Risk factors for the development of endometrial fluid in women undergoing IVF: A retrospective cohort study✰.

RESEARCH QUESTION: Presence of endometrial fluid (EF) is a poorly understood pathology and remains a challenge for clinicians, as very little data exists to explain its consequences and treatment. Our objective was to investigate risk factors for EF during IVF. DESIGN: This retrospective cohort study included all women with a freeze all embryos cycle (FAE) for EF between 2010 and 2016 at a university-affiliated private IVF center. Controls (2:1) were randomly selected out of the database of our fresh autologous IVF cycles during the same period. Main outcome measures were possible risk factors for EF, comprising polycystic ovarian syndrome (PCOS), ovarian hyperstimulation syndrome (OHSS), previous pelvic or endometrial surgery (polypectomy or synechia removal), cesarean section, myomas and severe endometriosis. A logistic regression model was used to assess independent risk factors for EF. RESULTS: Out of 9000 IVF cycles, 1204 were FAE cycles, among which we identified 86 EF cases. We then selected 171 controls. Independent risk factors for presence of EF were a history of previous myomectomy (adjusted odds ratio (aOR) 19.77, 95%CI [4.01-97.53]), severe endometriosis (aOR 5.97, 95%CI [2.09-17.05]), PCOS (aOR 5.72, 95%CI [2.66-12.33]) and previous cesarean section (aOR 5.17, 95%CI [1.84-14.49]). CONCLUSIONS: Our results are not only confirming the association between PCOS, severe endometriosis, previous cesarean procedure and EF, but also reporting for the first time an association between previous myomectomy and EF.

In freeze-all embryo cycles due to endometrial fluid (EF), live birth rates are comparable to those of controls, despite high rates of EF recurrence and cycle cancellation.

RESEARCH QUESTION: Do cumulative live birth rates (CLBRs) differ between women who have had a freeze-all embryo cycle (FAE) for endometrial fluid (EF) and controls? DESIGN: This retrospective cohort study included 83 women who had a FAE cycle due to the presence of EF between 2010 and 2016 at a university-affiliated private IVF center. The controls were 219 women who had FAE for other indications during the same period and were randomly selected. The main outcome measures were CLBRs, EF recurrence, cancellation and pregnancy loss rates. RESULTS: Population characteristics were comparable between the two groups. The CLBR was not significantly different between the EF and the control group: 39.8 % vs. 47.0 %, respectively, p=0.26. Cancellation rates in the two first FETs were higher in the EF group than the control group: 18.1 % vs. 4.1 % (p<0.001) and 22.9 % vs. 8.5 % (p=0.02). After FAE for EF, we observed a significant risk of EF recurrence (32/177 cycles, 18.1 %), allowing us to identify a poor prognosis subgroup. When EF was detected, the LBR per transfer was 7.1 % (1/14) when the transfer was finally performed (after EF aspiration or EF disappearance), compared to 25 % (32/128) in cycles without EF recurrence (p<0.05). Conversely, in the absence of EF recurrence (145/177, 81.9 %), the LBR was comparable to that of the control group. The type of endometrial preparation does not seem to be associated with EF recurrence. CONCLUSION: Despite higher rates of EF recurrence and cycle cancellation, women with FAE for EF ultimately have comparable LBRs to those who have had a FAE for other indications. However, women presenting with at least one EF recurrence during FETs seem to have a lower LBR.