Bladder cancer (BC) is a serious and common malignant tumor of the urinary system. Accurate and convenient diagnosis and treatment of BC is a major challenge for the medical community. Due to the limited medical resources, the existing diagnosis and treatment protocols for BC without the assistance of artificial intelligence (AI) still have certain shortcomings. In recent years, with the development of AI technologies such as deep learning and machine learning, the maturity of AI has made it more and more applied to the medical field, including improving the speed and accuracy of BC diagnosis and providing more powerful treatment options and recommendations related to prognosis. Advances in medical imaging technology and molecular-level research have also contributed to the further development of such AI applications. However, due to differences in the sources of training information and algorithm design issues, there is still room for improvement in terms of accuracy and transparency for the broader use of AI in clinical practice. With the popularization of digitization of clinical information and the proposal of new algorithms, artificial intelligence is expected to learn more effectively and analyze similar cases more accurately and reliably, promoting the development of precision medicine, reducing resource consumption, and speeding up diagnosis and treatment. This review focuses on the application of artificial intelligence in the diagnosis and treatment of BC, points out some of the challenges it faces, and looks forward to its future development.

Background: Bladder cancer (BC) threatens the health of human beings worldwide because of its high recurrence rate and mortality. As an actionable biomarker, fibroblast growth factor receptor 3 (FGFR3) alterations have been revealed as a vital biomarker and associated with favorable outcomes in BC. However, the comprehensive relationship between the FGFR3 alteration associated gene expression profile and the prognosis of BC remains ambiguous.Materials and Methods: Genomic alteration profile, gene expression data, and related clinical information of BC patients were downloaded from The Cancer Genomics database (TCGA), as a training cohort. Subsequently, the Weighted Gene Co-expression Network Analysis (WGCNA) was conducted to identify the hub modules correlated with FGFR3 alteration. The univariate, multivariate, and least absolute shrinkage and selection operator (LASSO) Cox regression analyses were used to obtain an FGFR3 alteration-related gene (FARG) prognostic signature and FARG-based nomogram. The receiver operating characteristic (ROC) curve analysis was used for evaluation of the ability of prognosis prediction. The FARG signature was validated in four independent datasets, namely, GSE13507, GSE31684, GSE32548, and GSE48075, from Gene Expression Omnibus (GEO). Then, clinical feature association analysis, functional enrichment, genomic alteration enrichment, and tumor environment analysis were conducted to reveal differential clinical and molecular characterizations in different risk groups. Lastly, the treatment response was evaluated in the immunotherapy-related dataset of the IMvigor210 cohort and the frontline chemotherapy dataset of GSE48276, and the chemo-drug sensitivity was estimated via Genomics of Drug Sensitivity in Cancer (GDSC).Results: There were a total of eleven genes (CERCAM, TPST1, OSBPL10, EMP1, CYTH3, NCRNA00201, PCDH10, GAP43, COLQ, DGKB, and SETBP1) identified in the FARG signature, which divided BC patients from the TCGA cohort into high- and low-risk groups. The Kaplan–Meier curve analysis demonstrated that BC patients in the low-risk group have superior overall survival (OS) than those in the high-risk group (median OS: 27.06 months vs. 104.65 months, p < 0.0001). Moreover, the FARG signature not only showed a good performance in prognosis prediction, but also could distinguish patients with different neoplasm disease stages, notably whether patients presented with muscle invasive phenotype. Compared to clinicopathological features, the FARG signature was found to be the only independent prognostic factor, and subsequently, a FARG-based prognostic nomogram was constructed with better ability of prognosis prediction, indicated by area under ROC curve (AUC) values for 1-, 3-, and 5-year OS of 0.69, 0.71, and 0.79, respectively. Underlying the FARG signature, multiple kinds of metabolism- and immune-related signaling pathways were enriched. Genomic alteration enrichment further identified that FGFR3 alterations, especially c.746C>G (p.Ser249Cys), were more prevalent in the low-risk group. Additionally, FARG score was positively correlated with ESTIMATE and TIDE scores, and the low-risk group had abundant enrichment of plasma B cells, CD8+ T cells, CD4+ naive T cells, and helper follicular T cells, implying that patients in the low-risk group were likely to make significant responses to immunotherapy, which was further supported by the analysis in the IMvigor210 cohort as there was a significantly higher response rate among patients with lower FARG scores. The analysis of the GDSC database finally demonstrated that low-risk samples were more sensitive to methotrexate and tipifarnib, whereas those in the high-risk group had higher sensitivities in cisplatin, docetaxel, and paclitaxel, instead.Conclusion: The novel established FARG signature based on a comprehensive FGFR3 alteration-related transcriptomic profile performed well in prognosis prediction and was also correlated with immunotherapy and chemotherapy treatment responses, which had great potential in future clinical applications.

Adenocarcinoma of the urinary bladder is a rare malignancy. Radical surgery is suggested as the best available treatment for early-stage disease, but there is currently no consensus on standard chemotherapy regimen for advanced stage. We assessed the feasibility and effect of neoadjuvant chemotherapy with gemcitabine and cisplatin (GC) plus S-1 for patients with locally advanced primary adenocarcinomas of the urinary bladder.