Is there an effect of the start day of luteal phase support on pregnancy outcomes in natural vitrified-warmed blastocyst transfer? Authors: Batuhan Turgay, Batuhan Özmen, Cem Somer Atabekoğlu, Yavuz Emre Şükür, Murat Sönmezer, Bülent Berker, Ruşen Aytaç Published: 2026 ABSTRACT AbstractObjectiveTo evaluate whether the timing of luteal phase support (LPS) initiation affects pregnancy outcomes in natural cycle vitrified-warmed blastocyst transfer (NC-FET) cycles.MethodsThis retrospective cohort study included NC-FET cycles performed between January 2022 and December 2024 at a tertiary university fertility center. Only true natural cycles with blastocyst transfer were analyzed. Patients were divided into two groups according to the initiation of LPS: two days before embryo transfer (Group 1) or on the day of embryo transfer (Group 2). All patients received vaginal and subcutaneous progesterone for LPS. Baseline characteristics, embryo features, serum progesterone levels on transfer day, and reproductive outcomes were compared. Multivariate logistic regression and receiver operating characteristic (ROC) analyses were performed to identify factors associated with clinical pregnancy.ResultsA total of 246 NC-FET cycles were analyzed (Group 1: n = 172; Group 2: n = 74). Baseline characteristics, embryo quality, and transfer parameters were comparable between groups. Serum progesterone levels on the day of transfer were significantly higher in Group 1 (32.1 ± 13.0 ng/mL) compared with Group 2 (10.3 ± 3.8 ng/mL; p = 0.001). However, total pregnancy rate (66.2% vs. 47.1%, p = 0.006), clinical pregnancy rate (56.8% vs. 40.7%, p = 0.020), and live birth rate (52.7% vs. 37.2%, p = 0.024) were significantly higher in Group 2. Logistic regression analysis showed that initiation of LPS on the day of transfer was independently associated with increased clinical pregnancy (OR 2.37, 95% CI 1.17–4.81). Serum progesterone level did not predict pregnancy outcomes.ConclusionIn NC-FET cycles, initiating luteal phase support on the day of embryo transfer is associated with improved pregnancy and live birth rates compared with earlier initiation, despite lower serum progesterone levels.Keywords: Natural cycle, Frozen embryo transfer, Luteal phase support, Progesterone, Pregnancy outcome Introduction Frozen embryo transfer (FET) has transformed the field of assisted reproductive technology (ART), providing the prevention of ovarian hyperstimulation syndrome, preimplantation genetic tests and more physiologically compatible method for embryo implantation [1]. The ideal FET protocol is controversial. Among the diverse FET methodologies, natural cycle FET (NC-FET) is especially preferred by physicians for patients with regular ovulatory cycles. In contrast to hormone replacement treatment (HRT) cycles, NC-FET employs the patient’s intrinsic hormonal environment, thereby diminishing the necessity for external drugs and consequently lowering the cost and potential adverse effects [2]. Luteal phase support (LPS) has been well established in HRT and stimulated cycles. LPS is essential in HRT-FET cycles due to the absence of endogenous progesterone production [3]. In NC-FET, the corpus luteum is present, so theoretically, exogenous progesterone may not be necessary. However, in some cycles, if sufficient progesterone is not released from the corpus luteum, implantation may be disrupted so pregnancy rate may be reduced. In the latest meta-analysis on this topic, it has been observed that progesterone supplementation in NC-FETs improves pregnancy outcomes [4]. The use of exogenous progesterone rescue in addition to luteal phase support was also evaluated, and although there was generally no positive effect, it was shown that patients with progesterone levels above 7–10 ng/ml had better pregnancy outcomes [5,6]. In this study, we aimed to compare the effects of starting luteal phase support two days before transfer without measuring progesterone levels and starting it on the day of transfer on pregnancy outcomes in NC-FET cycles. Materials methods This retrospective study