Comparison between oral dydrogesterone versus no drugs in clinical outcome within the natural cycle-frozen-thawed embryo transfer cycle: a retrospective analysis.

All day 2 cleavage embryos and blastocysts cryopreserved according to a vitrification protocol during the fresh cycle. Day 2 cleavage embryo freezing was the usual and routine procedure, and extended blastocyst culture was performed when the number of good quality day 2 cleavage embryo is more than 6 in the fresh cycle. Day 2 cleavage embryos were graded as one to six according to the evenness of each blastomere and the percentage of fragmentation [ 18 ]. Embryos of four cells and of grade one to four were regarded as good quality embryos in this study. Blastocysts were graded according to Gardner et al. [ 19 ]. Grade A was considered top quality blastocyst, grade B was median and grade C was low. Blastocysts with grade A or B were considered as good quality blastocyst.

The primary outcome was live birth rate per embryo transfer. The secondary outcomes included positive pregnancy rate, clinical pregnancy rate, ectopic pregnancy rate, embryo implantation rate and miscarriages. Positive pregnancy test was defined if the serum human chorionic gonadotropin, beta subunit (beta-hCG) concentration > 5.83 IU/L and the clinical pregnancy was confirmed by ultrasound examination 4–5 weeks after embryo transfer (presence of a gestational sac). Miscarriage was defined if the clinical pregnancy was lost spontaneously before 20 weeks of gestation. Implantation rate was defined as the ratios of the gestational sac number over the transferred embryo number. Shapiro-Wilk test was performed to examine the normality of continuous variables, indicating the variables were normally distributed. Thus, all the variables were expressed as a mean ± SD and analyzed with the student’s t-test. Additionally, categorical variables were expressed as percentages (frequency) and studied utilizing the Pearson Chi-square test or Fisher’s exact probability. Propensity score matching (PSM) was carried out to adjust for numerical differences and to balance between the two groups. The matching parameters were the woman’s age, the woman’s body mass index (BMI), the serum level of LH、E level, AFC and the number of transferred embryo, with a matching tolerance of 0.02, and non-repetitive sampling at a ratio of 1-to-1. All statistical analyses were run by software SPSS (version 23, IBM, USA), with P  < 0.05 (two-tailed) regarded as a significant difference.

A totally 2657 cycles were screened during the study period, 1497 cycles underwent the natural protocol, and 1160 cycles underwent the LPS protocol; and only the first cycle of NC-FET of each subject were included. The baseline characteristics before PSM are presented in Table  1 . Significant differences were found in woman’s age, antral follicle account, E2 level on the day of LH surge, the number of transferred embryos, the secondary infertility and the ratio of day 2 cleavage embryo/blastocyst, except for BMI, infertility duration, LH level, endometrial thickness, good quality embryo rate and infertility factors. To eliminate the bias from the unbalanced characteristics, all cycles were 1:1 matched in 8 variables using PSM. Thus 802 cycles in each group remained after PSM. The baseline characteristics after PSM are presented in Table 2 . No significant differences were found in each characteristic between two