Using 16S rRNA sequencing, we observed that urine and tissue exhibit distinct community structures. This was particularly seen in beta-diversity analysis, which demonstrated significant similarity (p<0.0001). Within same-patient samples, there was substantial divergence between tissue and urine bacterial populations. Overlap analysis identified a limited set of genera present in >75% of both sample types, whereas differential abundance testing identified multiple taxa preferentially enriched in tissue versus urine and vice versa. These results support that urine does not reliably reflect tissue-associated microbial composition in bladder cancer.
The principal implication is that microbiome investigations may benefit from prioritizing tissue-based sampling profiles rather than making the assumption that urine is a valid proxy. Since we observed differences between FFPE and frozen tissue profiles, it is clear that tissue preservation and standardized contamination prevention protocols should be implemented in future work in this field. Larger cohort studies with higher resolution sequencing platforms may be used to more accurately identify the discordance between and unique species associated with tissue vs urine.
Written by: Ahmed A. Hussein, Yakov Klugman, Jordan Carlson, Tariq A. Bhat, Zhe Jing, Eduardo Cortes Gomez, Prashant K. Singh, Jianmin Wang, Justine Jacobi, Gary Smith, David Goodrich, Khurshid A. Guru
- Roswell Park Comprehensive Cancer Center, Buffalo, NY.