Related Experiment Video
Updated: Apr 16, 2026

Mining Spatial Transcriptomics Datasets using DeepSpaceDB
Published on: September 5, 2025
Spatial transcriptomics of en bloc transurethral resection of bladder tumor specimens defines an invasive-front gene
Kentaro Yoshihara1,2, Fumihiko Urabe3,4, Miyaka Umemori5
1Laboratory of Integrative Oncology, National Cancer Center Research Institute, Tokyo, Japan.
Abstract:
Bladder cancer is classified by depth of invasion into muscle-invasive bladder cancer (MIBC) and non-muscle-invasive bladder cancer (NMIBC). Because some NMIBC cases progress to MIBC, it is important to elucidate the mechanisms of invasion. Spatial transcriptomics is a powerful tool for characterizing the tumor microenvironment. We applied Visium to seven high-grade NMIBC tumors using en bloc transurethral resection of bladder tumor (ERBT) specimens that preserve three-dimensional tumor architecture. Malignant regions were divided into the tumor-upper and tumor-boundary regions, and seven cases were further stratified by pathological subgroup based on the muscularis mucosae. We identified 15 genes with boundary-enriched expression in the higher-risk subgroup. These genes were enriched for epithelial-mesenchymal transition and myogenesis features. To assess the generalizability of this finding, we analyzed public bulk RNA-seq datasets and found that higher expression of the 15-gene signature was associated with worse overall survival, suggesting potential utility for prognostic stratification and therapeutic decision-making.