groups. Table 1 Demographic characteristics of women in the two groups before PSM Characteristic LPS Group ( n  = 1160) No drugs Group ( n  = 1497) P -value Woman’s age (years) 34.191 ± 4.167 33.500 ± 4.006 <0.001 Woman’s BMI (Kg/m2) 21.130 ± 2.579 20.866 ± 2.644 0.010 Infertility duration (years) 2.881 ± 2.393 3.082 ± 2.383 0.032 Antral follicle count 12.756 ± 6.765 11.823 ± 6.050 <0.001 LH surge (mIU/mL) 46.048 ± 108.552 42.113 ± 19.691 0.170 The endometrial thickness (mm) 8.674 ± 1.144 8.536 ± 1.100 0.002 E2 on the LH surge day(pg/L) 257.525 ± 115.774 318.380 ± 134.432 <0.001 Number of embryos transferred 1.265 ± 0.442 1.494 ± 0.5001 <0.001 Good quality cleavage embryo (%) 460 (58.300) 1131 (60.380) 0.317 Good quality blastocyst (%) 649 (95.580) 352 (96.700) 0.379 Infertility, n (%) 692 (59.655) 909 (60.721) 0.578 Primary infertility (%) Secondary infertility (%) 468 (40.344) 588 (59.655) 0.006 Cause of infertility, n (%) Male factor (%) 270 (23.276) 360 (24.048) 0.643 Female factor (%) 469 (40.431) 551 (36.807) 0.057 Double factor (%) 139 (11.983) 186 (12.424) 0.730 Unknown factor (%) 280 (24.138) 400 (26.720) 0.130 Percentage of Day 2 embryo (%) 41.81 75.68 < 0.001 Percentage of blastocyst (%) 58.19 24.32 < 0.001 P  < 0.05 has statistical difference Demographic characteristics of women in the two groups before PSM P  < 0.05 has statistical difference Table 2 Demographic characteristics of women in the two groups after PSM Characteristic LPS Group ( n  = 802) No drugs Group ( n  = 802) P -value Woman’s age (years) 33.877 ± 4.270 33.921 ± 4.085 0.830 Woman’s BMI (Kg/m2) 20.915 ± 2.537 20.932 ± 2.588 0.886 Infertility duration (years) 2.938 ± 2.301 3.025 ± 2.138 0.320 Antral follicle count 12.516 ± 6.526 12.333 ± 6.140 0.562 LH surge(mIU/mL) 41.892 ± 20.614 42.556 ± 19.557 0.568 The endometrial thickness (mm) 8.626 ± 1.125 8.566 ± 1.098 0.280 E2 on LH surge day (pg/L) 291.409 ± 122.564 294.974 ± 120.412 0.213 Number of embryos transferred 1.372 ± 0.484 1.369 ± 0.483 0.873 Good quality cleavage embryo (%) 419 (54.915) 421 (55.762) 0.740 Good quality blastocyst (%) 322 (96.119) 331 (95.942) 0.906 Infertility, n (%) Primary infertility (%) 337 (42.020) 332 (41.397) 0.800 Secondary infertility (%) 465 (57.980) 470 (58.604) 0.800 Cause of infertility, n (%) Male factor (%) 203 (25.312) 191 (23.816) 0.486 Female factor (%) 358 (44.638) 362 (45.137) 0.841 Double factor (%) 57 (7.107) 78 (9.726) 0.058 Unknown factor (%) 184 (22.943) 171 (21.322) 0.434 Percentage of Day 2 embryo (%) 58.98 57.48 0.544 Percentage of blastocyst (%) 41.02 42.52 0.544 P  < 0.05 has statistical difference Demographic characteristics of women in the two groups after PSM P  < 0.05 has statistical difference The pregnancy characteristics before and after PSM were shown in Tables 3   and 4 . Table  3 shows the clinical outcomes after the embryo transfer. No significant difference was found in live birth rate (30.351% vs.28.864%, P  = 0.402), the clinical pregnancy rate (38.298% vs. 34.537%, P  = 0.055), implantation rate (31.990%vs. 25.120%, P  = 0.141), miscarriage rate (5.095% vs. 4.231%, P  = 0.283), ectopic pregnancy rate (0.343% vs. 0.736%, P  = 0.181), and multiple birth rate (2.162% vs. 2.875%, P  = 0.256) between the LPS group and no drugs groups. Table  4 shows the clinical outcomes after PSM. No significant difference was found in live birth rate (27.930% vs. 32.173%, P  = 0.064), the clinical pregnancy rate (33.416% vs. 38.030%, P  = 0.054), implantation rate (27.322% vs. 29.636%, P  = 0.229), miscarriage rate (16.418% vs. 15.635%, P  = 0.798), ectopic pregnancy rate (0.374% vs. 0.748%, P  = 0.315), and multiple birth rate (2.868% vs. 2.369%, P  = 0.532) between the LPS group and no drugs groups. Table 3 Pregnancy outcomes before PSM Characteristic LPS Group ( n  = 1160) No drugs Group ( n  = 1497) P -value Live birth rate (%) 352(30.351) 432(28.864) 0.402 Clinical pregnancy rate (%) 443(38.198) 517(34.537) 0.055 Implantation Rate (%) 468(31.990) 562(25.120) 0.141 Miscarriage rate (%) 59(5.095) 63(4.231) 0.283 Ectopic pregnancy rate (%) 4(0.343) 11(0.736) 0.181 Multiple birth rate (%) 25(2.162) 43(2.875) 0.256 P  < 0.05 has statistical difference Pregnancy outcomes before PSM P  < 0.05 has statistical difference Table 4 Pregnancy outcomes after PSM Characteristic LPS Group ( n  = 802) No drugs Group ( n  = 802) P -value Life birth rate (%) 224 (27.930) 258 (32.173) 0.064 Clinical pregnancy rate (%) 268 (33.416) 305 (38.030) 0.054 Implantation Rate (%) 300 (27.322) 326 (29.636) 0.229 Miscarriage rate (%) 44 (16.418) 48 (15.635) 0.798 Ectopic pregnancy rate (%) 3 (0.374) 6 (0.748) 0.315 Multiple birth rate (%) 23 (2.868) 19 (2.369) 0.532 P  < 0.05 has statistical difference Pregnancy outcomes after PSM P  < 0.05 has statistical difference Additionally, there were no statistically significant difference in the live birth rate, the clinical pregnancy rate, implantation rate, miscarriage rate, ectopic pregnancy rate, and multiple birth rate neither in Day2 cleavage embryo transfer sub-group nor in blastocyst transfer sub-group (Table  5 ). Table 5 Clinical outcomes of day 2 cleavage stage and blastocyst transfer subgroup after PSM Outcome Day 2 cleavage embryo P -value Blastocyst P -value LPS Group ( n  = 473) No drugs Group ( n  = 461) LPS Group ( n  = 329) No drugs Group ( n  = 341) Live birth rate (%) 101 (21.353) 112 (24.295) 0.284 123 (37.386) 146 (42.815) 0.152 Clinical pregnancy rate (%) 121 (25.581) 134 (29.067) 0.232 147 (44.681) 173 (50.733) 0.117 Implantation rate (%) 141 (18.800) 149(19.867) 0.601 153 (45.266) 177(49.709) 0.245 Miscarriage rate (%) 20 (16.529) 22 (16.418) 0.980 26 (11.207) 27 (15.607) 0.194 Ectopic pregnancy rate (%) 2 (0.423) 2 (0.434) 0.979 1 (0.304) 4 (1.173) 0.191 Multiple birth rate 19 (4.017) 13 (2.820) 0.315 4 (1.216) 6 (1.760) 0.562 P  < 0.05 has statistical difference Clinical outcomes of day 2 cleavage stage and blastocyst transfer subgroup after PSM P  < 0.05 has statistical difference

A retrospective observational study was performed at the Center of Reproductive Medicine, the University of Hong Kong-Shenzhen Hospital from 2015 to 2023 and the Center of Reproductive Medicine, Shanghai Sixth People’s Hospital Affiliated to Shanghai Jiaotong University School of Medicine using data in 2023. The NC-FET cycles were groups by the use of LPS or not without special criteria. In this study, the following exclusion criteria were used: woman age more than 43 years old, uterine abnormality such as: adhesions, adenomyosis, congenital malformations, patients with azoospermia, women with severe endometriosis, serious endocrine or metabolic disorder, amenorrhea or recurrent pregnancy loss and LPS with injected or vaginal progesterone. The study was approved by the ethical committee of the University of Hong Kong-Shenzhen Hospital and the approval number was 2023 − 270.

