Scholarlink

Background <p>Muscle-invasive bladder cancer (MIBC) is an aggressive and heterogeneous malignancy with limited biomarkers to guide patient stratification and response to neoadjuvant chemotherapy (NAC). We sought to define the biological basis and clinical utility of a bladder cancer–associated ten-gene signature for predicting prognosis and therapeutic response.</p> Methods <p>Bulk transcriptomic data from The Cancer Genome Atlas (TCGA) and single-nucleus RNA sequencing (snRNASeq) of primary MIBC tumors were integrated to characterize the cellular context of the ten-gene bladder cancer (BLCA) signature. Machine-learning approaches were applied to qPCR-derived expression data to develop an exploratory model for survival stratification.</p> Results <p>An expression-based direction-weighted BLCA signature score was associated with worse OS, DSS, and PFI in TCGA-BLCA and marked a tumor state characterized by coordinated activation of oncogenic programs including EMT, cell-cycle, and MAPK/PI3K signaling. snRNASeq revealed enrichment of signature activity within the stromal compartment, marked by expansion of FAP<sup>+</sup> /PDPN<sup>+</sup> fibroblasts and MHCII<sup>+</sup> macrophages with endothelial remodeling. High-signature tumors exhibited an immune-dysregulated microenvironment characterized by increased immune checkpoint expression (CD274, HAVCR2, TIGIT) and reduced antigen-presentation machinery. This stromal-dominant phenotype was particularly associated with Luminal MIBC. Functionally, high expression of key signature components (e.g. SDC1, VEGFA, CXCL8) predicts a distinct, targetable vulnerability to MEK inhibitors in drug sensitivity databases.</p> Conclusions <p>Translating these findings, qPCR-derived expression of the ten-gene panel was used to build an exploratory machine-learning model for survival stratification in the NAC-treated MIBC cohort. The selected StepCox (forward) + RSF model showed moderate internal discrimination after bootstrap optimism correction (optimism-corrected C-index = 0.695, 95% CI 0.628–0.792), supporting the feasibility of qPCR-based risk modeling while highlighting the need for external validation.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A bladder cancer–associated gene signature for exploratory qPCR-based survival stratification in NAC-Treated muscle-invasive bladder cancer

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月21日

查看全文
  • Xingyu Chen,
  • Toru Sakatani,
  • Sunao Tanaka,
  • Alberto Contreras-Sanz,
  • Peter Black,
  • Charles J. Rosser,
  • Hideki Furuya

摘要

Background

Muscle-invasive bladder cancer (MIBC) is an aggressive and heterogeneous malignancy with limited biomarkers to guide patient stratification and response to neoadjuvant chemotherapy (NAC). We sought to define the biological basis and clinical utility of a bladder cancer–associated ten-gene signature for predicting prognosis and therapeutic response.

Methods

Bulk transcriptomic data from The Cancer Genome Atlas (TCGA) and single-nucleus RNA sequencing (snRNASeq) of primary MIBC tumors were integrated to characterize the cellular context of the ten-gene bladder cancer (BLCA) signature. Machine-learning approaches were applied to qPCR-derived expression data to develop an exploratory model for survival stratification.

Results

An expression-based direction-weighted BLCA signature score was associated with worse OS, DSS, and PFI in TCGA-BLCA and marked a tumor state characterized by coordinated activation of oncogenic programs including EMT, cell-cycle, and MAPK/PI3K signaling. snRNASeq revealed enrichment of signature activity within the stromal compartment, marked by expansion of FAP+ /PDPN+ fibroblasts and MHCII+ macrophages with endothelial remodeling. High-signature tumors exhibited an immune-dysregulated microenvironment characterized by increased immune checkpoint expression (CD274, HAVCR2, TIGIT) and reduced antigen-presentation machinery. This stromal-dominant phenotype was particularly associated with Luminal MIBC. Functionally, high expression of key signature components (e.g. SDC1, VEGFA, CXCL8) predicts a distinct, targetable vulnerability to MEK inhibitors in drug sensitivity databases.

Conclusions

Translating these findings, qPCR-derived expression of the ten-gene panel was used to build an exploratory machine-learning model for survival stratification in the NAC-treated MIBC cohort. The selected StepCox (forward) + RSF model showed moderate internal discrimination after bootstrap optimism correction (optimism-corrected C-index = 0.695, 95% CI 0.628–0.792), supporting the feasibility of qPCR-based risk modeling while highlighting the need for external validation.