L26/P-780 Natural cycles mitigate the adverse effects of thin endometrium on clinical outcomes following a frozen embryo transfer: A study of 23,221 single-blastocyst transfer cycles

Study question Does the association between endometrial thickness and clinical outcomes differ across three common endometrial preparation protocols in single-blastocyst frozen-thawed embryo transfer (FET) cycles? Summary answer Natural cycles (NC) appear to mitigate the adverse effects of thin endometrium on pregnancy and perinatal outcomes compared to hormone replacement therapy (HRT) -based protocols. What is known already Endometrial thickness is an established predictor of clinical outcomes in FET. Evidence suggests that the endometrial transcriptomic profiles of infertile patients during natural cycle FET demonstrate greater similarity to those of fertile controls than during hormone replacement therapy cycles. Nevertheless, it remains unclear whether the “optimal” EMT threshold remains consistent across different endometrial preparation protocols. Study design, size, duration This retrospective cohort study included 23,221 single-blastocyst transer cycles from 17,416 women at a single center between 2016 and August 2024. Endometrial preparation protocols included NC (1853 cycles), HRT (18086 cycles), or gonadotropin-releasing hormone agonist (GnRHa)-pretreated HRT (GnRHa+HRT; 3275 cycles). Participants/materials, setting, methods Generalized additive models were fitted to visualize EMT-outcome associations across the total cohort and specific cycles. EMT was stratified into five categories: ≤ 7, (7, 8], (8, 10], (10, 12], and >12 mm. Statistical associations were assessed using stabilized inverse probability of treatment weighting-weighted generalized linear mixed models (GLMMs), incorporating random intercepts to account for repeated cycles per patient. Primary outcomes included live birth and good perinatal outcome (GPO: singleton, ≥ 37 weeks, ≥ 2500g). Main results and the role of chance Substantial declines in LBRs were observed in HRT cycles at ≤ 7 mm (25.35%; aOR 0.40, 95% CI 0.33–0.48) and in GnRHa+HRT cycles at both ≤7 mm (18.18%; aOR 0.44, 95% CI 0.30–0.65) and (7, 8] mm (29.39%; aOR 0.70, 95% CI 0.53–0.93). A similar trend was manifested for GPO. Conversely, in NC cycles, an EMT ≤7 mm did not significantly impair LBR (aOR 0.88, 95% CI 0.54-1.46) or GPO (aOR 0.90, 95% CI 0.54-1.51), suggesting better resilience to thin endometrium. While clinical outcomes typically improved with increasing EMT across all strategies, the effect size was protocol-dependent. In GnRHa+HRT cycles, incremental increases in EMT up to 12 mm significantly enhanced both LBR and GPO. Conversely, NC cycles exhibited a minimal response to thickness variations; a significant benefit was observed only when reached (10, 12] mm, yielding peak odds for both live birth (aOR 1.37, 95% CI 1.08–1.74) and GPO (aOR 1.48, 95% CI 1.16–1.89). No detrimental effects were observed for EMT >12 mm. Sensitivity analyses controlling for oocyte retrieval age instead of age at transfer/pregnancy produced similar conclusions. Limitations, reasons for caution The retrospective design is a limitation, where unmeasured confounding cannot be entirely excluded. Wider implications of the findings The strategy for single-blastocyst transfer should be individualized based on the specific preparation protocol and EMT profile. For patients with refractory EMT < 7 mm, NC protocols appear to be a more effective management strategy, as they may mitigate the adverse effects of reduced thickness observed in HRT cycles. Trial registration number No