Women having regular cycles underwent NC-FET whereas those with irregular cycles underwent aFET. For NC-FET, women attended the hospital from 18 days before the next expected period. When the leading follicle is more than 14 mm in diameter, the serum oestradiol (E2), progesterone (P4) and luteinizing hormone (LH) concentrations was tested every day until the LH surge, which was defined as an elevation of the LH to 2 times the level of the average of the previous 3 days and the absolute level of the LH should be greater than 20 IU/L. FET with day 2 cleavage embryo was performed on the third day after the LH surge, blastocyst was transferred on the sixth day after the LH surge. Endometrial thickness was measured by Ultrasound scanning between the presence of leading follicle and the day on the LH surge. The threshold of endometrial thickness is at least 7 mm. There was no ovulation trigger used in each cycle. On third day after the LH surge, oral dydrogesterone (Duphaston 10 mg tablet, Abbott Healthcare Products, Netherlands) at 10 mg three times daily was administrated to one group, while the other group was not administrated any drugs. A blood pregnancy test of HCG was carried out 12–14 days after embryo transfer. If the pregnancy test was positive, women underwent pelvic ultrasound examination 14 days later to confirm intrauterine pregnancy, and the following several ultrasound examinations were performed up to 12 weeks of gestation.

In this retrospective cohort analysis, we found no significant difference in LBR or clinical pregnancy rate between the Dydrogesterone and no drugs groups in NC-FET cycle, neither before PSM nor after PSM. And there is no significant difference between the other clinical outcomes. We previously added progesterone to the NC-FET based on empirical grounds because the evidence for no LPS in NC-FET cycles was weak. In our center, day 2 cleavage embryo is routinely transferred, while most IVF centers in the world perform the transfer with day 3 cleavage embryo. Recent evidences demonstrated that there is no significant difference in pregnancy rate [ 20 ] or LBR [ 21 ] between day 2 and day 3 cleavage embryo transfer cycles. Since the clinical outcomes should be different between day 2 cleavage embryo and blastocyst transfer, different embryo subgroup comparisons of clinical outcomes were performed. And no significant difference was found between each subgroup. According to a recent meta-analysis, there is moderate-quality evidence that progesterone supplementation combined with LPS raised LBR and CPR in NC-FET. The evidence based on four RCTs, two of them use the HCG trigger, while the other two RCT performed the true NC-FET. Vaginal micronized progesterone was used in these studies, not Dydrogesterone as LPS [ 22 ]. Another meta-analysis showed that LPS was associated with higher CPR compared with no LPS group. However, few RCTs collected in this analysis, and most of data came from retrospective studies. In the true NC-FET cycles without HCG trigger, five studies demonstrated no significant difference between the LPS and no LPS groups [ 23 ]. Actually, Lee et al. reported a retrospective study demonstrated no difference between the no LPS and the HCG injection group in NC-FET [ 10 ]. Our study followed the same NC-FET procedure protocol as Lee el al’s study, except the administration of Dydrogesterone as LPS in our study. Recently, some authors have preferred NC-FET to Hormone Replacement Treatment (HRT) FET cycle, according to the “back to nature” attitude, and some updated evidence indicated more favorable perinatal outcomes when NC is performed [ 24 , 25 ]. Theoretically, a true NC entails no intervention or extra medication delivery. LPS might not be required because ovulation is the key component in the NC for timing embryo transfer and the subsequent corpus luteum for embryo implantation. As a result, there was little evidence supporting the use of LPS in actual NC-FET. Theoretically, in contrast to the HRT-FET cycle, corpus luteum in the natural menstrual cycle can support the embryo implantation end early pregnancy. The luteal phase defect usually is found in stimulation IVF cycles resulting from the supraphysiological steroid hormones, negative