Summary Current transcriptomic classification systems for bladder cancer do not consider the level of intra-tumor subtype heterogeneity. Here we present an investigation of the extent and possible clinical impact of intra-tumor heterogeneity across early and more advanced disease stages of bladder cancer. We performed single nucleus RNA-sequencing of 48 bladder tumors and four of these tumors were additionally analyzed using spatial transcriptomics. Total bulk RNA-sequencing and spatial proteomics data were available from the same tumors for comparison, along with detailed clinical follow-up of the patients. We demonstrate that tumors display varying levels of intra-tumor subtype heterogeneity and show that a higher class 2a weight estimated from bulk RNA-sequencing data is associated with worse outcome in patients with molecular high-risk class 2a tumors. Our results indicate that discrete subtype assignments from bulk RNA-sequencing data may lack biological granularity and continuous class scores could improve clinical risk stratification of patients. Highlights Single nucleus RNA-sequencing of tumors from 48 bladder cancer patients. Tumors display varying levels of intra-tumor subtype heterogeneity at single nucleus and bulk tumor level. The level of subtype heterogeneity could be estimated from both single nucleus and bulk RNA-sequencing data with a high concordance between the two. High class 2a weight estimated from bulk RNA-sequencing data is associated with worse outcome in patients with molecular high-risk class 2a tumors.

Single nucleus and spatially resolved intra-tumor subtype heterogeneity in bladder cancer