group compared to the HRT group, this difference was not statistically significant (66.0% vs 53%, P = 0.061). Consequently, we performed exploratory subgroup analyses to determine whether specific subgroups benefited from LE application. Past research suggests that using LE for endometrial preparation may offer advantages. Several potential mechanisms for these benefits have been proposed. Firstly, LE inhibits the conversion of androgens to estrogens in granulosa cells and resulted the reduced estrogen levels [11]. Subsequently, the decreased estrogen levels reduce the ubiquitination of estrogen receptors, promoting faster proliferation of the endometrium and increased uterine blood flow, positively impacting pregnancy outcomes [20]. What is more, evidence suggests that LE may improve endometrial receptivity [24]. The lack of endometrial αvβ3 integrin expression is associated with adverse outcomes in IVF cycles, and this may be improved with LE adjunctive therapy [26]. Additionally, women with unexplained infertility undergoing ovulation stimulation with LE showed significantly increased expression of endometrial receptivity markers, including integrin, leukemia inhibitory factor, and L-selectin, compared to natural cycles [27]. In women with PCOS, endometrial ultrasonic parameters, integrin αvβ3, and VEGF concentrations during the implantation window were significantly higher in the LE group compared with the clomiphene citrate group and natural cycle group [28]. Collectively, these studies indicate that LE could potentially have a positive impact on endometrial receptivity and may improve the pregnancy outcomes. We observed a slight increase in endometrial thickness in the LE group. Similarly, studies by Hu et al. [20], Zhang et al. [24], and Yuan et al. [13] reported that endometrial thickness on the day of initiating progesterone treatment or on the day of ET was significantly greater compared to the HRT group. However, some studies reported no significant differences in endometrial thickness between the LE and HRT groups [23, 29, 30]. Therefore, further large-sample randomized trials are still needed to draw definitive conclusions and insights regarding this matter. Additionally, the potential for natural conception during LE-induced ovulation cycles should not be overlooked. Letrozole stimulates mono-ovulation, which may synchronize endometrial development with natural hormonal fluctuations, potentially enhancing embryo-endometrium dialogue. This physiological alignment, combined with the possibility of spontaneous conception during the treatment window, might partially contribute to the higher pregnancy rates in the LE group. Exploratory subgroup analyses based on BMI showed higher biochemical pregnancy rates and live birth rates in the LE group compared to the HRT group in the normal weight subgroup. Our findings partially align with a previous large prospective study [23]. In the normal androgen subgroup, the rates of clinical pregnancy, biochemical pregnancy, and live birth were also higher in the LE group. However, the pregnancy outcomes were similar in the overweight subgroup and the HA subgroup. In the overweight subgroup, metabolic abnormalities may play a major role in affecting pregnancy outcomes [31]. In the HA subgroup, the effect of letrozole on increasing androgens may affect pregnancy outcomes [32, 33]. Due to the limited sample size, the subgroup results are exploratory only and require validation in larger-scale studies. Beyond pregnancy outcomes, the endometrial preparation protocol may have implications for obstetric outcomes. Emerging evidence suggests that FET cycles with ovulation induction are associated with a lower risk of pregnancy-induced hypertension compared to artificial cycles[34, 35], likely due to the presence of a corpus luteum and its vasoactive mediators. This potential benefit further supports the consideration of LE protocols in clinical practice, particularly for patients at high risk of vascular complications. Although our study observed two cases of hypertensive disorders in the HRT group versus none in the LE group, the small sample size precludes definitive conclusions regarding