O-241 Elevated serum progesterone levels before frozen embryo transfer do not negatively impact reproductive outcomes: a large retrospective cohort study

Abstract

Abstract Study question Do patients with high serum progesterone levels before frozen embryo transfer (FET) under hormonal replacement therapy (HRT) present worse reproductive outcomes? Summary answer Elevated serum progesterone levels before FET in artificially prepared cycles with vaginal or vaginal plus subcutaneous progesterone do not impair reproductive outcomes. What is known already Low serum progesterone levels before FET do negatively affect reproductive outcomes in terms of live birth rate. However, there is not robust data regarding the impact of high serum progesterone levels in the luteal phase of patients who undergo HRT for FET. Study design, size, duration Retrospective cohort study of 3183 blastocyst FET cycles under HRT performed in a university-affiliated fertility centre between March 2009 and December 2020. All the cycles presented adequate serum progesterone levels before FET (≥10.6 ng/ml). A total of 1360 cycles corresponded to frozen homologous embryo transfer (ET) (hom-FET), 1024 were euploid ET (eu-FET) after preimplantational genetic testing for aneuploidies (PGT-A), and 799 cycles were frozen heterologous ET (het-FET). The primary objective was live birth rate (LBR). Participants/materials, setting, methods Standard HRT was used. Luteal phase was covered with vaginal progesterone 200 mg/8h, or vaginal plus a daily subcutaneous injection of progesterone (25 mg). Serum progesterone levels were measured the day before FET. Elevated progesterone levels were considered in the 90th and 95th centiles. A generalized additive model (GAM) was performed to study the functional relationships between progesterone and LBR. A multivariable logistic regression was used to evaluate the effect of high progesterone over LBR. Main results and the role of chance Mean serum progesterone level before FET was 16.77±8.43 ng/ml. Progesterone levels were significantly higher in the group under vaginal plus subcutaneous progesterone (21.87±14.17 vs. 15.56±5.72, p < 0.001). No differences in clinical pregnancy, miscarriage and LBR were found according to the use of vaginal or vaginal plus subcutaneous progesterone for each of the groups (hom-FET, eu-FET and het-FET). Live birth rates were comparable among patients in the highest centile of serum progesterone levels (≥p90, ≥22.33 ng/ml) and the rest of patients (p < 90) (43.9 vs 41.3%; p = 0.381). Patients with progesterone levels ≥p90 presented lower BMI compared to those in the lower centiles (<p90) (22.62±3.82 vs. 23.32±4.06; P = 0.009). After dividing patients in deciles according to serum progesterone levels before, no differences in LBR were observed among groups (P = 0.938). No association was observed with GAM model between progesterone levels and LBR. A multivariable logistic regression adjusted by oocyte age, type of treatment and number of embryos transferred was applied for centile 90 and centile 95 of progesterone, and showed that serum progesterone in their highest levels did not negatively impact LBR. Limitations, reasons for caution The main limitation of this study is its retrospective design. The results only apply for patients under HRT with vaginal micronized progesterone alone or plus subcutaneous progesterone. Progesterone determination was measured before blastocyst FET. Extrapolation to other HRT protocols or timings of progesterone measurement needs to be validated. Wider implications of the findings The results of this study suggest that once a threshold of serum progesterone before FET is achieved, progesterone levels are not predictive of the clinical outcome. Actually, LBR are not negatively affected when progesterone levels are found in their highest centiles after luteal phase rescue with vaginal plus subcutaneous progesterone. Trial registration number Not applicable