Evaluating the use of biobanked urine specimens for human urobiome studies

Case-control studies focused on the urinary tract microbiome, or urobiome, have consistently reported significant associations with disease. However, clinical urobiome studies have typically been small, averaging ~50 patients per study. While these sample sizes are sufficient to detect large effect sizes, they have not been able to differentiate disease phenotypes within a larger disease complex (e.g. different types of kidney stones), which have unique etiological origins. Biobanked urine specimens can help to fill this void. However, since these specimens were not collected specifically for urobiome studies, they must be validated before drawing any strong conclusions. The objective of this study was to evaluate microbiome data derived from metagenomic analysis of biobanked urine specimens against the following criteria: 1) Level of contaminants; 2) Retention of high-quality DNA; 3) Overgrowth of a few dominant bacteria; and 4) Preservation of sex-specific taxa. A total of 171 samples were assessed from biobanked or freshly collected specimens, in addition to multiple positive and negative controls. While there were significant differences in diversity (alpha/beta; p<0.001) based on whether or not samples were biobanked, these differences can largely be explained by study-specific variation. With these criteria, we find that biobanked urine specimens provide similar data to fresh specimens collected using standardized protocols and can be used for clinical urobiome studies.