group differences. Our study has several strengths. First, we included the patients with good embryo quality to control the embryo factor. Second, we conducted exploratory subgroup analyses and found some interesting results. However, there were limitations to our study. First, the data came from only one assisted reproductive center; thus, heterogeneity among centers has not been estimated. Second, we calculate the sample size for between-group differences based on a previous study’s primary outcome (clinical pregnancy rate: 65.0% in the LE group and 40.8% in the HRT group) from a different reproductive center [20]. However, our center had a higher CPR in the HRT group (53%), which might be due to optimal estradiol in the HRT procedure [36] compared to the previous study. Therefore, a larger sample size is needed in future studies. Third, we only performed subgroup analyses based on BMI and androgen levels and did not analyze PCOS phenotypes due to the sample size. Fourth, we did not detect the endometrium receptive biomarkers, so the role of LE on endometrium is unknown. In conclusion, this randomized clinical trial found no statistically significant differences in clinical pregnancy rates and live birth rates between the LE application and HRT cycles for FET in women with PCOS. The LE protocol may be a superior option for those with normal weight or normal androgen levels. Future RCTs with larger sample sizes and mechanistic studies are needed to validate our results. Supplementary Information Below is the link to the electronic supplementary material. Author contribution LLJ contributed to data analysis and original draft writing; JH contributed to conceptualization, planning, and patient treatment; YXX contributed to randomization and patient treatment; LL contributed to conceptualization and patient treatment; PP contributed to patient treatment; DZY contributed to conceptualization and review of manuscript; RYZ contributed to data acquisition; YL contributed to conceptualization, planning, data analysis, supervision, and review of manuscript. Funding Grant support was provided by Basic and Applied Basic Research Foundation of Guangdong Province (2021 A1515010299), Fertility Research Program of Young and Middle-aged Physicians IN 2023 (BJHPA-2023-SZHYXZHQN-002), and Sun Yat-sen Memorial Hospital Yixian Scientific Research Launch Funding (SYSQH-II-2024–06). Data availability The data that support the findings of this study are available from the corresponding author upon reasonable request. Declarations Ethics approval The trial protocol was approved by the Ethics Committee of Sun Yat-sen Memorial Hospital (No. 2017[14] and was registered at Chinese Clinical Trial Registry (http://www.chictr.org.cn/; ChiCTR-IOR-17014124). Consent to participate Written informed consent was obtained from all participants. Competing interests The authors declare no competing interests. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

- 1.Homburg R. Management of infertility and prevention of ovarian hyperstimulation in women with polycystic ovary syndrome. Best Pract Res Clin Obstet Gynaecol. 2004;18(5):773–88. 10.1016/j.bpobgyn.2004.05.006. [DOI] [PubMed] [Google Scholar] - 2.Chen ZJ, Shi Y, Sun Y, Zhang B, Liang X, Cao Y, et al. Fresh versus frozen embryos for infertility in the polycystic ovary syndrome. N Engl J Med. 2016;375:523–33. [DOI] [PubMed] [Google Scholar] - 3.Teede HJ, Tay CT, Laven JJE, Dokras A, Moran LJ, Piltonen TT, et al. Recommendations from the 2023 international evidence-based guideline for the assessment and management of polycystic ovary syndrome. J Clin Endocrinol Metab. 2023;2023:2447–69. [DOI] [PMC free article] [PubMed] [Google Scholar] - 4.Yarali H, Polat M, Mumusoglu S, Yarali I, Bozdag G. Preparation of endometrium for frozen embryo replacement cycles: a systematic review and meta-analysis. J Assist Reprod Genet. 2016;33:1287–304. [DOI] [PMC free article] [PubMed] [Google Scholar] - 5.Konc J, Kanyo K, Varga E, Kriston R, Cseh S. The effect of cycle regimen used for endometrium preparation on the outcome of day 3 frozen embryo transfer cycle. Fertil Steril. 