Qi Zhang, Zhipeng Wen, Ziyang Yang, Xuezhi Long, Jieshen Xian, Yueting Huang, Yufei Liu, Yubo Wang, Jiahao Cheng, Gengjia Chen, Xuekun Zhang, Tuxiong Huang, Di Gu
Basal UTUC has a largely stage-independent, immune‒stromal-enriched background that is spatially reorganised with muscle invasion, with an SPP1-associated myeloid‒stromal boundary program as its most reproducible feature. Spatial, functional and clinical evidence supports the prognostic value of SPP1 and warrants its evaluation as a therapeutic target.
BACKGROUND: Upper tract urothelial carcinoma (UTUC) is a rare, aggressive malignancy representing 5%‒10% of urothelial carcinomas. Its basal subtype remains understudied but is preferentially associated with muscle invasion and poorer disease-specific survival (DSS). Although the genomic landscape of UTUC has been extensively profiled, current bulk transcriptomic classification cannot resolve cellular heterogeneity or spatial organisation, potentially obscuring subtype-specific microenvironmental programs.
METHODS: We integrated single-cell RNA sequencing and spatial transcriptomics to resolve relationships among molecular subtype, muscle invasion and cellular heterogeneity within the UTUC microenvironment. Cellular communication and multiscale spatial analyses identified spatial programs, with tissue-level validation by multiplex immunofluorescence. These programs were further examined in external bladder cancer datasets spanning single-cell, spatial and GeoMx platforms. SPP1 knockdown in J82 cells tested epithelial-intrinsic effects. We evaluated the subtype-discriminatory and prognostic relevance of SPP1 in the Japan-UTUC cohort.
RESULTS: Basal UTUC exhibited an immune‒stromal-enriched microenvironment primarily associated with molecular subtype rather than pathological stage. Within basal tumours, muscle invasion coincided with increased heterogeneity and two spatial programs. A conserved SPP1+ tumour-associated macrophage‒FAP+ myofibroblast boundary program was associated with myeloid immunosuppression and extracellular matrix remodelling. External datasets provided convergent support for its components and localisation near basal-high epithelium at invasive margins. A more heterogeneous angiogenic program involved VEGFA+ tumour-associated neutrophil‒CXCR4+ tip endothelial interactions. At the epithelial level, SPP1 knockdown reduced J82 cell viability, migration and invasion. SPP1 discriminated basal from luminal tumours (area under the receiver operating characteristic curve (AUC) = .839), and higher expression remained associated with poorer DSS after pathological stage adjustment (hazard ratio = 1.17, 95% confidence interval = 1.01‒1.35, p = .036).
CONCLUSIONS: Basal UTUC has a largely stage-independent, immune‒stromal-enriched background that is spatially reorganised with muscle invasion, with an SPP1-associated myeloid‒stromal boundary program as its most reproducible feature. Spatial, functional and clinical evidence supports the prognostic value of SPP1 and warrants its evaluation as a therapeutic target.