Study design and patients
This retrospective cohort study was conducted at the Department of Assisted Reproduction of the Ninth People's Hospital of Shanghai Jiao Tong University School of Medicine (Shanghai, China) from January 2015 to February 2017. It was approved by the Ethics Committee (Institutional Review Board) of the Ninth People's Hospital of Shanghai. All participants provided informed consent after counselling for infertility treatments and routine IVF procedures. All patients underwent artificial endometrial preparation with a combination of oral dydrogesterone and vaginal progesterone for LPS only. Inclusion criteria were as follows: age < 50 years; BMI < 30 kg/m2; and no systemic diseases. Due to the potential bias from recurrent implantation failure, analysis was limited only to patients undergoing their first or second FET cycles. The exclusion criteria were recurrent miscarriages, infertility due to severe male factors, uterine diseases (e.g., fibroids, polyps, and previously diagnosed Müllerian abnormalities) or the presence of hydrosalpinx. Likewise, records with missing data were excluded. Therefore, a final cohort of 3659 cycles was analysed.
Protocol of endometrial preparation and embryo transfer
Endometrial preparation was induced with sequential provision of oral E2 (Fematon-red tablets, Abbott Biological, USA) 8 mg/d from cycle day 3 onwards. After 12–14 days of oestrogen therapy, a blood sample was collected, and a vaginal ultrasound were performed for measurements of oestradiol, LH, and progesterone levels and endometrial thickness, respectively. If the endometrial thickness was < 7 mm, oestrogen therapy was extended for as long as 7 days if required. If a triple-line endometrium of ≥ 7 mm thickness was observed with serum progesterone concentrations < 1.0 ng/ml, P was administered both orally (40 mg dydrogesterone and 8 mg E twice per day; Fematon-yellow tablets, Abbott Biologicals, USA) and vaginally (400 mg/d; Utrogestan, Besins Manufacturing, Belgium). The time of thaw and transfer was set as the 3rd or 5th day after P administration depending on the embryo stage. The maximum number of embryos transferred was two per patient in each FET cycle. Given the absence of corpora lutea, exogenous P4 administration was continued until 10-week gestation. A blood sample was obtained between 7 and 8 am on the day of frozen-thawed embryo transfer for serum oestradiol and progesterone measurements. Embryos were generated from either IVF or intracytoplasmic sperm injection cycles and were vitrified and warmed as previously described [32, 33] on day 3 or at the blastocyst stage. All embryo transfers were performed under ultrasound guidance. The number of transferred embryos and the stage of the embryo were recorded.
Serum hormonal assays
Hormone levels were measured by chemiluminescence (Abbott Biologicals B.V.). Intra- and inter-assay coefficients of variation were 7.9 and 10% for P. The synthetic progestogens used (dydrogesterone) did not have any cross-reaction with the progesterone assay.
Outcome measures
The primary endpoint was the relationship between progesterone levels on the embryo transfer day and the live birth rate per cycle. Secondary endpoints included the implantation rate, positive β-HCG test rate, clinical pregnancy rate, ongoing pregnancy rate at 12 weeks of gestation and early miscarriage (first-trimester pregnancy loss) rate. A human chorionic gonadotrophin (hCG) test was considered positive if HCG was > 10 IU/l. The clinical pregnancy rate (CPR) was defined as the proportion of patients with a gestational sac with or without foetal heart activity under ultrasound examination 7 weeks after ET among all transfer cycles. The implantation rate was defined as the number of gestational sacs divided by the number of embryos transferred. The early miscarriage rate was defined as the proportion of patients with spontaneous pregnancy termination before the gestational age of 12 weeks. The ongoing pregnancy rate (OPR) was defined as the proportion of patients with a gestational sac with foetal heart activity assessed by ultrasound examination at 12 weeks of gestation among all transfer cycles.
The live birth rate (LBR) was defined as the proportion of patients with live birth among all transfer cycles. The main neonatal outcomes of the study included preterm birth (PTB), low birthweight (LBW), and major congenital malformations. PTB was defined as delivery before 37 completed weeks of gestation. LBW was defined as a birthweight below 2500 g. Major congenital malformations were defined and coded according to the Q codes (Q00–Q99) of the International Classification of Diseases, 10th Revision (ICD-10).
Statistical analysis
Serum P on the day of ET was classified into four quartiles according to the 25th, 50th and 75th percentiles. Continuous variables are presented as the mean plus/minus standard deviation, and the Kruskal-Wallis test was used to compare continuous variables. Categorical variables were described as the frequency with the rate, and between-group differences were assessed by chi-square test or Fisher’s exact test, as appropriate. The association between serum P on the day of ET and pregnancy outcome was evaluated by univariable and multivariable logistic regression analysis. All potential confounders were introduced into the regression equation for adjustment by the entry method, whether significant differences between groups were observed. These included maternal age (continuous), maternal BMI (continuous), duration of infertility (continuous), type of infertility (primary, secondary), infertility diagnosis (tubal factor, male factor, ovulatory, endometriosis, unexplained or combined), ovarian stimulation protocol (GnRH-a short, mild stimulation or progestin-primed ovarian stimulation (PPOS)), fertilization method (IVF, ICSI or IVF + ICSI), endometrial thickness, progesterone level on embryo transfer day, number of embryos transferred (single or double), embryo stage at transfer (cleavage or blastocyst) and duration of embryo cryopreservation (continuous).
All P values were based on two-sided tests, and P < 0.05 was considered statistically significant. Statistical analysis was performed with the Statistical Package for the Social Sciences (version 20.0; SPSS Inc., USA).