Frozen-thawed embryo transfers in natural cycles with spontaneous or induced ovulation: the search for the best protocol continues

STUDY QUESTION: Is natural cycle frozen-thawed embryo transfer (NC-FET) associated with better clinical pregnancy rates (CPR) when compared to modified natural cycle frozen-thawed embryo transfer (mNC-FET)? SUMMARY ANSWER: NC-FET is associated with a higher CPR compared to mNC-FET. WHAT IS KNOWN ALREADY: There is conflicting evidence regarding the impact of hCG triggering on clinical outcomes after frozen-thawed embryo transfer (FET), and information on the effect of luteal phase support (LPS) is lacking. STUDY DESIGN, SIZE, DURATION: This retrospective study included all (n = 2353) consecutive cycles with FET of vitrified cleavage and blastocyst stage embryos warmed between January 2010 and April 2015 in a tertiary centre. The FET cycles were grouped by type as follows: NC (n = 501), NC + LPS (n = 828) or mNC + LPS (n = 1024). Artificial cycles were excluded from the analysis. PARTICIPANTS/MATERIALS, SETTING, METHODS: We performed mixed-effect multilevel multivariable regression analysis to account for the clustering of FETs using embryos derived from the same patient and/or ovarian stimulation cycle. Adjustment for the following potential confounders was also performed: female age at oocyte retrieval, number of oocytes retrieved, fresh cycle pregnancy outcome, embryo transfer rank, number of embryos transferred, embryo stage and grade and endometrial thickness. Bonferroni adjustment for multiple comparisons was performed whenever indicated. MAIN RESULTS AND THE ROLE OF CHANCE: The unadjusted CPR per cycle was significantly higher in the NC-FET group (46.9%) when compared with the mNC-FET + LPS groups (29.7%, P < 0.001) but not the NC-FET + LPS group (39.9%, P = 0.069). The lower clinical performance of mNC-FET + LPS remained significant even after adjusting for potential confounders [adjusted odds ratio (95% CI) compared to the NC-FET groups: 2.18 (1.64-2.90) and 1.67 (1.31-2.12) for the NC-FET and NC-FET + LPS groups, respectively]. A sensitivity analysis restricting the sample only to the first FET performed by the couple in our centre was also performed. The predicted CPR in this multivariable logistic regression model remained significantly higher in the NC-FET (53.9%) and NC-FET + LPS (44.9%) groups when compared to mNC-FET + LPS (34.2%, all Bonferroni-adjusted pairwise comparisons with P = 0.01). LIMITATIONS, REASONS FOR CAUTION: The interpretation of the findings of this study is limited by the retrospective nature of the analysis and the potential for unmeasured confounding. WIDER IMPLICATIONS OF THE FINDINGS: The use of mNC-FET, using hCG triggering or progesterone supplementation, should be reconsidered in light of the potential negative effect on pregnancy outcome.