Does the urinary microbiome reflect the bladder-cancer-associated microbiome? Characterizing the microbiome in urine and cancer tissue in bladder cancer.

introductionWe sought to characterize the microbiome in bladder cancer tissue samples and to compare it with the microbiome in simultaneously collected urine.

methodsBladder cancer tissue and transurethral urine specimens were collected simultaneously from consecutive patients with bladder cancer at the time of transurethral resection of bladder tumor (TURBT) or radical cystectomy (RC). Samples were analyzed using 16S rRNA sequencing. Urinary and tissue microbiome were compared. Overlaps among bladder cancer tissue and respective urine samples were described. Microbiome was further described in terms of alpha (diversity within a sample measured by Observed, Chao1, Shannon, and Simpson indices), beta diversities (diversity among different samples measured by Bray Curtis Diversity index) and differential abundance of bacteria at the genus level.

resultsTwenty-one patients were included in the study (15 males and 6 females). Transurethral urine samples were available for all but 3 patients, where voided samples were used. Nineteen patients had high grade urothelial carcinoma and 2 had low grade. Looking at the overlapping genera among the urine and tissue samples, only Pseudomonas, Staphylococcus, Acinetobacter, Corynebacterium, Escherichia-Shigella, Anaerococcus, Streptococcus, and Prevotella were present in >75% of both urine and tissue samples. Comparing tissue and urine specimens, there was no significant difference across all alpha diversity indices, while Bray Curtis for beta diversity showed significant dissimilarity (p<0.0001). There was significantly higher abundance of Moraxella, Herbaspirillum, Clostridium sensu stricto 8, Cellulomonas, Pleomorphomonas, Conchiformibius, Prevotella_9, Lachnospiraceae, Marmoricola, Pseudoglutamicbacter, Helicobacter, Jeotgalicoccus, Roseburia, Granulicatella, Lachnoclostridium, Odoribacter, Dermabacter, Akkermansia, Abiotrophia, and Reinbacterium in the urine samples. On the other hand, there was significantly higher abundance of Conexibacter, Cnuella, Mobilitalea, Fulvimonas, Pedomicrobium, Pectobacterium, Weissella, Selenomonas, Tannerella, Aliterella, Xanthobacter, Sporosarcina, Gordonia, Bosea, Pantoea, SM1A02, Vibrio, Pediococcus, Lacticaseibacillus and Blastococcus in the tissue specimens.

conclusionIn this cohort, bladder-cancer tissue associated microbiome exhibited a distinct microbial signature when compared to urine. These results suggest that the urinary microbiome may not provide an accurate representation of the bladder-cancer associated microbiome. Validation in larger, standardized cohorts with contamination control is warranted.