P-418 What is the optimal midluteal phase circulating progesterone level during a fresh IVF cycle? The answer differs from what one expects in a natural cycle

Abstract

Abstract Study question Do midluteal phase circulating progesterone (P4) levels relate to the fresh embryo transfer clinical outcome and is rescue P4 useful in such cases? Summary answer Live birth rate (LBR) was significantly lower when P4 levels were below 20ng/ml and did not seem to benefit from rescue intramuscular P4. What is known already Preceding literature on the predictive value of midluteal P4 for LBR has shown conflicting results and frequently included small sample sets and heterogeneous ovarian stimulation protocols. The findings reported ranged from the lack of an association between P4 serum levels and clinical outcome to worse success rates seen either in the groups with the highest or the lowest P4 in midluteal phase. Furthermore, while rescue P4 has been empirically proposed, it has lacked formal assessment thus far. Study design, size, duration In total, 652 patients who underwent a fresh single blastocyst transfer in an IVF cycle in our center between 2014 and 2021 were included in this retrospective cohort study. All were GnRH-antagonist cycles using hCG trigger and luteal support with micronized vaginal progesterone 200mg twice daily. Rescue intramuscular progesterone was systematically administered whenever low midluteal P4 (below 1 standard deviation of the center’s mean) levels were detected. Participants/materials, setting, methods The cycles were subdivided into three groups according to the midluteal P4 levels: <10ng/ml, 10-20ng/ml and >20ng/ml. The primary outcome assessed was LBR, with secondary outcomes including clinical pregnancy (CPR) and pregnancy loss rates. Maternal age and body mass index (BMI), year of treatment, ovarian response, endometrial thickness, cause of infertility and embryo quality were controlled for using multivariable logistic regression. Main results and the role of chance The mean age ± standard deviation of patients was 36.8±3.8 years, while mean weight and BMI were 63.2±11.5kg and 23.3±3.9kg/m2, respectively. LBR in the group with P4>20ng/ml (46.5%) was significantly higher than in the groups with low and intermediate P4 serum levels (23.1% and 28.1%, respectively, p < 0.001). The same was true for CPR (54.5%, 30.8% and 37.9%, respectively, p < 0.001). Conversely, the rate of pregnancy loss in the group with P4>20ng/ml (23.8%) was significantly lower than in the groups with low and intermediate P4 levels (42.6% and 37.5%, respectively, p = 0.018). The multivariable logistic regression confirmed that, after controlling for the confounding variables described above, P4>20ng/ml remained significantly associated with a higher LBR (OR: 2.285; 95% CI: 1.506-3.469) and a lower pregnancy loss rate (OR: 0.488; 95% CI: 0.273-0.873), irrespective of the use of rescue IM P4. Finally, a multivariable linear regression analysis of the patients baseline characteristics suggested that the only parameter predictive of midluteal P4 levels was the female patients’ weight, with an inverse relationship (adjusted regression coefficient: -0,012; 95% CI: -0,016/-0,009). Limitations, reasons for caution As the pituitary suppression, ovulation trigger and luteal support protocols used in the cycles included in the study were homogeneous (in order to maximize the internal validity of the results), this limits the extrapolibility of the results to other stimulation protocols. Furthermore, we caution that this is a retrospective study. Wider implications of the findings The usefulness of both midluteal P4 serum measurements and rescue P4 should continue to be assessed. The reasons behind the need for higher levels of P4 in a fresh cycle, when compared to natural cycles, must also be investigated. Luteal support may require personalization according to female body weight. Trial registration number Not Applicable