2010;94:767–8. [DOI] [PubMed] [Google Scholar] - 6.Zhang Y, Fu X, Gao S, Gao S, Gao S, Ma J, et al. Letrozole use in vitrified single-blastocyst transfer cycles is associated with lower risk of large for gestational age infants in patients with polycystic ovary syndrome. J Assist Reprod Genet. 2023;40:2885–94. [DOI] [PMC free article] [PubMed] [Google Scholar] - 7.Wu H, Zhou P, Lin X, Wang S, Zhang S. Endometrial preparation for frozen-thawed embryo transfer cycles: a systematic review and network meta-analysis. J Assist Reprod Genet. 2021;38:1913–26. [DOI] [PMC free article] [PubMed] [Google Scholar] - 8.Mitwally MF, Casper RF. Use of an aromatase inhibitor for induction of ovulation in patients with an inadequate response to clomiphene citrate. Fertil Steril. 2001;75:305–9. [DOI] [PubMed] [Google Scholar] - 9.Bao SH, Sheng SL, Peng YF, Lin QD. Effects of letrozole and clomiphene citrate on the expression of HOXA10 and integrin alpha v beta 3 in uterine epithelium of rats. Fertil Steril. 2009;91:244–8. [DOI] [PubMed] [Google Scholar] - 10.Legro RS, Brzyski RG, Diamond MP, Coutifaris C, Schlaff WD, Casson P, et al. Letrozole versus clomiphene for infertility in the polycystic ovary syndrome. N Engl J Med. 2014;37:119–29. [DOI] [PMC free article] [PubMed] [Google Scholar] - 11.Tatsumi T, Jwa SC, Kuwahara A, Irahara M, Kubota T, Saito H. Pregnancy and neonatal outcomes following letrozole use in frozen-thawed single embryo transfer cycles. Hum Reprod. 2017;32:1244–8. [DOI] [PubMed] [Google Scholar] - 12.Zhang Y, Wu L, Li TC, Wang CC, Zhang T, Chung JPW. Systematic review update and meta-analysis of randomized and non-randomized controlled trials of ovarian stimulation versus artificial cycle for endometrial preparation prior to frozen embryo transfer in women with polycystic ovary syndrome. Reprod Biol Endocrinol. 2022;20:62. [DOI] [PMC free article] [PubMed] [Google Scholar] - 13.Yuan Y, Chang Q, Wen Y, Gao J, Huang S, Xu Y, et al. Letrozole during frozen embryo transfer in women with polycystic ovarian syndrome: a randomized controlled trial. Obstet Gynecol. 2023;142:1087–95. [DOI] [PubMed] [Google Scholar] - 14.Hosseini-Najarkolaei A, Moini A, Kashani L, FaridMojtahedi M, Hosseini-Najarkolaee E, Salehi E. The effect of letrozole versus artificial hormonal endometrial preparation on pregnancy outcome after frozen-thawed embryos transfer cycles: a randomized clinical trial. Reprod Biol Endocrinol. 2020;18:115. [DOI] [PMC free article] [PubMed] [Google Scholar] - 15.Harbin Consensus Conference Workshop G, Conference C, Legro RS, Wu X, Scientific C, Barnhart KT, et al. Improving the reporting of clinical trials of infertility treatments (IMPRINT): modifying the CONSORT†‡. Hum Reprod. 2014;2014(102):2075–82. [DOI] [PMC free article] [PubMed] [Google Scholar] - 16.Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop group. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome (PCOS). Hum Reprod. 2004;19(1):41–7. 10.1093/humrep/deh098. [DOI] [PubMed] [Google Scholar] - 17.Wang H, Zhai F. Programme and policy options for preventing obesity in China. Obes Rev. 2013;14:134–40. [DOI] [PMC free article] [PubMed] [Google Scholar] - 18.Alpha Scientists in Reproductive Medicine and ESHRE Special Interest Group of Embryology. The Istanbul consensus workshop on embryo assessment: proceedings of an expert meeting. Hum Reprod. 2011;26(6):1270–83. 10.1093/humrep/der037. [DOI] [PubMed] - 19.Schoolcraft WB, Gardner DK, Lane M, Schlenker T, Hamilton F, Meldrum DR. Blastocyst culture and transfer: analysis of results and parameters affecting outcome in two in vitro fertilization programs. Fert Ster. 1999;72(4):604–9. [DOI] [PubMed] [Google Scholar] - 20.Hu YJ, Chen YZ, Zhu YM, Huang HF. Letrozole stimulation in endometrial preparation for cryopreserved–thawed embryo transfer in women with polycystic ovarian syndrome: a pilot study. Clin Endocrinol. 2014;80(2):283–9. [DOI] [PubMed] [Google Scholar] - 21.Palomba S. Aromatase inhibitors for ovulation induction. J Clin Endocrinol Metab. 2015;100(5):1742–7. 