Abstract
Bladder cancer, specifically urothelial carcinoma originating from the bladder lining, encompasses two distinct pathologies: non-muscle invasive bladder cancer (NMIBC), and muscle invasive bladder cancer (MIBC). Approximately 75% of patients present with NMIBC. Due to bladder cancers proximity to urine, research has focused on developing non-invasive tests that assay nucleic acids, proteins or cells found in urine to diagnose bladder cancer. Currently, MIBC detection routinely reaches >95% sensitivity. However, NMIBC, especially low-grade NMIBC, struggles to achieve >70% sensitivity limiting the clinical utility of these non-invasive tests.
The research presented herein aimed to characterise well-known urinary biomarkers to better understand the reduced sensitivity observed in NMIBC detection. To achieve this, we utilised in situ hybridisation and immunocytochemistry on cells isolated from bladder cancer patient urine and FFPE tumour tissue to explore the frequency and locality of these biomarkers. Furthermore, we adapted in situ hybridisation for potential development into a stand-alone diagnostic test.
We next investigated how various cell types in urine influence bladder cancer biomarker signals. To achieve this, we used single-cell RNA-sequencing (scRNAseq). This required in-house development of a protocol to isolate, cryopreserve, and store patient urine cells from a distant site without compromising RNA or cell viability. Having successfully sequenced three patient samples, we identified lymphocytes, monocytes, macrophages, urothelial, renal, squamous, and progenitor cells in all samples assayed. In addition, by profiling urothelial cell clusters against known biomarkers of NIMBC and MIBC, we identified putative cancer cells. Using differentially expressed genes originating from these cells, we generated two pipelines to identify both over-expressed and highly specific genes that enabled the discrimination between cancer cells and all other cell types present in the dataset. Finally, a shortlist of identified biomarkers was benchmarked against gene expression data from 406 MIBC samples, identifying multiple candidate biomarkers for future validation.