Abstract Study question Does subcutaneous injection of human chorionic gonadotropin (hCG) at the onset of progesterone supplementation affect clinical pregnancy rate in frozen embryo transfer cycles? Summary answer Clinical pregnancy rates are higher when hCG injection is administered at the onset of 300 mg/day progesterone in patients undergoing frozen embryo transfer. What is known already HCG and progesterone play crucial roles in embryo transfer and implantation during in vitro fertilization. Exisitng studies have demonstrated that pre-administration of hCG at the onset of progesterone supplementation for frozen embryo transfer increases clinical pregnancy rates in patients with endometriosis. However, the synergistic effects of hCG and progesterone co-supplementation in a broader range of patients remain unexplored and require further investigation. A multifactorial analysis of hormone replacement therapy, covering various patient-specific aspects, is necessary to understand the implications of this approach. Study design, size, duration This retrospective cohort study analyzed data from 2,251 single frozen embryo transfers performed between Jan 2019-Aug 2023 at the INVICTA Fertility Centre, Poland. The dataset was selected by the practitioner and included age, AMH, BMI, embryo morphology grade, total number of transfers, medical and embryological procedures, recurrent implantation failure, preimplantation genetic testing, endometrial thickness before transfer (≥ 7 mm), and the menstrual cycle day on which progesterone was started. Participants/materials, setting, methods Patients (21-47 years) receiving 300 or 600 mg/day progesterone for five days were divided into the control and hCG (6,500 IU hCG injection on progesterone onset) groups. A machine learning model was trained using a gradient boosting technique (100 decision trees, four leaves, maximum depth of two nodes, learning rate of 0.05) to identify factors affecting the clinical pregnancy rate. Statistical analysis was performed to compare clinical pregnancy rates between the groups. Main results and the role of chance Our machine-learning model revealed that hCG injection (512 processes vs. 1739 processes without it) at the onset of progesterone administration is an important feature for the clinical pregnancy rate. The performance accuracy of the model was approximately 70%. Of the 1,294 protocols where patients received 300 mg/day progesterone, 311 included hCG injection while 983 did not. The clinical pregnancy rate was significantly higher in the hCG group than that in the control group (54% vs. 46%, P < 0.05), with 167 clinical pregnancies observed in patients receiving hCG injections against 452 pregnancies without it. In contrast, among the 957 protocols where 600 mg/day progesterone was administered, the previously observed effect was not significant (P > 0.05). Comparing 201 processes with hCG injection to 756 processes without it, the incidence of clinical pregnancies was 46.8% and 47.2% (with 94 and 357 confirmed pregnancies), respectively. Other factors such as the morphology grade of the embryo used for transfer, patient’s age, AMH, BMI, total number of transfers performed on the patient, menstrual cycle day on which progesterone was started, occurrence of recurrent implantation failure, and the use of preimplantation genetic testing were not significantly different between the groups. Limitations, reasons for caution This study focused on frozen embryo transfers and validation across other protocols using fresh embryo transfers, and different doses or routes of hCG and progesterone administration are needed. The impact of hCG administration on patients with various causes of infertility should also be explored. Wider implications of the findings Pre-administration of hCG injection along with a daily dose of 300 mg/day progesterone enhanced clinical pregnancy rates for frozen embryo transfer. This holds the potential for the optimization of hormon supplementation protocols to increase overall success rates in assisted reproductive technologies. Trial registration number not applicable
Read full abstract