was conducted at the Department of Reproductive Endocrinology, Ankara University Faculty of Medicine, Department of Obstetrics and Gynecology. IVF cycles performed between January 2022 and December 2024 were evaluated. Fresh transfer cycles, modified natural cycles, artificial endometrial preparation protocols and day 3 embryo transfers were excluded. Among these, only NC-FET cycles were included in the analysis. The analysis was restricted to true natural cycles, defined as cycles without any exogenous hormonal endometrial preparation except for luteal phase support, in which spontaneous ovulation occurred without any exogenous trigger agent for endometrial preparation (Fig. 1). The first cycle of each patient was included in the study. Flowchart of the study Endometrial preparation procedure On the second or third day of menstruation, transvaginal ultrasonography was done to rule out any cysts that might have developed during the previous cycle. Cycle cancellation was performed in cycles with serum P4 levels over 1.5 ng/mL on days 2 or 3 of menstruation. Daily transvaginal ultrasonography and endocrine monitoring (E2, LH, and P4 measures) were carried out once the leading follicle reached a mean diameter of 15–16 mm. Follicle collapse or follicle volume reduction and/or the presence of pelvic fluide were used to check the day of ovulation. LH surge and P4 level (> 1.5 ng/ml) 2–3 days after the LH surge were used to confirm the ovulation [7]. Five days following follicular failure, a warmed blastocyst transfer was planned. Luteal phase support (LPS) was administered either on the day of embryo transfer or two days before transfer, reflecting two distinct clinical management approaches applied at different time periods within the same center. The allocation to each group was not guided by individual patient characteristics such as ovulatory dynamics, endometrial development, or hormonal profile. Patients were divided into two groups according to whether luteal phase support was started two days before transfer or on the day of transfer. Serum progesterone levels were measured on the transfer day. The progesterone concentration was assessed using the Progesterone Elisa Kit (Neogen Corp, USA), with a maximum reportable value of 40 ng/mL. Demographic data, including age, body mass index (BMI), duration and etiology of infertility (unexplained, male factor, female factor), were recorded. All IVF stimulation cycles were conducted using recombinant FSH (Gonal-F®, Merck Serono, Switzerland) and/or highly purified human menopausal gonadotropin (hMG) (Menopur®, Ferring Pharmaceuticals, Switzerland). Two ovarian stimulation protocols were used: GnRH antagonist and progestin-primed ovarian stimulation (PPOS). Final oocyte maturation ovulation triggering was achieved using GnRH agonist alone or dual trigger methods based on follicular response and clinician preference. Embryos were classified based on Gardner criteria, taking into account the categories of ‘good’ ‘fair’ and ‘poor’.All embryo transfers were performed with a soft embryo transfer catheter (Cook® Medical, USA) under abdominal ultrasound guidance by experienced embryologists. Luteal support included vaginal micronized progesterone (Progestan®, Koçak Farma, 200 mg twice daily) and subcutaneous progesterone (Prolutex®, IBSA Institut Biochimique SA, Switzerland, 25 mg once daily). Embryo characteristics (day, quality), endometrial thickness, and hormonal markers (basal FSH, LH, E2, AMH) were documented. Clinical outcomes including pregnancy, clinical pregnancy, live birth, and miscarriage rates were compared between the two groups. Ten days following the ET, the serum hCG level was measured to confirm pregnancy; a level of > 15 IU was considered a positive result and the clinical pregnancy was determined as the presence of gestational sac. Study approval was obtained from the Ankara University Faculty of Medicine Department of Obstetrics and Gynecology Academic Board (Approval No: İ25-98745), and all procedures were conducted in accordance with the Declaration of Helsinki. Statistical