10.1210/jc.2014-4235. [DOI] [PubMed] [Google Scholar] - 22.Jirge PR, Patil RS. Comparison of endocrine and ultrasound profiles during ovulation induction with clomiphene citrate and letrozole in ovulatory volunteer women. Fert Ster. 2010;93(1):174–83. [DOI] [PubMed] [Google Scholar] - 23.Wang X, Li Y, Tan H, Cai S, Ma S, Peng Y, Guo H, Li X, Tang Y, Zhang S, Lin G. Letrozole-stimulated endometrial preparation protocol is a superior alternative to hormone replacement treatment for frozen embryo transfer in women with polycystic ovary syndrome, a cohort study. Reprod Biol Endocrinol. 2023;21(1):101. [DOI] [PMC free article] [PubMed] [Google Scholar] - 24.Zhang J, Liu H, Wang Y, Mao X, Chen Q, Fan Y, et al. Letrozole use during frozen embryo transfer cycles in women with polycystic ovary syndrome. Fert Ster. 2019;112(2):371–7. [DOI] [PubMed] [Google Scholar] - 25.Aslih N, Dorzia D, Atzmon Y, Estrada D, Ellenbogen A, Bilgory A, et al. Ovulatory-based FET cycles may achieve higher pregnancy rates in the general population and among anovulatory women [Internet]. J Clin Med. 2021. Available from: https://www.ncbi.nlm.nih.gov/pubmed/33670133 [DOI] [PMC free article] [PubMed] - 26.Miller PB, Parnell BA, Bushnell G, Tallman N, Forstein DA, Higdon HL 3rd, et al. Endometrial receptivity defects during IVF cycles with and without letrozole. Hum Reprod. 2012;27:881–8. [DOI] [PMC free article] [PubMed] [Google Scholar] - 27.Ganesh A, Chauhan N, Das S, Chakravarty B, Chaudhury K. Endometrial receptivity markers in infertile women stimulated with letrozole compared with clomiphene citrate and natural cycles. Syst Biol Reprod Med. 2014;60(2):105–11. [DOI] [PubMed] [Google Scholar] - 28.Wang L, Lv S, Li F, Bai E, Yang X. Letrozole versus clomiphene citrate and natural cycle: endometrial receptivity during implantation window in women with polycystic ovary syndrome. Front Endocrinol (Lausanne). 2020;11:532692. [DOI] [PMC free article] [PubMed] [Google Scholar] - 29.Peigné M, Devouche E, Ferraretto X, Gricourt S, Luton D, Patrat C, Epelboin S. Higher live birth rate with stimulated rather than artificial cycle for frozen-thawed embryo transfer. Euro J Obstetr Gynecol Reprod Biol. 2019;243:144–9. [DOI] [PubMed] [Google Scholar] - 30.Aleyasin A, Aghahosseini M, Safdarian L, Noorzadeh M, Fallahi P, Rezaeian Z, et al. Can letrozole plus HMG protocol improve pregnancy outcomes in frozen-thawed embryo transfer? An RCT. Int J Reprod Biomed. 2017;15:83–6. [PMC free article] [PubMed] [Google Scholar] - 31.Si M, Xu W, Qi X, Jiang H, Zhao Y, Li R, et al. Metabolic syndrome rather than other phenotypes in PCOS as a predictive indicator for clinical outcomes in IVF: comprehensive phenotypic assessment across all PCOS classifications. J Clin Med. 2023;12(15):5073 (https://www.ncbi.nlm.nih.gov/pubmed/37568475). [DOI] [PMC free article] [PubMed] [Google Scholar] - 32.Zhao W, Li Z, Cai B, Zhou C, Mai Q. Impact of dehydroepiandrosterone sulfate and free androgen index on pregnancy and neonatal outcomes in PCOS patients. Reprod Biol Endocrinol. 2024;22(1):43. [DOI] [PMC free article] [PubMed] [Google Scholar] - 33.Bülow NS, Skouby SO, Warzecha AK, Udengaard H, Andersen CY, Holt MD, et al. Impact of letrozole co-treatment during ovarian stimulation with gonadotrophins for IVF: a multicentre, randomized, double-blinded placebo-controlled trial. Human Reprod. 2022;37(2):309–21. [DOI] [PubMed] [Google Scholar] - 34.Saito K, Kuwahara A, Ishikawa T, Morisaki N, Miyado M, Miyado K, et al. Endometrial preparation methods for frozen-thawed embryo transfer are associated with altered risks of hypertensive disorders of pregnancy, placenta accreta, and gestational diabetes mellitus. Hum Reprod. 2019;34:1567–75. [DOI] [PubMed] [Google Scholar] - 35.Zhang J, Wei M, Bian X, Wu L, Zhang S, Mao X, Wang B. Letrozole-induced frozen embryo transfer cycles are associated with a lower risk of hypertensive disorders of pregnancy among women with polycystic ovary syndrome. Am J Obstetr Gynecol. 2021;225(1):59-e1. [DOI] [PubMed] [Google Scholar] - 36.Alsbjerg B, Jensen MB, Elbaek HO, Laursen R, Povlsen BB, Anderson R, et al. Midluteal serum estradiol levels are associated with live birth rates in hormone replacement therapy frozen embryo transfer cycles: a cohort study. Fertil Steril. 2024;121:1000–9. [DOI] [PubMed] [Google Scholar] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request.