feedback at hypothalamic-pituitary level, preventing LH surge and the removal of granulosa cells during oocyte retrieval [ 26 ]. There is an argument if the natural cycle encounters the luteal phase defect, however, it is subjected to controversy and diagnostic difficulty. Single intake of progestogen should be more convenient for the patient. According to research on the clinical randomized control trial of Lotus I and Lotus II for IVF fresh embryo transfer cycles, the dosage of 30 mg oral dydrogesterone is now the lowest in the ART LPS, and their clinical results were deemed acceptable and appropriate [ 27 , 28 ]. Moreover, the single use of 30 mg oral dydrogesterone in HRT-FET treatment was performed by our center previously. In that retrospective study, 30 mg/d oral dydrogesterone and 180 mg/d MVP gel revealed similar reproductive outcomes in HRT-FET cycles [ 29 ]. Single use of dydrogesterone should be enough and effective in LPS, even in the natural cycle. Interestingly, there is rare report of oral dydrogesterone in NC-FET treatment, and the significant clinical benefits were reported in the NC-FET cycle with the vaginal progesterone administration [ 22 ]. In our study, no difference in clinical outcomes was demonstrated, which may result from the low dosage or no necessity of LPS in NC-FET. Single dosage of 30 mg oral dydrogesterone was lowest in the LPS of ART treatment. Additionally, safer clinical treatments involve lower drug dosages. The Lotus I and Lotus II studies both demonstrated the safety and tolerability of dydrogesterone [ 27 , 28 ]. Since dydrogesterone was widely and safely used in clinical treatment, our study attempted a low dosage of medication in NC-FET. And we found rare side effect or allergic reaction of dydrogesterone in our study. Moreover, RCT should be performed in the future to get better evidence in the NC-FET LPS with oral dydrogesterone. As we know, the LH surge triggers ovulation, which is followed by fertilization, embryo implantation and natural conception. FET cycle need not the ovulated oocyte for fertilization but the reference timepoint for transferring embryo. There is a marked inter-individual variation in duration between the onset of LH surge and ovulation [ 30 ]. As a result, in the true NC-FET cycle, determination the time of embryo transfer usually bases on the onset of the LH surge, not the ovulation. However, there is no consensus developed the definition of the onset of the LH surge. Summarily, the onset of LH surge defined as increase of 1.8 to 6-fold from the LH baseline level [ 31 – 35 ], increase of two [ 36 ] or three [ 37 ] times above the mean of preceding LH level, and at least 30% of the amplitude of the LH surge [ 6 ]. In our study, the onset of LH surge was defined as an elevation of the LH to 2 times the level of the average of the previous 3 days and the absolute level of the LH should be greater than 20 IU/L. it was reasonable and similar with the previous description. This study’s clinical results were comparable to those of other research on FET treatment. In order to date the ovulation and subsequent transfer, the real NC-FET cycle requires extensive ultrasound monitoring and hormone detection. One study reported that 6% of NC-FET cycles were cancelled because ovulation was not accurately identified [ 38 ]. Therefore, the NC-FET cycles using HCG were performed to trigger the ovulation and determine the day of embryo transfer, which is called modified NC-FET. Modified NC-FETs are more adjustable, patient-friendly, and require less hormone and ultrasonographic monitoring than the original NC-FET. In modified NC-FET cycle, HCG trigger is performed when the dimeter of dominant follicle is more than 16 mm or reaches 16–18 mm. The maturation and ovulation with HCG takes places 36–40 h later [ 39 ]. Theoretically administration of HCG result in the effect of LPS and increasing the clinical pregnancy rate. However, the recent Cochrane review did not demonstrate an advantage in one approach of NC-FET above others, although this based on the low number of high-quality studies [ 40 ]. In 2020, an RCT focusing on trigger or not to trigger ovulation in NC-FET reported no difference in clinical outcomes. This