analysis Statistical analysis was conducted using IBM SPSS Statistics for Windows, Version 25.0 (IBM Corp., Armonk, NY, USA). Normality of distribution was assessed using the Kolmogorov–Smirnov test. Continuous variables were presented as mean ± standard deviation (SD) for normally distributed data and as median with interquartile range (IQR) for non-normally distributed data. Categorical variables were summarized as frequencies and percentages. Comparisons between groups were made using the independent samples t-test for normally distributed variables, and the Mann–Whitney U test for skewed distributions. The Chi-square test or Fisher’s exact test was used for categorical comparisons. A multivariate logistic regression analysis was performed to evaluate the independent effects of endometrial thickness, serum progesterone level, and embryo quality on clinical pregnancy outcomes. All variables were entered simultaneously into the model, and odds ratios (Exp(B)) with 95% confidence intervals were calculated. Cycles involving double embryo transfer were excluded from the logistic regression analysis to avoid potential confounding. A p-value < 0.05 was considered statistically significant. Results Luteal phase support was started two days before transfer for patients in group 1, while it was started on the day of transfer for patients in group 2. Baseline characteristics and treatment parameters of both groups were similar (Table 1). Only progesterone level on the transfer day was statistically different between the groups. The mean serum progesterone level on the day of transfer was 32.1 ± 13.0 ng/mL in Group 1 and 10.3 ± 3.8 ng/mL in Group 2 (p = 0.001). Table 1Baseline characteristics and treatment parametersVariableGroup 1 (n = 172)Group 2 (n = 74)p-valueAge (years), mean ± SD32.4 ± 4.631.9 ± 4.80.48BMI (kg/m²), mean ± SD24.1 ± 2.923.8 ± 3.10.55Duration of infertility (month), median (IQR)34 (22–40)36 (23–42)0.62AMH (ng/mL), mean ± SD3.1 ± 1.23.0 ± 1.30.51Basal FSH (mIU/mL), mean ± SD6.2 ± 1.86.4 ± 1.70.43Basal E2 (pg/mL), mean ± SD45 ± 1247 ± 110.39Basal LH (mIU/mL), mean ± SD4.8 ± 1.64.7 ± 1.50.67Previous live birthYes31 (18.0%)15 (20.3%)0.67Stimulation duration (days), mean ± SD9.4 ± 1.89.6 ± 1.90.44Total gonadotropin dose (IU), mean ± SD2350 ± 5402280 ± 5100.37Protocol:PPOS, n (%)48 (27.9%)20 (27.0%)0.88GnRH antagonist, n (%)124 (72.1%)54 (73.0%)Trigger method: GnRH agonist only, n (%)158 (91.9%)67 (90.5%)0.72Dual trigger, n (%)14 (8.1%)7 (9.5%)Infertility etiology: Unexplained infertility, n (%)63 (36.6%)25 (33.8%)0.67 Female factor, n (%)58 (33.7%)24 (32.4%)0.84 Male factor, n (%)51 (29.7%)25 (33.8%)0.64Transfer day Progesterone (ng/mL), mean ± SD32.1 ± 13.010.3 ± 3.80.001Transfer day endometrial thickness(mm), mean ± SD9.2 ± 1.49.8 ± 1.30.92 Baseline characteristics and treatment parameters Previous live birth Yes Protocol: PPOS, n (%) No statistically significant differences were observed between the two groups in terms of embryo day, embryo quality, PGD application, or double embryo transfer (all p > 0.05) (Table 2). Table 2Embryo transfer featuresVariableGroup 1 (n = 172)Group 2 (n = 74)p-valueEmbryo day, n (%) • Day 5155 (90.1%)67 (90.5%)0.92 • Day 617 (9.9%)7 (9.5%)Embryo quality, n (%) • Good54 (31.4%)22 (29.7%)0.88 • Fair87 (50.6%)39 (52.7%) • Poor31 (18.0%)13 (17.6%)PGD, n (%)28(16.3%)13(17.6%)0.95Double embryo transfer22(12.8%)10(13.5%)e0.98 Embryo transfer features Group 2 demonstrated significantly higher reproductive outcomes compared to Group 1 (Table 3). The total pregnancy rate was 66.2% in Group 2 versus 47.1% in Group 1 (p = 0.006), and the clinical pregnancy rate was also higher in Group 2 (56.8% vs. 40.7%, p = 0.020). Miscarriage rates were comparable between the groups (4.1% vs. 3.5%, p = 0.83). Furthermore, live birth rates were significantly greater in Group 2 than in Group 1 (52.7% vs. 37.2%, p = 0.024). Table 3Pregnancy outcomes of the groupsVariableGroup 1 (n = 172)Group 2 (n = 74)p-valueTotal pregnancies, n (%)81 (47.1%)49 (66.2%)0.006Clinical pregnancies, n (%)70 (40.7%)42 (56.8%)0.020Miscarriages, n (%)6 (3.5%)3 (4.1%)0.83Live births, n (%)64 (37.2%)39 (52.7%)0.024 Pregnancy outcomes of the groups Logistic regression analysis revealed that endometrium thickness, serum progesterone level, and embryo quality were not independently associated with clinical pregnancy but but the group in which the patients are is significantly associated with pregnancy rates. The results showed that being in the second group was significantly associated with an increased likelihood of the pregnancy rates (B = 0.865, SE = 0.384, Wald = 5.063, p = 0.024; OR = 2.374) (Table 4). Table 4Logistic regression analysis of factors associated with clinical pregnancyVariableBS.E.WalddfSig.Exp(B)Being group 2 (LPS start on the day of ET)0,8650.0245.06310.0242.374Endometrium thickness-0.0720.0940.57910.4470.931Progesterone level0.0120.0092.03910.1531.013Embryo quality-0.0620.1890.10910.7420.940Constant0.3660.9580.14610.7031.442 Logistic regression analysis of factors associated with clinical pregnancy ROC analysis was conducted for each subgroups and all patients to determine the ability of the serum progesterone level to predict the presence of pregnancy and there was no statistically significant value (p:0.151; AUC: 0.553). To evaluate the independent association between transfer-day progesterone levels and clinical pregnancy, an additional multivariable logistic regression analysis was performed after excluding treatment group (beeing group 2) assignment from the model. In this sensitivity analysis, transfer-day progesterone levels were not significantly associated with clinical pregnancy outcomes (OR = 0.998, p = 0.856). Discussion This study assessed the impact of the initiation day of luteal phase support in natural vitrified-warmed blastocyst transfer on pregnancy outcomes. Initiating luteal phase support two days prior to transfer dramatically elevated blood progesterone levels; nevertheless, this adversely impacted pregnancy outcomes. The average progesterone level of patients who started luteal phase support on the day of transfer was found to be 10.3 ± 3.8 ng/mL, but it did not show a significant value in ROC curve analysis to predict pregnancy. Luteal phase support is essential in hormone replacement therapy cycles, while its requirement in natural cycles remains debatable [8, 9]. Some clinicians believe that the corpus luteum in the NC secretes appropriate progesterone hormone for the implantation and the others consider that approximately 8% of normo-ovulatory patients have corpus luteum insufficiency [10]. Consequently, the use of LPS is considered as a distinct clinical approach. Mizrachi et al. reported that progesterone administration for the LPS increases the CPR and LBR rate, however, administering hCG for LPS has no positive effect on the pregnancy outcomes [11]. According to the meta-analysis conducted in 2023, moderate-quality evidence indicated that LPS was correlated with enhanced LBR and CPR in NC-FET cycles [12]. The lack of a control group not administered LPS in the current study limited our ability to assess the impact of LPS on pregnancy; however, the pregnancy rates in the groups (47.1% vs. 66.2%) were considered satisfactory. Mid-luteal serum progesterone level is considered important for the successful reproductive outcomes. In the retrospective study conducted in 2020 revealed that the patients with serum P levels < 10 ng/ml one day before embryo transfer during NC-FET had a lower live birth rates than the patients with serum P levels > 10 ng/ml [13]. In a study involving a total of 181 patients, it was shown that approximately half of the patients had a progesterone level < 10 on the day of embryo transfer, and the pregnancy rates of these patients with subcutenous progesterone support were similar to those of the group with a progesterone level > 10 [14]. Erden et al. reported that serum P levels > 7 ng/ml on one day prior the embryo transfer secures the live birth rate in NC-FET. Generally, the literature indicates that blood P4 levels ranging from 8.75 ng/ml to 10.6 ng/ml constitute a crucial threshold, beneath which