OBJECTIVE: To evaluate the uterine microbiome among women with endometrial polyps and submucosal fibroids and to compare results between endometrial sampling techniques. DESIGN: Patients with polyps or fibroids were prospectively recruited before hysteroscopy, whereas patients undergoing retrieval for planned oocyte cryopreservation were recruited prospectively as controls. Three specimen types obtained for each patient were the distal 5 mm of an embryo catheter passed to the uterine fundus (C), endometrial tissue from an endometrial biopsy (T), and formalin-fixed paraffin-embedded (FFPE) endometrial tissue from the same endometrial biopsy. 16S ribosomal RNA gene amplicon sequencing was performed to analyze the structure of the endometrial microbiome. SUBJECTS: Thirty-seven participants including 28 women with polyps and/or fibroids and 9 controls. EXPOSURE: None. MAIN OUTCOME MEASURES: Microbial taxonomic alpha and beta diversity
differential abundance of taxa. RESULTS: Across all sample types, participants with polyps had higher microbial alpha diversity than controls (4.3 vs. 5.1, q = 0.049), and microbial communities were significantly different (pairwise Permutational Multivariate Analysis of Variance (PERMANOVA) pseudo-F = 2.1, q = 0.003). These differences were observed when examining C specimens alone (5.4 vs. 6.4, q = 0.001
pairwise PERMANOVA pseudo-F = 2.5, q = 0.003), although they did not reach significance when examining either T or FFPE specimens alone. Participants with fibroids had similar alpha diversity yet significant differences in beta diversity compared with controls in analyses combining all specimens (pairwise PERMANOVA pseudo-F = 1.475, q = 0.030)
however, these differences did not achieve significance when analyzing C, T, or FFPE specimens alone. When comparing C and T specimens vs. FFPE specimens overall, alpha diversity was significantly higher (q <
0.001 and q <
0.001, respectively) and there were significant differences in beta diversity (q <
0.003 and q <
0.003, respectively). Analyses of C specimens generated a larger number of significantly differentially abundant taxa compared with other sampling methods. Although not statistically significant, relative abundance of putative pathogens was higher in participants with polyps than controls regardless of sampling technique. CONCLUSIONS: Results of this exploratory study suggest that significant microbial differences exist among patients with endometrial polyps vs. healthy controls. However, results varied by sampling technique, highlighting a need to identify optimal sampling methods before validating findings in larger prospective cohort studies.