study suggested that the choice of NC-FET protocol can be made according to the preference of physician or patient, the modified NC-FET is more patient-friendly and cost-effective because of the fewer clinic visits and serum hormone detection [ 41 ]. Currently, there is few studies focus on whether LPS is necessary in the NC-FET, although much data demonstrated the administration of different progesterone in NC-FET. Our study indicated reasonable clinical outcomes in true NC-FET with no drugs. And LPS with the conveniently oral drugs was practiced to improve the clinical outcomes. Our study encourages the clinical research, especially the RCT, in the LPS of NC-FET. We aim to find the naturel, without intervention or convenient protocol in NC-FET. In this study, we retrospectively investigated whether LPS was necessary in NC-FET. This is the first comparison between LPS with oral dydrogesterone and no drugs in NC-FET. The limitation of this study is its retrospective nature despite its large sample size. The prospective RCT can be designed in the future to substantiate the evidence of the necessity of LPS and oral dydrogesterone effectiveness in the NC-FET treatment. Second, we did not examine ovulation with ultrasound after the LH surge. Third, the women recruited in this study are Chinese, which may limit generalizability to other ethnicities. In conclusion, 30 mg/d oral dydrogesterone and no LPS revealed similar reproductive outcomes in NC-FET cycles in the study.

Frozen-thawed embryo transfer (FET) were commonly performed around the world from 1983 [ 1 ]. The FET treatment can prevent the risk of ovarian hyperstimulation syndrome in contrast to the super ovulation cycle with fresh embryo transfer [ 2 ]. Additionally, FET is necessary for patients to have preimplantation genetic screening and diagnosis. Furthermore, the notable global rise of FET cycles over the past few decades has been largely attributed to safe and effective vitrification techniques [ 3 , 4 ]. A larger RCT demonstrated that the live birth rate (LBR) of FET was higher than the fresh embryo transfer cycle [ 5 ]. Usually, FET protocol includes natural cycle frozen-thawed embryo transfer (NC-FET), ovulation induction cycle, or artificial FET (a-FET). The NC-FET cycle’s successful embryo implantation and pregnancy are dependent on spontaneous ovulation and the corpus luteus’s subsequent synthesis of progesterone. However, the level of serum progesterone vary considerably among different women [ 6 , 7 ], and the ovulatory nature of the cycle may be somewhat predictive of conception and live birth in infertile women [ 8 ]. Typically, the IVF fresh stimulation cycle features a luteal phase defect [ 9 ]. However, the normal natural menstrual cycle should not encounter the same defect since there is follicle develops and ovulates. The argument about the luteal phase defect in natural cycle should be diagnosed difficult and controversy [ 10 ]. Although there isn’t much evidence to support luteal phase support (LPS) in the natural cycle, it is really frequently utilized in NC-FET with varying routes of administration. Using vaginal progesterone has been shown in some RCTs to improve LBR and clinical pregnancy (CP) [ 11 , 12 ]. And, LPS with vaginal progesterone did not decrease the miscarriage rate (MR) in these studies [ 11 , 12 ]. A recent meta-analysis showed that the LPS in NC-FET can decrease the risk of preterm birth while the other clinical outcomes did not show difference between the LPS or not in NC-FET [ 13 ]. One randomized study in 2011 demonstrated the use of vaginal progesterone increased the significant higher LBR in NC-FET [ 14 ]. However, Lee et al.‘s RCT revealed no difference in NC-FET between hCG-using and non-using groups [ 10 ]. Additionally, the use of intramuscular progesterone also did not increase the clinical pregnancy rate in NC-FET compared with the placebo group from another RCT [ 15 ]. Furthermore, several retrospective studies did not demonstrate any improvement in PR, ongoing pregnancy or LBR with LPS in NC-FET [ 16 , 17 ]. This retrospective study aimed to compare the LBR and clinical pregnancy rate in NC-FET with oral dydrogesterone as luteal phase support or without any drugs.