pregnancy outcomes worsen [15–17].In the present study, all patients were administered the LPS regardless of the serum progesterone level according to the clinical approach. It was observed that the average serum progesterone levels of patients who started progesterone support two days ago were 32.1 ± 13.0 ng/mL, which was significantly higher than the other group. While progesterone supplementation is well established to support pregnancy outcomes in NC-FET cycles, the higher pregnancy rates observed in the group that initiated LPS on the day of transfer may suggest that markedly elevated progesterone levels at the time of transfer could potentially interfere with endometrial receptivity or embryo implantation. It is possible that exogenous progesterone administered prior to transfer may advance endometrial maturation beyond the optimal implantation window, thereby affecting endometrial–embryo synchrony. These observations should be interpreted with caution, as our retrospective data do not allow causal conclusions to be drawn. No significant cut-off progesterone level was found in ROC analysis for predicting pregnancy rates, regression analysis showed that progesterone levels did not have a significant effect on pregnancy rates, statistically. Additionally, Lawrenz et al. reported that serum progesterone level on the transfer day is not associated with the pregnancy rate if routine LPS is administered consistence with our study [18]. However, starting LPS on the day of transfer was shown to have a positive impact on pregnancy rates in the present study. Previously, it has been demonstrated that increasing serum progesterone levels with intramuscular injections improves pregnancy outcomes [19]. Nowadays, because of vaginal progesterone can avoid the liver’s first-pass metabolism and achieve a higher tissue level, it is becoming more preferable. Studies on LPS in NC-FET cycles have generally evaluated the outcome of vaginal progesterone use considering the progesterone use [20]. Although there is no study investigating the efficacy of oral dydrogesterone in this regard on NC-FET cycles, it can be said to have similar pregnancy outcomes to vaginal progesterone in fresh and HRT IVF cycles [21, 22]. In our study, however, no difference could be evaluated between the two methods of administration because patients received both vaginal and subcutaneous progesterone treatment. The retrospective nature of the study is one of the most significant limitations. LPS was not administered to patients based on their progesterone levels on the day of embryo transfer; LPS was administered to all patients. This could be another limitation of the study. The fact that the study was conducted in the same clinic with similar clinical views and strict inclusion criteria highlights the study’s strengths. In conclusion, it has been shown that NC-FET patients receiving LPS on the day of transfer in cycles increases pregnancy rates. Although tailored LPS is recommended, it has been observed that starting vaginal and subcutaneous progesterone on the day of transfer, without tailoring, leads to better pregnancy outcomes. Authors’ contributions BT, BO, CSA contributed to designing, planning and manuscript writing. YES, BB, MS, RA contributed to recruitment of subjects and conduct of study. BT, BO, CSA drafted the manuscript and did data analysis. YES, BB, MS, RA helped in recruitment of subjects and data analysis. All authors approved the final manuscript. Data availability The data that support the findings of this study are available from the corresponding author upon reasonable request. Declarations Ethics approval and consent to participate Ankara University Faculty of Medicine Department of Obstetrics and Gynecology Academic Board (Approval No: İ25-98745), Informed constent was obtained from all participants. All procedures were conducted in accordance with the Declaration of Helsinki. Consent for publication None. 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. References Associated Data Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request.