P592 Microbiota concordance between mid-vaginal swabs and both clean- and random-catch urine samples. (14th July 2019)

• Record Type: Journal Article
• Authors: Robinson, Courtney; Holm, Johanna; Brown, Sarah; Ravel, Jacques; Ghanem, Khalil; Brotman, Rebecca
• Is Part Of: Sexually transmitted infections. Volume 95(2019)Supplement 1
• Page Start: A264
• Page End: A264
• Publication Date: 2019-07-14
• Subjects: microbiome -- diagnosis; Sexually transmitted diseases -- Periodicals; HIV infections -- Periodicals; 616.951005
• DOI: 10.1136/sextrans-2019-sti.663
• Languages: English
• ISSNs: 1368-4973

Abstract

Abstract : Background: While urine has successfully been used for STI testing, it has not been routinely used in urogenital microbiota studies. This work explores whether random-catch and/or clean-catch urine could be a proxy for assessing the vaginal microbiota. Methods: In two studies, urinary and vaginal microbiota from women ages 17–45 were compared for (1) 91 participants with paired mid-vaginal swabs and random catch urine samples and (2) 99 participants with paired mid-vaginal swabs and clean catch urine samples. Microbiota composition was characterized by amplicon sequencing of the V3-V4 regions of the 16S rRNA gene. Taxonomic classification was assigned based on SILVA and SpeciateIt. Community State Types (CST) were assigned using an algorithm trained on 13, 000 well-characterized samples. CST I, II, III, and V were dominated by: Lactobacillus crispatus, L. gasseri, L. iners, and L. jensenii, respectively. CST IV-A, IV-B, and IV-C represented low- Lactobacillus states. Similarity of paired urine and vaginal samples was measured at the CST-level by kappa statistics and the population-level with the Yue-Clayton theta indices. Results: We obtained 12 and 7.8 million sequences from urine and vaginal samples, respectively. At the CST-level, random-catch and clean-catch urines were 82.4% and 81.8% concordant with paired mid-vaginal swabs, respectively. Substantial agreement was observed between urine and paired vaginal specimen (Κrandom-catch = 0.770 and Κclean-catch =0.743). At a population-level, average similarity of random- and clean-catch samples to paired vaginal samples indicated a high degree of similarity (θ=0.7496 and 0.7565, respectively). Comparison of the distributions of random-catch and clean-catch θ similarity scores showed no differences (p=0.86). Conclusion: Bacterial compositions of random catch and clean catch urine samples showed substantial agreement to paired mid-vaginal samples assessed by CST- and community-level analyses. Random and clean catch urine samples could potentially be used as a proxy for vaginal microbiota in studies assessing the urogenital microbiota. Disclosure: No significant relationships.