NATURAL PROLIFERATIVE PHASE FROZEN EMBRYO TRANSFERS (NPP-FET) – A NEW APPROACH THAT FACILITATES SCHEDULING WITHOUT HINDERING PREGNANCY OUTCOMES

Abstract

Despite increasing evidence pointing towards a potential benefit in performing natural cycle frozen embryo transfers (NC-FET), multiple centers still prefer artificial cycles (AC-FET) to avoid difficulties in cycle monitoring and scheduling. Specifically, conventional NC-FET frequently require consecutive patient visits to determine the ideal FET timing according to ovulation (either following a spontaneous LH surge or exogenous hCG triggering of an adequately-sized dominant pre-ovulatory follicle). This study assessed progesterone-programmed FET (PP-FET), a new approach in regularly cycling women in which vaginal progesterone for luteal phase support is initiated in the late-follicular phase as soon as an endometrial thickness of at least 7 mm is reached after endogenous estradiol stimulation and without LH surge monitoring and/or hCG triggering. This was a retrospective analysis of FETs performed in our center between 2010-2019 using either autologous of donated oocytes. All conventional NC-FET or cycles in which the endometrial thickness on the day of planning was below 7 mm were excluded. Our main outcome measure was livebirth. We performed multivariable generalized estimating equations (GEE) regression analysis to account for the following potential confounding variables: female age, oocyte source (autologous versus donated), embryo developmental stage (cleavage versus blastocyst) and number of embryos transferred. Finally, we also performed a subgroup analysis according to oocyte source. When compared to AC-FET (n=1155), the PP-FET (n=79) subgroup presented, on average, a higher female age (31.8±5.8 vs 33.2±5.5, p=0.04), performed less oocyte donation (36.2% vs 21.5%, p=0.01) and blastocyst stage (83.8% vs 69.6%, p<0.01) cycles, while having more single embryo transfers (26.7% vs 45.6%, p<0.01) performed. Despite, at first glance, having more unfavorable baseline characteristics, the livebirth rates were higher in the PP-FET group (32.7% vs 46.2%, p=0.02), an association which remained statistically significant even following confounder adjustment in the multivariable GEE regression model (adjusted odd-ratio 1.79, 95% CI 1.11-2.87). This tendency for better livebirth rates in PP-FET was also found in the subgroup analysis according to whether the oocyte used were autologous (30.1% vs 42.6%, respectively, p=0.04) or donated (37.3% vs 58.8%, respectively, p=0.07). PP-FET were associated with better livebirth outcomes.