O-127 The endometrial tissue microbiota of women who did or did not achieve a live birth within 12 months after a first failed IVF/ICSI cycle

Abstract

Abstract Study question After one failed IVF/ICSI cycle, does the endometrial microbiota composition differ between women who will or will not reach a live birth within 12 months? Summary answer The endometrial microbiota composition did not significantly differ in women with one failed IVF/ICSI cycle with or without live birth, but statistical power was low. What is known already Evidence for the presence of an indigenous endometrial microbiome is mounting, and its composition may be associated with implantation success. However, a ‘core’ endometrial microbiome has not yet been defined, and its role in embryo implantation is still poorly understood. Further investigation of this topic may allow improvement and personalisation of clinical care for infertile couples. Endometrial microbiome analysis in infertile women has not yet been performed using transcervically obtained endometrial tissue. Using endometrial tissue instead of swabs or fluid may increase the bacterial DNA yield and therefore the precision of microbiome analyses. Study design, size, duration Endometrial tissue was obtained from a cohort of 141 infertile women undergoing endometrial scratching within a randomised controlled trial (RCT) (SCRaTCH trial, NL5193/NTR5342). Briefly, women aged 18-44 years with failed implantation after one full IVF/ICSI cycle and planning a subsequent IVF/ICSI cycle, were eligible. Participants were followed-up until 12 months after randomisation, with the primary outcome being live birth, defined as the delivery of at least one live foetus after 24 weeks of gestation. Participants/materials, setting, methods Endometrial tissue was obtained with an endometrial biopsy catheter in the midluteal phase of a natural cycle preceding subsequent IVF/ICSI, snap-frozen and stored at -80 °C until use. Total DNA was isolated from these biopsies, followed by 16S rRNA sequencing (V3-V4 region) to determine the endometrial microbiota composition. Positive (mock communities) and negative controls (DNA extraction and PCRs) were included. QIIME2 and DADA2 were used for the data analysis, followed by statistical analysis in R studio. Main results and the role of chance During the 12-month follow-up, 61/141 women (43.3%) reached a live birth. While endometrial microbiota profiles of all 141 women were analysed, only samples with ≥100 reads were included in the analysis, resulting in a total of 46 samples (32.6%) that were included in the analysis, which consisted of samples from 25 women who did not have and 21 women who did have a live birth within 12 months. The median number of reads per sample was not significantly different between the two groups (respectively 2,317 (IQR 651-19,031) and 1,335 (IQR 296-3,180), p = 0.29 by Mann-Whitney test). The endometrial microbiota detected, were bacterial genera frequently reported within the vaginal microbiota (e.g. Lactobacillus, Atopobium and Gardnerella). A clear dominance of Lactobacillus (relative abundance 55-100%, n = 22) or an unclassified bacterium genus (relative abundance 52-76%, n = 18) was observed in the majority of the samples; however, this dominance was not associated with the outcome of live birth. In addition, the samples dominated by Lactobacillus genera were mostly dominated by one species of Lactobacillus each (L. crispatus, L. iners, L. gasseri or L. jensenii). Limitations, reasons for caution The low biomass and the low ratio of bacterial versus human DNA in endometrial tissue were limiting factors in endometrial microbiota analysis. Furthermore, tissue was obtained transcervically, and contamination with vaginal/cervical microbiota could therefore have occurred. In the SCRaTCH trial no vaginal swabs were taken to serve as internal controls. Wider implications of the findings Future endometrial microbiota studies should consider the use of samples with a lower proportion of human DNA to maximize bacterial DNA yield. Furthermore, for endometrial microbiota research, sampling devices avoiding cervicovaginal contamination are desirable and may be developed in the future. Trial registration number SCRaTCH trial, NL5193/NTR5342