The effect of an immediate frozen embryo transfer following a freeze-all protocol: a retrospective analysis from two centres

Study question: Does the timing of the first frozen embryo transfer (FET) after gonadotropin-releasing hormone (GnRH) agonist triggering with the elective cryopreservation of all embryos affect pregnancy outcome?

Summary answer: FETs performed immediately after a freeze-all cycle did not vary significantly from delayed FETs in terms of pregnancy rates.

What is known already: As interest in, and use of, the freeze-all strategy expands in the field of reproductive medicine, the optimal timing to perform the subsequent FET has become increasingly important. Thus far, all clinical trials evaluating the efficacy of the segmentation strategy have opted to electively defer the first FET for at least one menstrual cycle. However, this mere empirical approach may cause unnecessary distress to infertile patients eager to conceive as soon as possible.

Study design, size and duration: This retrospective cohort study included the first FET cycle of all women who underwent a freeze-all protocol between October 2010 and October 2015 in two reproductive medicine centres (in Belgium and Vietnam, respectively).

Participants/materials, setting and methods: A total of 333 FET cycles were included in the analysis. Following the freeze-all cycle, the preparation of the endometrium consisted of the sequential administration of oestradiol valerate and micronized vaginal progesterone. The start of the FET was classified as either immediate (following the GnRH agonist withdrawal bleeding) or delayed (by at least one menstrual cycle). Clinical pregnancy rate (CPR) was the main outcome of our study.

Main results and the role of chance: Women in the immediate FET group were slightly younger on average (30.9 ± 4.1 versus 31.8 ± 4.3, P = 0.045) on the date of oocyte retrieval. Moreover, women in the immediate FET group received a blastocyst transfer more frequently (53.4% versus 41.6%, P = 0.038) and had fewer embryos transferred on average compared to the delayed FET group (1.8 ± 0.8 versus 2.0 ± 0.8, P = 0.013). CPR/FET was marginally significantly higher in the immediate FET group in our crude analysis (52.9% after immediate FET versus 41.6% after delayed FET, P = 0.046). In order to assess if CPR/FET remained unaltered after adjusting for measured confounding, we performed mixed-effects multivariable regression analysis. Timing of the FET no longer affected significantly the CPR of the first FET in the adjusted analysis (adjusted odds ratio (aOR): 0.62, 95% CI: 0.38-1.00; predicted CPR of 52.5% for immediate FET versus 41.8% for delayed FET).

Limitations, reasons for caution: The results are limited by the retrospective design and the potential for unmeasured confounding. Furthermore, we only evaluated the effect of the FET timing of the first FET on CPRs in artificially supplemented cycles and, thus, the results should not be extrapolated to live birth rates or natural-cycle FETs.

Wider implications of the findings: This study offers a simple but potentially relevant measure to increase patient satisfaction and adherence in couples who seek to become pregnant both safely and as soon as possible.

Study funding/competing interests: No funding was received for this study. The authors have no conflicts of interest to declare.