Abstract
Neuroendocrine bladder carcinoma (NEBC) is an aggressive and therapy-resistant cancer with poor prognosis. Although lineage plasticity drives urothelial-to-neuroendocrine transdifferentiation, intermediate states remain poorly characterized. Through transcriptomic profiling of public datasets, we define a lineage plasticity-associated signature in bladder cancer. We analyze 64,756 in-house and 201,720 public single-cell transcriptomes from NEBC and urothelial carcinoma (UC), identifying TPX2(high) bladder cancer cell subpopulation with increased lineage plasticity, supported by multi-cohort histological validation. TPX2(high) cells lack canonical neuroendocrine markers and are associated with adverse clinical outcomes. TPX2(high) cells exhibit dysregulated cell cycle progression and occupy a suppressive niche associated with CD8(+) T cell exhaustion. In vitro and in vivo, TPX2 promotes lineage plasticity and induces T cell exhaustion. In preclinical patient-derived organoids (PDOs) and patient-derived xenograft (PDX) models, CDK4/6 inhibition plus immune checkpoint blockade demonstrates potent antitumor efficacy. Overall, our findings suggest this combination therapy as a promising therapeutic strategy for TPX2(high) and NEBC patients.